First Order Draft
1
IPBES Deliverable 3c
DO NOT CITE
Overview and vision Coordinating Lead Authors: Simon Ferrier and Karachepone N. Ninan Lead Authors: Paul Leadley, Rob Alkemade, Grigorii Kolomytsev, Monica Moraes, Yongyut Trisurat and Essam Yassin Mohammed
5
1.1 Introduction
10
15
20
Scenarios and models of biodiversity and ecosystem services play important roles in assessments, policy support and decision making because they help to "better understand and synthesize a broad range of observations; alert decision makers to undesirable future impacts of global changes such as land use change, invasive alien species, overexploitation, climate change and pollution; provide decision support for developing adaptive management strategies; and explore the implications of alternative social-ecological development pathways and policy options. One of the key objectives in using scenarios and models is to move away from the current reactive mode of decision-making in which society responds to the degradation of biodiversity and ecosystem services in an uncoordinated, piecemeal fashion to a proactive mode in which society anticipates change and thereby minimizes adverse impacts and capitalizes on important opportunities through thoughtful adaptation and mitigation strategies" (IPBES/2/16/Add4). The most fundamental message emerging from this assessment is that scenario analysis and modelling can, and should, contribute significantly to achieving the overarching goal of IPBES “to strengthen the science-policy interface for biodiversity and ecosystem services for the conservation and sustainable use of biodiversity, long-term human well-being and sustainable development”.
1.1.1 Background and context 25
30
35
40
Scenarios and models of biodiversity and ecosystems have been a key component of most global, regional and national environmental assessments carried out over the last decade. The IPCC has amply demonstrated the power of scenarios and models as a cornerstone of the science-policy dialog surrounding climate change, but despite the use of scenarios and models of biodiversity and ecosystem services in several major global assessments they have yet to impact on decision making to the extent achieved in the climate domain. Scenarios and models in assessments of biodiversity and ecosystem services have done a good job of alerting the scientific community, natural resource managers and politicians to the possible future risks for biodiversity and ecosystem services. Examples include the Millennium Ecosystem Assessment which called attention to the possibility of greatly increased species extinction risk by 2050 driven primarily by land use and climate change (MA 2005, Alkemade et al. 2009), and the most recent Global Biodiversity Outlook which showed that progress on reducing pressures on biodiversity and ecosystems is currently insufficient to attain most of the Convention on Biological Diversity's Aichi Targets by 2020 (SCBD GBO4 2014, Leadley et al. 2014). Scenarios and modelling of biodiversity and ecosystems services in assessments have done less well than climate change scenarios in being translated into actions at national to global levels to reduce degradation of biodiversity and ecosystem services. Scenarios and models of biodiversity and ecosystems are also increasingly being used in decision contexts outside of global, regional and national environmental assessments. In particular, a wide Page 1 of 35
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
range of policy support methodologies have been developed to allow more direct use of scenarios and models in policy design, implementation and evaluation. The bulk of this work has been done at local scales, but some methodologies are also pertinent at national to global scales. Experience shows that successful application of models and scenarios to policy design, implementation and evaluation requires sustained interactions between decision makers and modellers. The science behind scenarios and models of biodiversity and ecosystem services is young. The first global assessment with substantial component of biodiversity scenarios was the Millennium Ecosystem Assessment released in 2005 (MA 2005). Assessments with significant use of scenarios and models to evaluate ecosystem services are even more recent (e.g., UK NEA 2011). Very rapid progress in the development and use of scenarios and modelling of biodiversity and ecosystem services over the last decade (Fig 1.1) means that IPBES is now well positioned to make substantial use of these methodologies in all of its activities.
Figure 1.1 Change over time in the number of articles published in scientific journals related to scenarios and models of biodiversity. These were compared to the total number of publications on biodiversity over the same period to calculate the percentage of articles related to scenarios and models. The analysis was carried out by the Foundation for Biodiversity Research in 2012 (FRB 2013) using the Web of Science and the following search command: TOPIC = ((projection* or prediction* or forecast* or scenario*) AND ((ecosystem and service*) or (ecological and service*) or “species loss” or biodiversity or (biological diversity) or (species richness) or (species diversity) or (functional diversity) or (biological conservation) or (species conservation) or (habitat conservation) or (genetic resource*) or (genetic diversity) or (plant diversity) or (microbial diversity) or (bacterial diversity) or (fung* diversity) or (weed diversity) or (animal diversity) or (mammal diversity) or (insect* diversity) or (functional trait*) or (virus diversity) or (bird diversity) or (invasive species) or (biological invasion*) or (landscape diversity) or (habitat diversity) or “cultural diversity” or “local knowledge” or “traditional knowledge” or “traditional local knowledge” or “environmental knowledge”))TIMESPAN = From 1975 to 2011.
25 A variety of approaches have been used for developing and presenting scenarios and models in environmental assessments. In some cases, assessment bodies have opted to support the development of a common set of scenarios of direct and indirect drivers, as well as accompanying Page 2 of 35
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
models of impacts on biodiversity and ecosystems. Examples include the global assessments such as the Millennium Ecosystem Assessment (MA 2005), early Global Biodiversity Outlooks (SCBD GBO2 2006), and Global Environmental Outlooks (e.g. UNEP GEO4 2007), as well as some national and regional assessments (UK NEA 2011). At the opposite end of the spectrum some assessments have focused on synthesizing a broad range of published analyses of scenarios and modelling studies available in the literature (e.g., CBD GBO3 2010, Leadley et al. 2010; UNEP GEO5 2012). Still others fall in between these extremes, for example the IPCC relies on a common set of scenarios of direct and indirect drivers developed specifically for the assessment, while assessment of projected impacts on biodiversity and ecosystems is primarily based on analyses of peer-reviewed literature (e.g., IPCC 2014a, b). The advantage of using a common set of scenarios and models is that they provide a clear and homogenous analysis that may be easier for non-specialists to understand: the disadvantages are that these typically are useful for a very limited range of spatial and temporal scales and decision contexts. The advantage of analyses based a broad spectrum of published work is that they provide much greater insight into assumptions underlying scenarios and models and their associated uncertainties and can cover a wide variety of scales and decision contexts, but very diverse assumptions and indicators used in published work make synthesis difficult (Pereira et al. 2010). Several biases need to be addressed to improve the usefulness of scenarios and models for decision makers. Published studies of scenarios and models show a strong bias towards terrestrial ecosystems with climate change as a driver: stronger near term drivers such as habitat loss, invasive species and overexploitation have received insufficient attention. Aquatic ecosystems, especially freshwater ecosystems are under-represented in scenarios and modelling analyses compared to terrestrial ecosystems (FRB 2013). There is also a strong bias towards scenarios in the literature and assessments exploring mid- and end-21st century outcomes (FRB 2013), whereas many decision makers are more focused on nearer terms goals (e.g., Aichi Targets for 2020). Spatial scale ranges in assessments typically cover the range of national to global. Few assessments account for the vast amount of information in sub-national scale scenarios and models, which may be a more pertinent spatial scale for some decision makers. Finally, most assessments have weak use of scenarios and modelling of ecosystem services outside of food production, even though other types of ecosystem services are key elements in decision-making (but see PBL 2012). One of the objectives of this document is to provide guidelines for addressing these biases.
1.1.2 Purpose and scope 35
40
The IPBES Deliverable 3c includes three elements: • A "fast-track assessment of methodologies for scenario analysis and modelling of biodiversity and ecosystem services" (subsequently referred to as the "assessment for scenarios and models") to be completed in 2015. • Creation of a task force following the completion of the assessment in order to promote "methods for the use of different types of knowledge and catalyze the development of databases, geospatial data, and tools and methodologies for scenario analysis and modelling" (IPBES/2/17 Annex IV). In particular, this task force will provide expertise and advice on the use of scenarios and models for other deliverables, and to ensure that
Page 3 of 35
First Order Draft
•
IPBES Deliverable 3c
DO NOT CITE
scenarios and modelling work uses lessons learned and expertise from other deliverables (see 1.1.3 below). The task force is also charged with the creation and maintenance of an "evolving guide" based on the assessment.
5
10
15
20
The methodological assessment on scenarios and models was initiated in order to provide expert advice on “the use of such methodologies in all work under the Platform to ensure the policy relevance of its deliverables” (IPBES/2/17 Annex IV). It is one of the first assessment activities of IPBES because it lays the foundations for the use of scenarios and models in the regional, global and thematic assessments. The primary audiences are the participants in the expert groups and task forces associated with other deliverables of the IPBES Work Programme. For these experts, this methodological assessment provides an overview of scenarios and models, a critical analysis of the types and uses of scenarios and models currently available and perspectives on the development of new methods in the near future. This assessment also addresses several audiences in addition to experts involved in IPBES activities. The overview of scenarios and models, particularly Chapter 1, has been written with nonexperts in mind so that it is accessible for a broad audience including members of the IPBES plenary, stakeholders and policy makers. The critical analysis and perspectives sections of this assessment are more technical in nature and address the broader scientific community in addition to the expert groups and task forces of IPBES. They are written for a scientific audience working on issues related to biodiversity and ecosystem services. Highly technical descriptions and jargon have been kept to a minimum, because this audience has widely varying degrees of familiarity with the full spectrum of scenarios and models.
25
30
35
40
The scope of the assessment covers a broad range of scenarios and models. The objective is to provide guidance for "evaluating alternative policy options using scenarios and models; including multiple drivers in assessments of future impacts; identifying criteria by which the quality of scenarios and models can be evaluated; ensuring comparability of regional and global policies; including input from stakeholders at various levels; implementing capacity-building mechanisms to promote the development, use and interpretation of scenarios and models by a wide range of policymakers and stakeholders; and communicating outcomes of scenario and model analyses to policymakers and other stakeholders" (IPBES/2/16/Add.4). Follow-up work by a task force will start following the completion of this assessment in 2015 and will continue through 2017 and possibly beyond. As such, one of the deliverables of this task force will be an "evolving guide". The exact nature of this remains to be defined, but since methods for scenarios and models are changing very rapidly, it is important that guidance is updated on a regular basis. The task force will also interact with other IPBES deliverables and the broader scientific community to stimulate work on scenarios and models that are specifically targeted at IPBES objectives. It is envisaged that this will be similar to the interactions between the IPCC and the scientific community when developing scenarios and models for climate change assessment.
1.1.3 Links with other IPBES deliverables Page 4 of 35
First Order Draft
5
10
15
20
25
30
35
IPBES Deliverable 3c
DO NOT CITE
This assessment provides guidance on the contributions of scenarios and models to the four functions of IPBES: assessment, capacity building, policy support and knowledge generation. Assessment activities of IPBES focus on the delivery of global, regional and thematic assessments. Key elements of the assessment of scenarios and models have been summarized in a guide for production and integration of assessments (deliverable 2a). The guide furnishes a brief overview of important concepts and highlights the potential roles of scenarios and models in IPBES assessment activities. The guide also gives a short list of recommendations for regional and global assessments. The guide does not provide in-depth analyses or advice on scenarios and models. The assessment of scenarios and models provides a detailed analysis of scenarios and models, and is essential reading for experts involved in regional and global assessments (deliverables 2b & 2c) and thematic assessments (deliverables 3a & 3b). The assessment of scenarios and models is intended to facilitate the use of scenarios and appropriate models in the thematic, regional and global assessments. Development of common of vocabulary, concepts and methods with the assessment on values and valuation of biodiversity and ecosystem services (deliverable 3d) will ensure that guidance on these issues are coherent. Capacity building activities of IPBES include prioritizing key capacity building needs, providing financial and other support for these priority needs, and establishing mechanisms to mobilize additional support. This methodological assessment highlights the capacity building that is require to fill the gaps in the scientific and policy communities and will involve close collaboration with the Task Force on Capacity Building (deliverables 1a & 1c). IPBES also focuses on mobilising indigenous and local knowledge (ILK) as part of efforts to strengthen the capacity and knowledge foundations to implement key functions of IPBES. This assessment provides guidance on the participation of ILK in scenarios and models as well as their use by indigenous and local communities. The Task Force on Scenarios and Modeling will work with the Task Force on Indigenous and Local Knowledge (deliverable 1b) to this end. Policy support activities of IPBES include identifying policy relevant tools/methodologies, facilitating their use, and promoting and catalysing their further development. This assessment illustrates the broad variety of ways in which scenarios and models can contribute to decision-making. Because of the key role of scenarios and models in policy support tools, tight links with the Task Force on Policy Support Tools and Methodologies (deliverable 4c) have been established and will be pursued. Knowledge generation activities of IPBES focuses on identifying knowledge needs of policymakers, and catalyse efforts to generate new knowledge. As such, the methodological assessment of scenarios and models does not outline research to be carried out by IPBES. However, it is hoped that key research gaps identified in this assessment will help catalyse efforts by funding agencies and the scientific community put in place research programs to fill these gaps. This will require close collaboration with the IPBES Task Force on Data and Knowledge Generation (deliverable 1d).
40
Page 5 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
1.2 Scenarios and models: definition and their role in assessment and decision support 1.2.1 Overview 5
10
15
20
The IPBES Conceptual Framework (Figure 1.2) provides a logical starting point for introducing, and explaining, the respective roles of scenarios and models within the context of IPBES. This framework emerged from an extensive process of consultation and negotiation, leading to formal adoption by the second IPBES Plenary (Decision IPBES-2/4, http://www.ipbes.net/), and therefore represents a key foundation for all IPBES activities. It is a simplified representation of the complex interactions between the natural world and human societies. IPBES recognizes and considers different knowledge systems, including indigenous and local knowledge systems, which can be complementary to science-based models. The Conceptual Framework is therefore intended to serve as a tool for achieving a shared working understanding across different disciplines, knowledge systems and stakeholders that are expected to be active participants in IPBES.
Figure 1.2 The IPBES analytical conceptual framework. This depicts the main elements and relationships for the conservation and sustainable use of biodiversity and ecosystem services, human well-being and sustainable development. Similar conceptualizations in other knowledge systems include “living in harmony with nature” and “Mother Earth”, among others. In the main panel, delimited in grey, “nature”, “nature’s benefits to people” and “good quality of life” (indicated as black headlines) are inclusive of all these world views; text in green denotes the concepts of science; and text in blue denotes those of other knowledge systems. Solid arrows in the main panel denote influence between elements; the dotted arrows denote links that are acknowledged as important, but are not the main focus of the Platform. The thick coloured arrows below and to the right of the central panel indicate different scales of time and space, respectively.
25
Page 6 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
As explained in the full text of Decision IPBES-2/4, this framework provides a conceptual foundation for the science-policy interface through which knowledge from science, and other knowledge systems, flows through to policy and decision-making via the four main functions of IPBES – i.e. knowledge-generation, assessment, policy-support, and capacity-building. 5
10
15
20
The high-level roles that scenarios and models play in enabling this flow of knowledge to policy and decision-making are depicted in Figure 1.3. Modelling offers a means of explicitly describing, and quantifying, interactions between major elements of the IPBES Conceptual Framework, based on best-available knowledge. In the original framework (Figure 1.2) arrows are used simply to indicate the existence of relationships between elements but convey very little about the precise nature of these relationships. Replacing these conceptual links with models allows observed, or projected, changes in the state of one element to be used to estimate, or project, resulting changes in other elements. The methodological assessment presented in this report focuses on models addressing three main links within the IPBES Conceptual Framework: • the effects of changes in indirect drivers (e.g. socio-political, economic, technological and cultural factors) on direct drivers of change in, and therefore pressures on, biodiversity and ecosystems (e.g. habitat conversion, exploitation, climate change, pollution, species introductions); • the impacts of changes in direct drivers – both negative, and positive (e.g. through policy or management intervention) – on nature, including various dimensions and levels of biodiversity, and ecosystem properties and processes; and • the consequences of changes in biodiversity and ecosystems for the benefits that people derive from nature, and that therefore contribute to good quality of life (human well-being) – including, but not limited to, ecosystem goods and services.
25
30
As indicated in Figures 1.2 and 1.3, good quality of life can also be affected directly by changes in anthropogenic assets (built, human, social, and financial) that are not mediated by changes in biodiversity or ecosystems. Comprehensive assessment of human wellbeing is likely to involve modelling of impacts of indirect socio-economic drivers both on nature’s benefits to people, and on anthropogenic assets. While models of the latter fall largely outside the scope of this document, the report does illustrate how indicators of good quality of life (e.g., people not suffering from hunger, access to electricity, etc) can be associated with scenarios and models of nature and nature's benefits. These are critical for understanding the tradeoffs and synergies between conserving biodiversity and ecosystems and attaining a broad range of sustainable development goals.
35
Page 7 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 1.3 High-level roles of scenarios and models in assessment and decision support. The rectangular boxes represent key elements from the IPBES Conceptual Framework (see Figure 1.2).
5
10
15
20
25
All three types of modelling outlined above require, as input, information on the state of one element of the framework, which a model then uses to predict, or project, the state of another element. In the simplest case, models can be applied to actual observations (data) for the first of these two elements of interest. For example, spatially-explicit data on forest habitat loss obtained from remote sensing (satellite imagery) might be used to model the expected loss of species within this forest. Such application of models to observations can play an important role in assessing the present status of nature, or nature’s benefits to people, and in assessing changes, or trends, in this status past-to-present (explored further in section 1.2.2 below). However policy and decision-making processes often also require looking beyond the present to the future. Questions raised in these processes might include, for example: What is the risk of future loss of nature, or nature’s benefits to people? How would alternative policy or management interventions alter this outcome? Using models to address questions such as these relating to possible changes in the future, rather than to actual changes in the present or recent past, poses special challenges. In this situation observations of change (e.g. in drivers) are not available to use as inputs to models, because these changes are yet to occur. Furthermore there is often considerable uncertainty associated with the future trajectory of any given input variable, because this trajectory will be affected by events and decisions that have also not yet occurred, and are often unpredictable. Scenarios provide a useful means of dealing with the reality that not just one, but many, futures are possible. In policy and decision-making around nature, and nature’s benefits, scenarios are most commonly used to address possible futures for indirect and/or direct drivers (Pereira et al. 2010; Cook et al. 2014).
Page 8 of 35
First Order Draft
5
10
15
20
25
30
35
IPBES Deliverable 3c
DO NOT CITE
Within this broad context scenarios can play two main roles, either separately or in combination. In the first of these roles, “explorative (or non-intervention) scenarios” (van Vuuren et al. 2012) are used to explore a range of plausible futures reflecting different socio-economic assumptions, and/or uncertainties associated with key drivers, thereby informing agenda setting and high-level strategy development. In the second role, “policy (or intervention) scenarios” (ibid) for a driver of interest are aligned directly with possible policy or management interventions, and therefore represent choices that can be made within a given policy or decision-making process. Explorative scenarios may also be integrated with intervention scenarios in a decision-support context, as a means of addressing uncertainties associated with drivers that might affect the outcome of a given policy or decisionmaking process but are external to, and therefore not amenable to control or influence by, that process (Peterson et al. 2003). For example, in assessing intervention scenarios involving establishment of new protected-areas, modelling of outcomes for biodiversity expected from alternative reserve configurations might need to consider the effects of a plausible range of explorative climate scenarios, to account for uncertainties in future climate impacts on biodiversity. Scenarios and models each play different, but highly complementary, roles in informing and supporting policy and decision-making (Figure 1.3). Scenarios are used to describe possible futures for drivers of change (indirect and/or direct), and options for altering the course of drivers through policy and management interventions. Models then enable scenarios of change in drivers to be translated into expected impacts on nature (biodiversity and ecosystems) and consequences for nature’s benefits to people (including ecosystem services). As depicted in Figure 1.3, the interaction of policy and decision-making processes with scenarios and models will nearly always be mediated by some form of assessment or decision-support system or process, here referred to generically as an “interface”. This interface manages the translation of high-level policy and decision-making needs into explicit scenarios for analysis by appropriate models and, in turn, interprets and communicates outputs from this modelling back to the world of policy and decision-making. The following four subsections describe major components of the linked system depicted in Figure 1.3 in more detail, laying out the scope and main attributes of, and options for, each of these components, and highlighting important dependencies between them – starting with “policy and decision-making context”, and then moving on to “assessment and decision-support interface”, “scenarios”, and “models”. The “knowledge (scientific, indigenous, local)” component is not considered further in this chapter (NOTE: a subsection dealing with this is likely to be added in the Second Order Draft). However, a recurring message throughout this report (and especially in Chapters 7 and 8) is that the quality of support provided by scenario analysis and modelling to policy and decision-making is very much dependent on the quality, and relevance, of underpinning knowledge and data. The importance placed by IPBES on this issue is also reflected by the establishment of two key activities under the IPBES Work Programme: the Task Force on Knowledge and Data Generation; and the Task Force on Indigenous and Local Knowledge.
40
1.2.2 Policy and decision-making context
45
What exactly is meant by “policy and decision-making”? The adoption of this term in Figure 1.3 follows its use in various other IPBES documents including, for example, documentation of the Conceptual Framework (Decision IPBES-2/4, http://www.ipbes.net/). However, policy and decisionmaking can encompass a very broad range of processes and activities conducted in a wide variety of contexts across multiple scales. A reasonable understanding of this diversity is required to better Page 9 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
appreciate key differences in the needs of different policy and decision-making activities, and implications of these differences for the appropriateness of different approaches to assessment and decision-support, scenarios, and models, discussed in the following three subsections (Figure 1.4). 5
10
1.2.2.1 Roles of scenarios and models across phases of the policy cycle Numerous frameworks have been proposed over recent decades for conceptualising phases or elements of the policy cycle, and similar frameworks have also been developed for describing adaptive planning or management cycles. There is considerable commonality between most of these frameworks, and this is reflected in the synthesised framework adopted recently by the IPBES Expert Group Developing a Catalogue of Policy Tools and Methodologies (Deliverable 4c; cite document when available), which contains just three broad phases or elements: 1) agenda setting and review (evaluation); 2) policy design and decision-making; and 3) policy implementation (which is also referred to as “planning and management” in parts of this report, e.g. in Chapter 2). Scenario analysis and modelling can inform and support activities across all three of these elements.
15
20
25
30
35
40
45
Agenda setting and review A sizeable proportion of previous efforts in scenario analysis and modelling of biodiversity and ecosystem services have been targeted at agenda setting – i.e. identifying and promoting the need for action to address detrimental changes in nature and nature’s benefits to people. While Deliverable 3c, and therefore this report, is focused primarily on the integrated use of scenarios and models (in combination), and therefore on modelling change into the future, it is worth noting in passing that modelling can, and does, play an important role in agenda setting even in the absence of scenarios. As indicated in section 1.2.1, models can be applied to actual observations (data) for input variables, rather than to future scenarios for these variables, thereby shedding valuable light on the present status of nature and its benefits, and on changes, or trends, in this status past-to-present (Leadley et al. 2014). Several elements of the IPBES conceptual framework align well with major categories of indicators within the widely adopted DPSIR (driverspressures-states-impacts/benefits-responses) approach to status-and-trend assessment (Feld et al. 2010; Sparks et al. 2011). Modelling can add considerable value to such assessments in two important ways. Firstly, modelling can be used to help fill gaps in data needed to underpin key indicators. While ongoing data acquisition is clearly of vital importance (as reflected by the establishment of an IPBES Taskforce on Knowledge and Data) data are much easier and/or less costly to obtain for some elements of the IPBES conceptual framework than for others. For example, advances in remote sensing have now made it possible to track temporal changes in a number of direct drivers (pressures), including habitat conversion and climate change, at relatively fine spatial resolutions across extensive regions. On the other hand, most components of biodiversity, particularly at the species and genetic levels, are not detectable through remote sensing, and changes in their state can be observed only through direct field survey. Such data therefore tend to be sparsely, and unevenly, distributed across both space and time. Modelling offers a cost-effective means of filling gaps in this coverage by using remotely sensed, and therefore geographically complete, information on drivers to estimate changes in the state of biodiversity (past to present) expected across unsurveyed areas (Ferrier 2011). Using modelling to fill gaps in information can play an equally valuable role in assessing status and trends in nature’s benefits to people – e.g. by estimating changes in the supply Page 10 of 35
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
of ecosystem services from remotely-sensed land cover classes and structural or functional ecosystem attributes (biomass, net primary production etc) (Tallis et al. 2012; Andrew et al. 2014). Secondly, modelling can enable the integration of multiple pressure-state-response elements into composite indicators. Applications of the DPSIR framework typically generate multiple indicators (Butchart et al. 2010; Sparks et al. 2011), distinguished not only by their focus on different high-level components of this framework (e.g. pressure indicators versus state indicators versus response indicators) but also by differences in the focus of indicators within each component – e.g. indicators of habitat-conversion pressures versus species-introduction pressures; or indicators of habitatprotection (reservation) responses versus introduced-species-control responses. To provide a better sense of the overall status of, and trends in, the condition or “health” of the system as a whole these individual indicators are sometimes aggregated to produce one, or a small number of, composite indicators or indices (Halpern et al. 2012). While aggregation will often be most readily achieved through simple summation or multiplication (Butchart et al. 2010), this may fail to adequately address the often complex, non-linear nature of interactions between multiple pressure, state and response elements in real-world systems. Modelling offers an alternative means of integrating data, and indicators, describing past-to-present changes across multiple system elements, to generate composite indicators that better account for complexities and dynamics in these interactions (Vackar et al. 2012; Pereira et al. 2013; Tett et al. 2013). As just described, applications of modelling to actual observations can play an important role in agenda setting by providing information on the present status of nature, and nature’s benefits to people, and on trends in this status past-to-present. Linking models to scenarios builds on this role by extending the focus of assessment from changes that are known to have already occurred pastto-present, to changes that might occur into the future. Using explorative scenarios (introduced in section 1.2.1) to project possible changes beyond the present provides a powerful means of assessing future risks to nature and its benefits, and therefore the need for action (Pereira et al. 2010; Cook et al. 2014). If the range of explorative scenarios considered in a given process is sufficiently broad, and scenarios are formulated with sufficient care, then the role played in agenda setting may extend beyond simply identifying the need for action, to also shedding light on the potential effectiveness and feasibility of broad options for policy intervention. By additionally incorporating outcomes from previous policy design and implementation decisions (see next subsection), this same general form of scenario analysis can be used to inform the review of implemented policies.
Page 11 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 1.4 Key attributes of, and options for, four major components of the linked system depicted in Figure 1.3 – policy & decision-making context, assessment & decision-support interface, scenarios, and models.
5
10
15
Policy design and implementation Moving from assessing the need for action in agenda setting, to actual decision-making around specific actions in policy design and implementation, shifts the focus of scenario analysis and modelling from explorative scenarios to intervention scenarios (introduced in section 1.2.1; see case studies in Boxes 1.2 and 1.3). Intervention scenarios can play a role in both policy design and policy implementation, and these are therefore considered together here. While the boundary between policy design and implementation is often rather fuzzy, the requirements for intervention scenarios on either side of this boundary can be quite different, especially in terms of the level of specificity, and spatial explicitness, with which potential actions are defined. This is particularly the case for policies allowing choice in the location of actions implemented under these policies – e.g. establishment of new protected areas to meet a high-level percentage-reservation target, or allocation of funding under various economic instruments (e.g. an environmental stewardship scheme). In such situations, lower-level decisions made during the implementation of a high-level policy can have significant implications for the effectiveness of the outcome actually achieved by Page 12 of 35
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
that policy – not just in biophysical terms, but also in terms of implementation costs, and socioeconomic consequences for people affected by these decisions. For example, in relation to the above cases, decision-making around the precise location of new protected areas, or funded stewardship actions, may require spatially-explicit intervention scenarios at a much finer spatial resolution than those needed to inform the initial design of these high-level policies.
1.2.2.2 Diversity in scope of policy and decision-making processes: values and spatial/temporal scales 10
15
20
25
Activities relating to all three policy-cycle elements discussed in the previous subsection can occur at a wide range of spatial scales – global, regional, national, sub-national and local. The spatial extent and resolution of scenarios and models employed in any given policy or decision-making process must therefore be aligned carefully with the scale of interest for that process. Policy and decisionmaking processes can also address quite different temporal scales of concern – ranging from processes focused on short-term outcomes (changes made over a few years) through to those focused on achieving longer-term change (e.g. over several decades) – which again has clear implications for the temporal scale of any scenarios and models employed. The focus placed on different values associated with nature, or nature’s benefits to people, can also vary markedly across policy and decision-making processes – e.g. intrinsic values associated with different levels or dimensions of biodiversity and ecosystem properties and processes, or utilitarian values associated with different types of ecosystem goods and services (see IPBES Conceptual Framework, Decision IPBES-2/4, http://www.ipbes.net/). Many of these processes – particularly those based in domains or sectors not focused primarily, or exclusively, on biodiversity and ecosystem services – will also involve consideration of a broader range of environmental, social and economic values (including anthropogenic assets, as depicted in Figures 1.2 and 1.3).
1.2.3 Assessment and decision-support interface 30
35
40
The application of scenario analysis and modelling to policy and decision-making is usually mediated by an interface of some form – i.e. a process, or system, that manages the translation of policy and decision-making needs into explicit scenarios for analysis by models and, in turn, interprets and communicates outputs from this modelling back to policy and decision-making (Figure 1.3). The form and complexity of this interface depends very much on the precise nature of the policy or decisionmaking process being served, and particularly on the phase of the policy cycle being addressed (from section 1.2.2 above). For processes focused on agenda setting, this interface needs simply to take care of selecting and formulating any explorative scenarios to be assessed, analysing these scenarios using an appropriate set of models, and reporting results from these analyses in terms of projected outcomes for nature, or nature’s benefits to people. The interface employed in such situations will often take the form of an “assessment”, typically communicating results in technical reports and/or published papers. The Regional (and subsequent Global) Assessments of Biodiversity and Ecosystem Services being planned by IPBES are likely to take this form. A rather different type of interface may be needed to manage the application of intervention scenarios to actual policy design and implementation (as opposed to agenda setting), requiring a shift from relatively static assessment to more dynamic, and interactive, “decision support” (see Page 13 of 35
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Chapter 2). This is because the number of potential options for intervention can be very large, particularly within the policy-implementation phase – e.g. picking up on the examples from section 1.2.2.1, a large number of possible configurations of protected areas, or of funded stewardship actions. If all possible options of interest are known at the outset of a decision-making process then various forms of mathematical (computer-based) optimisation might be used to automate the search for a ‘best-solution’ intervention, or set of interventions, based on modelling of the consequences of these for nature, or nature’s benefits to people (Williams & Johnson 2013). However, many policy design and implementation processes – especially at lower (more local) levels of decision-making – require consideration of intervention options that are not necessarily known in advance, but instead arise dynamically from interactions and negotiations within the process itself. This means that intervention scenarios must be formulated, and analysed, progressively throughout the decision-making process. Searching for, and reaching agreement on, effective policy or management interventions in such situations becomes more a process of interactive trial-and-error, involving adaptive evaluation and modification of intervention scenarios informed by feedback on the modelled consequences of these options. Growing recognition of this need for more interactive, and inclusive, involvement of decision-makers and stakeholders in the formulation and evaluation of intervention scenarios is reflected in the recent proliferation of planning approaches, both qualitative and quantitative, based around so-called “participatory scenarios” (Walz et al. 2007; Sandker et al. 2010; Priess & Hauck 2014). The basic idea of using models to evaluate consequences of intervention scenarios, as a foundation for decision-making, is already well established within several existing methodological paradigms or frameworks including, for example: “structured decision making” (Addison et al. 2013), “management strategy evaluation” (Fulton et al. 2014), “scenario planning” (Peterson et al. 2003), and “strategic foresight” (Cook et al. 2014). Tools associated with these, and related, paradigms are often called upon to fulfil the role of decision-support interface depicted in Figure 1.3. An important issue when considering alternative approaches to formulating and evaluating intervention scenarios for decision-support is the distinction between forecasting and backcasting strategies (Dreborg 1996; van Vuuren et al. 2012). In a forecasting strategy, scenarios are formulated and modelled for each of the intervention options being considered, and these are then evaluated and compared in terms of the relative change achieved for some aspect of nature or its benefits. A backcasting strategy instead first defines an end-point or goal that must be achieved – e.g. a desired level of change in nature or its benefits – and then searches for one or more intervention scenarios that fulfil this goal (see case study in Box 1.1). Both strategies can be applied either using some form of mathematical optimisation, or through more interactive, participatory engagement. This distinction between forecasting and backcasting has strong parallels with the well-established use of “maximum coverage” versus “minimum set” strategies in systematic conservation planning (Kukkala & Moilanen 2013) – where a maximum-coverage strategy searches for protected-area configurations that maximise the level of biodiversity conservation achieved given a fixed budget, while a minimum-set strategy searches for configurations that achieve a fixed conservation goal, or set of goals (end-points), at minimal cost. A final important characteristic of some methodologies operating within the “assessment and decision-support interface” (Figure 1.3) is the ability to aggregate, and thereby synthesise, results Page 14 of 35
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
from modelling of different values, through various forms of multi-criteria analysis (Arhonditsis et al. 2002). Both explorative scenarios employed in agenda setting, and intervention scenarios employed in policy design and implementation, are often evaluated using multiple models dealing with different values associated with nature (e.g. multiple biodiversity or ecosystem attributes), or nature’s benefits to people (e.g. multiple ecosystem services). As mentioned in section 1.2.1, these same scenarios may also be evaluated using models focused on anthropogenic assets (built, human, social, and financial) that are not mediated by changes in biodiversity or ecosystems. Multi-criteria analysis can play a crucial role in aggregating modelled outcomes across different values into composite indices of quality of life (human wellbeing) (Ding & Nunes 2014), particularly if this analysis is well integrated with other rigorous assessment and decision-support methodologies such as those described above.
1.2.4 Scenarios 15
20
25
30
35
40
The IPBES emphasises the importance of scenario analyses for decision making and notes that “…Scenarios of future socioeconomic development and models of the impacts of these development pathways are key elements of nearly all environmental assessments…. scenarios and projections of future trends are crucial for anticipating future changes in biodiversity and ecosystem services and for developing proactive strategies to minimize future degradation of, or restore, biodiversity and ecosystem service” (IPBES 2/3,p.10,12). Scenarios are like ‘crystal gazing’ where one tries to depict or visualize plausible futures under alternative contexts, assumptions, risks and opportunities, and policy interventions/management options. Scenarios aim to foresee various pathways of changes in indirect and direct drivers so that impacts on nature and nature's benefits can be evaluated. Comparisons of alternative scenarios can then be used to guide and formulate policies such as increasing protected areas, reducing fossil fuel use, or establishing hunting or fishing restrictions. Scenarios, in the sense that is used throughout this document, include changes in indirect drivers such as human population, per capita use of energy and development of infrastructure, and changes in direct drivers such as climate change, land use change, pollution or fishing pressure. Scenarios are often used to capture complexity, understand uncertainties, to assess interactions of drivers of change, or to test alternative development trajectories (Priess and Hauck, 2014). There are varying definitions of ‘scenarios’ but on one point there is consensus – scenarios are not predictions of the future (van der Heijden et al, 2002). The EEA (2009) defines a scenario as a consistent and plausible picture of a possible future reality that informs the main issues of a policy debate. The IPCC describes scenarios as “a coherent, internally consistent and plausible description of a possible future state of the world. It is not a forecast; rather, each scenario is one alternative image of how the future can unfold” (IPCC 2014). Scenario development emerged following World War II in US military strategic planning with the RAND Corporation. Scenarios achieved prominence in the 1970s in speculation about the future of society, the economy and the environment (van Notten, 2005). Today scenarios are used in a wide range of contexts by small, medium and large enterprises; in regional, national foresight studies; and in environmental assessments for public policy for example the UNEP’s Global Environmental Outlook, CBD’s Global Biodiversity Outlooks , OECD’s Environmental Outlook to 2050, UK’s National Ecosystem Assessment (van Notten, 2005; Bateman et al, 2014).
Page 15 of 35
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
As introduced in Section 1.2.1, it is useful to distinguish between two broad classes of scenarios (van Vuuren et al. 2012) – “explorative scenarios” and “intervention scenarios”. Each of these can be implemented using a variety of methodologies, effectively yielding a larger number of sub-classes of scenario types. For explorative scenarios, the most widely used approach has been to develop “plausible futures” involving the building of narratives or storylines of socio-economic and environmental pathways including assumptions regarding, for example, technological development (Pereira et al, 2010). Thus these storylines are the backbones of the scenarios (Spangenberg et al, 2012). They are the qualitative component, defining philosophies, policies and instruments, which is then complemented by a quantitative component (Spangenberg et al, 2012). Examples of this general approach are the SRES scenarios of IPCC, the Millennium Ecosystem Assessment scenarios and the scenarios developed in GEO. Sparrow (2000) argues that planners advising decision makers should interpret such scenarios as more exploratory so that a scenario is less a strategy and more a coherently structured speculation (cited in van Notten, 2005). Of late scenarios that are highly quantitative, including those based on econometric analyses, have gained prominence (Pereira et al, 2010). In recent years, the plausible-futures approach has been increasingly complemented by alternative approaches to the development of explorative scenarios, for example: “statistical extrapolation” into the future of past observed trends in the state of biodiversity and ecosystems (e.g. Tittensor et al 2014); and “probabilistic scenarios”, employing similar process-based models to those employed in modelling plausible futures, but using inputs drawn from probability distributions for each parameter based on best-available empirical data or expert knowledge, in place of discrete “plausible” combinations of parameter values, thereby allowing probabilities to be attached to resulting projections (e.g. Abt Associates 2012).
25 As discussed in Section 1.2.3, intervention scenarios can be developed and applied in either a “forecasting” or a “backcasting” mode (van Vuuren et al. 2012). An example of the back-casting approach is the analysis for the Rio+20 conference, described in Box 1.1 (PBL, 2012 ). Examples of forecasting are the analyses of policy options presented in Boxes 1.2 and 1.3. 30
40
Scenarios can be developed in either a top-down or bottom-up manner – where, once scenarios are developed at a higher scale, these are then used as a guide to develop the scenarios at lower scales (say, from global to regional and national scales) or vice versa where after developing scenarios at lower scales these are aggregated to higher scales. The disadvantage of the first approach is that drivers or factors that may be relevant at, say, a global scale may not be that important at lower scales and vice versa. Another approach is the inclusive approach where participatory methods and scenarios are developed in consultation with stakeholders and users/decision makers. No approach can be considered as superior to the other and the choice of which approach to adopt for scenario development depends on the objectives of the scenario analysis and the needs and preferences of the users/decision makers.
45
The scenario development process involves a number of stages, which include: consulting stakeholders, users and decision makers; exploiting and improving the knowledge base for developing scenarios; and evaluating their credibility and validity through peer reviews. Once the contours of the scenarios are set and mapped, drivers are selected and interactions between drivers
35
Page 16 of 35
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
are analysed. The consequences of these social, economic, and cultural developments, along with relevant data and other additional information, can be analyzed by using models to project and estimate the values of the variables under study. Scenario analysis and models can enable governments and societies to anticipate future events (e.g. global warming, loss of biodiversity/species) and the driving forces behind them (e.g. fossil fuel use, habitat loss, land use changes or pollution) and take remedial measures or proactive policies (e.g. reducing fossil fuel use, increasing protected area coverage, conservation programs for species under threat of extinction) in anticipation of these environmental challenges.
1.2.5 Models The models that can be used by IPBES fall into three broad categories (from IPBES 2 /17): • Models projecting changes in direct drivers of biodiversity and ecosystem function (e.g., land use change, fishing pressure, climate change, invasive alien species, nitrogen deposition), see chapter 3; • Models assessing the impacts of drivers on biodiversity and ecosystems (e.g., species extinctions, changes in species abundance and shifts in ranges of species, species groups or biomes), see chapter 4; and • Models assessing the impacts of drivers, and changes in biodiversity and ecosystems, on ecosystem services (e.g., ecosystem productivity, control of water flow and quality, ecosystem carbon storage, cultural values) see chapter 5. • This section first provides a general overview of the elements of scientific models, then broadly outlines the different types of models that are most commonly used for assessing the impacts of direct and indirect drivers on nature and nature's benefits and finishes with comments on the importance of the use of multiple types of models in IPBES activities.
1.2.5.1 What is a model?
30
35
40
There are various broad definitions of scientific models, such as: • “an abstract representation of reality”; • “an approximation or simulation of a real system that omits all but the most essential variables of the system” ( http://www.yourdictionary.com/scientific-model) ; or • “a schematic description of a system, theory, or phenomenon that accounts for its known or inferred properties and may be used for further study of its characteristics: a model of generative grammar; a model of an atom; an economic model” (http://www.thefreedictionary.com/model). These definitions all involve two necessary aspects of models, namely that 1) a scientific model must contain variables that can take on a range of values or states and 2) one must be able to establish relationships either between the variables or between the value ranges of these variables (from Ritchey, 2012). This implies that a model is a simplified description of a real system by establishing relationships between the most essential variables of that system. Mathematical models play an important role in all fields of science because they cumulate and summarize knowledge, and improve the consistency and repeatability of analyses. A scientific model aims to address specific questions about the when (temporal), where (spatial), what (pattern) or processes (behaviour) of Page 17 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
real world phenomena (Borner et al., 2012). In addition, it is essential to relate these key questions and model outputs to policy and decision-making processes of IPBES.
5
10
15
20
Models of a specific system differ in the choice of essential variables and relationships, properties of the variables they describe, and the way relationships are described. The types of models combine several properties. Relevant properties for models are the type of variables used, the types of relationships described, and how these relationships are described. Variable types can include variables with or without specified value ranges. Value ranges may include continuous (e.g. 0.8, 0.9, 1.5), discrete ordered (e.g. 1, 2, 5, or low, medium, high) or categorical values (e.g. grassland, forest, urban). The types of relationships include quantitative vs. qualitative; dynamic vs. static; and spatial vs. non-spatial relationships. The way the relationships are established may be 1) mathematical or functional relationships using established scientific understanding and mathematical formulation of relevant underlying processes (e.g. metapopulation modelling Gordon et al., 2012, or mechanistic models of ecosystem function e.g. Harfoot et al. 2014b); 2) correlative (statistical) analysis of available empirical data (e.g. species distribution modelling, Elith and Leathwick, 2009); 3) probability assignment methods (e.g. Bayesian Belief Network, Haines Young, 2011); and 4) ‘quasi’ causal methods, using expert knowledge to capture and represent stakeholder knowledge (e.g. using participatory techniques (e.g. Priess and Hauck, 2014; Walz et al. 2007). Modelling approaches can combine these properties in different ways to yield, for example, a ‘quantitative, dynamic, non-spatial, mechanistic model’.
1.2.5.2 Types of models that may be used in IPBES activities 25
30
35
40
45
Figure 1.5 shows examples of the variables and relationships between them that are accounted for in models that may be used by IPBES in its various activities. This figure illustrates a number of key points about models: • There is a tremendous variety of variables simulated by models of Nature and Nature's benefits. In some studies only a single variable is simulated. For example, empirical species distribution models are often used to simulate only the spatially explicit response of a single variable, species presence or absence, to environmental change. In many other cases, models simulate several variables, but typically only a small subset of the variables listed in Figure 1.5. For example, biodiversity models simulate dynamics of genes, or species, or functional groups, or communities; but many focus on only one of these levels and none simulate biodiversity dynamics at all these levels. • In practice, there are privileged sets of relationships between variables when linking the three main components of Nature and Nature's benefits. For example, models of ecosystem function, especially at large spatial scales, typically represent biodiversity using a small number of groups of species that have similar characteristics (i.e., functional groups). A few models of ecosystem function use species level variables, but very few incorporate variables related to genetic adaptation (but see Kramer et al. 2010). Models of Nature's benefits typically rely on empirical relationships between habitat type and ecosystem services (arrow directly from habitat), or use inputs from variables simulated by models of ecosystem function, but few account for the contribution of species diversity to ecosystem function (Cardinale et al. 2012, but note that some models do account for a small set of key species interactions). Page 18 of 35
First Order Draft
•
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
Policy and decision-making can be supported by models in the various parts of the policy cycle: agenda setting, exploring policy options (ex-ante evaluation), evaluating the effectiveness of policies (ex-post evaluation), and in policy implementations, for example in norm-setting and spatial planning. These applications need the full range of modelling steps from indirect driver – direct drivers – impacts on biodiversity and ecosystem services and their relevance for human well being. For example, Integrated Assessment Models (IAMs) are widely used in global and regional assessment activities. These models often account for some ecosystem functions and ecosystem services, but usually lack representation of biodiversity below the functional group or habitat type level, do not include key ecosystem functions and omit cultural services (Harfoot et al. 2014a, but see Harfoot et al. 2014b and Alkemade et al. 2009 for examples of including species diversity in global ecosystem models).
Figure 1.5 Variables commonly simulated by models of impacts of drivers on the three main components of Nature and Nature's benefits (boxes). Privileged relationships between variables in models are indicated by blue arrows. The thickness of the arrows indicates the frequency with which variables are connected in the most commonly used modelling approaches.
An important set of models that are not illustrated in this figure are models addressing the relationships between indirect drivers and direct drivers. These models are mainly developed for purposes other than the assessment of biodiversity and ecosystem services. There is a broad set of models available for relating indirect and direct drivers. The most relevant models are those that forecast climate changes, land use changes, hydrological changes based on economic and demographic changes, and the impacts of human management, and construction Linking models together is one means of increasing the number of variables that can be studied simultaneously (see Chapter 6); however, this comes at the cost of increasing model complexity.
Page 19 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
The modelling steps may be integrated in a single modelling framework (e.g., IMAGE/GLOBIO, Stehfest et al. 2014) or realized in a chain of linked models where the input for one model is derived from the output of another model (e.g., Bateman et al. 2013, or InVEST, Nelson et al. 2009) . 5
10
1.2.5.3 Selecting types of models for a given purpose The questions posed in a policy or decision-making process are a key factor in determining the types of modelling approaches that are the most appropriate. Other key factors include the availability of data (see chapter 8) and the capacity to develop and use models (see chapter 7). If a model is mainly used for agenda setting, explaining the main relationships in a conceptual manner using qualitative, expert-based, models may be sufficient. If models are used to explore effects of human behaviour on potential policies it might be sufficient to have semi-quantitative causal models. If models are used for ex-ante or ex-post evaluation of policies, quantitative, spatial explicit models may be needed. Models that are used for specific policy implementations where large economic interests are involved may need to be precise and quantitative.
15
20
25
30
35
40
For a given question, a variety of modelling approaches may be available. It is the position taken throughout this methodological assessment that there is usually no a priori single best modelling approach for a given question. In particular, debates about the use of models with correlative vs. mechanistic vs. 'quasi'-casual relationships are frequently polluted by misconceptions about the usefulness of these various types of models. Many modelling exercises have clearly illustrated the benefits of examining multiple model types in terms of understanding of underlying processes, improving the ability to simulate biodiversity and ecosystem functions, providing complementary sets of variables and estimating uncertainty (Cheaib et al. 2010, Gritti et al. 2013, van Oijen et al. 2013). The use of multiple models does not necessarily require quantitative comparisons among models. However, in some cases IPBES may want to stimulate work on quantitative multi-model comparisons since, as the IPCC has amply demonstrated for climate models and some models of impacts on ecosystems (IPCC 2014a), they often have much more weight in decision making than individual models. This does not mean that all models are equally good. As such, the strengths and weaknesses of models should be included when presenting model outcomes. The following chapters provide more specific guidelines for model selection and evaluation of model strengths and weaknesses. For the global and regional assessment as envisioned in the IPBES work programme, various modelling approaches can be applied. For example in large global assessments combinations of quantitative, dynamic and spatial models were used to explore socio-economic scenarios (e.g. UNEP GEO4 20, MA 2005, SCBD GBO4 2014). In addition, qualitative models were used to explore and suggest policy responses. It should be noted that all models have limitations and no model can perfectly explain or predict observed dynamics of Nature or Nature's benefits. This is an unavoidable outcome of many factors including lack of knowledge about key variables and relationships; loss of information when simplifying complex real world systems to models; uncertainty in the estimation of the values of parameter and variables; and error propagation, especially complex models.
Page 20 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Box 1.1: Case study – Rio+20 scenarios Project Title
Rio+20 scenarios
Type of value
Global terrestrial and aquatic biodiversity
Driver
Human pressures
Temporal extent
Current to 2050
Spatial extent
Global
Model use
IMAGE, GLOBIO3
Client
CBD, Governments
Multiple challenges, multiple targets In 1992, governments worldwide agreed to work towards a more sustainable development that would eradicate poverty, halt climate change and conserve ecosystems. Although progress has been made in some areas, actions have not been able to alter the trend in other, critical areas of sustainable development, such as providing access to sufficient food and modern forms of energy, preventing dangerous climate change, conserving biodiversity and controlling air pollution. Without additional effort, these sustainability objectives also will not be achieved by 2050. Different pathways towards the targets To jointly reach the long term targets on human well-being (eradicating hunger and ensuring full access to modern energy sources), climate change (temperature rising less than 2 °C) and biodiversity conservation (no further loss by 2050), three scenarios were developed. The long term targets for sustainability were the starting point, for these back-casting scenarios (van Vuuren et al., 2012). Three types of scenario were defined based on different strategies of sustainable development as follows (PBL, 2012): Global Technology: focus on large-scale technologically optimal solutions, such as intensive agriculture and a high level of international coordination; for instance, though trade liberalization Decentralized Solutions: focus on decentralized solutions, such as local energy production, agriculture that is interwoven with natural corridors and national policies that regulate equitable access to food Consumption Change focus on changes in human consumption patterns, most notably by limiting meat intake per capita, by ambitious efforts to reduce waste in the agricultural production chain and through the choice of a less energy-intensive lifestyle These pathways towards the 2050 targets use a different mixtures of policies to enhance productivity and reduce biodiversity loss (Figure Box 1.1), as well as different mixtures to enhance the use of modern energy and reducing climate change. Models The scenarios were evaluated up to 2050 using the IMAGE modelling framework (Stehfest et al., 2014) combined with the GLOBIO3 model (Alkemade et al., 2009). IMAGE is an integrated assessment model of global environmental change, and enables assessment of the impacts of socioeconomic development on the environment, including land use, climate and water flow and pollution. GLOBIO3 is linked to IMAGE and calculates the impacts of environmental changes on some biodiversity indicators by using cause-effect relationships.
Page 21 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure Box 1.1. Three scenarios contrasting broad policy actions needed to eradicate hunger, maintain a stable food supply, and reduce climate change, and their impact on biodiversity loss reduction (PBL, 2012). The lefthand graph indicates the response of biodiversity (as measured by Mean Species Abundance) of a business-asusual development pathway ("Trend") and the pathway (green dotted line) that achieves a stabilization of biodiversity at the global scale by 2050 (green dot). The right-hand graph indicates the contributions of different components of the three sustainability scenarios to the difference between the "Trend" scenarios and the biodiversity stabilization pathway.
The results of scenario analyses show that different combinations of policy actions, grouped in the three scenarios, may lead to achieving the sustainability targets of eradicating hunger and maintaining a stable and sufficient food supply, and ensuring the access to modern energy sources; conservation of biodiversity; and to limit global climate change and air pollution. So these quantitatively coherent scenarios indicate that eradicating hunger as well as providing full access to modern energy, on the one hand, and achieving environmental sustainability, on the other, is possible. However, marginal improvements will not suffice; large, transformative changes are needed to realize sustainable development. The role of the Rio+20 scenarios in policy support Initially a contribution to the Rio+20 conference held in Rio de Janeiro in 2012, the scenarios and their main messages were taken up in the 4th Global Biodiversity Outlook or GBO4 (SCBD, 2014). The parties to the CBD adopted the conclusions of the GBO4, and committed to step up actions to achieve the Aichi Biodiversity targets, including to double funding for necessary actions (CBD, 2014a). Additional initiatives were launched to enhance the biodiversity perspective in sustainable commodity production (CBD, 2014b). The outcomes from the scenario analyses provided underlying arguments for these decisions and initiatives.
Page 22 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Box 1.2: Case study – Thadee watershed, Thailand Project Title
Thadee watershed, Thailand
Type of value
Watershed services
Driver
Land-use change
Temporal extent
2009-2020
Spatial extent
Catchment (112 km2)
Model use
CLUEs, InVEST, RIOS
Client
Local stakeholders, local government
The Thadee watershed located in southern Thailand covers approximately 112 km2. Water supply from the watershed is mainly used for agriculture by upstream farmers and household consumption by downstream people in the Nakhon Srithammarat municipality. However, natural forests in the watershed have been degraded and transformed to mono-cultures (fruit tree and rubber plantations) due to a governmental subsidy program. The ECO-BEST project co-funded by EU, German Government (GIZ) and Thailand (Department of National Parks, Wildlife and Plant Conservation and Kasetsart University) worked with scientists to quantify water yield and sediment load according to different land use and rainfall scenarios during 2009-2020. The CLUE-s model was used to allocate future land demands based on three scenarios – trend, agriculture development and conservation. In addition, InVEST and USLE models were employed to estimate water yield and soil erosion, respectively. The modelling results clearly showed that intensifying land use change due to rapid expansion of rubber plantation and extreme rainfall will generate a high risk of major sediment loadings and overland water-flows due to the force of rainfall and decreased evapotranspiration from vegetation. With the application of the RIOS tool, the project team together with stakeholders could identify which conservation activities (e.g., protection, reforestation and promotion of mixedcropping system) should be implemented, and where, to yield the highest return on investments and to enhance watershed services. The municipality has agreed in principle to find the best practical mechanism for collecting payments from tap water clients and downstream, so called “payment for watershed services”, to implement the above activities.
Page 23 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure Box 2.1. Integrated scenarios and modelling in ecosystem services assessment for the Thadee Watershed, Thailand.
Figure Box 2.2. Three scenarios of land use for 2030 within the Thadee watershed in Nakhon Srithammarat Province and consequences on water yields and sediment loads.
Page 24 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Box 1.3: Case study – Guyana Road Projects Impact. Project Title
Potential Impact of Road Projects on Habitat Loss and Greenhouse Gas Emissions in Guyana
Type of value
Tropical forest
Driver
Mining and road development
Temporal extent
2012 to 2022
Spatial extent
Country (211,000 km2)
Model use
Terra-I, “Training Models”
Client
Government, Inter-American Development Bank
Guyana covers approximately 1% of the terrestrial surface of South America (21.1 million ha). The country has one of the most intact tropical forests and has lower deforestation rate (0.020-0.056% per year) than most countries in South America (0.41% per year, on average) partly due to its low population size. The main drivers of deforestation include mining and road development. Potential deforestation and its consequences on GHG emission for the year 2000 were assessed based on three road development scenarios: 1) improvement of the Georgetown-Linden corridor; 2) linking Linden with the Amaila Falls; and 3) paving of the Linden-Lethem corridor (Figure Box 3.1). In addition, two models of recent deforestation rates (model 1: similar to Guyana’s region; model 2: growing mining in the Peru’s Madre de Dios region) detected by satellite-based rainfall and vegetation data (Terra-i) were combined to map deforestation risk areas. The results derived from the scenarios on a national scale show that the implementation of the explored infrastructure projects alone may result in a deforestation rate of 0.036% (best-case scenario) to 0.092% (worse-case scenario) per year. Although deforestation and its consequences for GHG emission rates might look low in comparison with other countries in Latin America, it is higher than the current rate in Guyana. It is important to maintain emission levels at low levels in order to receive the full compensation payment from the REDD+ scheme agreed to with Norway. An increase of deforestation more than 0.09%, as seen in the worse-case scenarios, will mean a loss of more than 70% in payments. Therefore, in order to reduce the risk of losses in REDD+ payments, it is highly important to design and enforce a careful mining licensing and land management policy in the event that the roads are built or upgraded.
Page 25 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure Box 3.1. Potential Deforestation under Scenario of the baseline (left) and under Scenarios of Improvement and Construction of Both Roads with Current Mining (middle) and Potential Mining or the worst-case scenario (right). Besides delivering concrete results that are useful for ongoing Inter-American Development Bank projects in Guyana, the study further explores the possibility of using the scenarios and models as a basis for land-use management and in the development of infrastructure projects and maintaining intact tropical forests in Guyana.
Page 26 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
1.3 Structure of this report
5
10
Methodologies for modelling different components of socio-ecological systems (i.e. elements of the IPBES Conceptual Framework) are increasingly being integrated within a single modelling framework (e.g. through so-called “integrated assessment models”, IAMs). Likewise, the boundary between methodologies for modelling, and methodologies for scenario development, assessment and decision-making, is becoming increasingly fuzzy as a result of closer coupling of approaches across these domains. However, in the interests of breaking the overall challenge down into manageable pieces, Chapters 2 to 5 each focus, in turn, on a particular aspect or component of this challenge (Figure 1.6). Linkages and dependencies between these topics, and the need for any given application of scenarios and models to consider these issues together, rather than sequentially, are emphasised throughout.
Figure 1.6 Relationship of chapters to the components depicted in Figure 1.3
15
20
25
Chapter 2 examines issues around “using scenarios and models to inform decision-making in diverse policy, planning and management contexts”. It provides an overview of policy, planning and management contexts in which scenarios and models can aid decision-making, and considers lessons learnt from established decision-support paradigms and frameworks that make strong use of scenarios and models. Particular emphasis is placed on the importance of aligning the design of scenarios and models with the particular needs and objectives of different decision-making processes, and of dealing with uncertainty in scenarios and models employed in decision-making Chapter 3 addresses challenges associated with “building scenarios and models of indirect and direct drivers of change in biodiversity and ecosystems”. It explores approaches to modelling plausible, or alternative, trajectories of indirect drivers through socioeconomic scenarios and lessons learnt from previous development and application of such scenarios in assessments at global, regional and subregional scales. It then reviews methods for modelling expected consequences of socioeconomic Page 27 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
scenarios for direct drivers of change in biodiversity and ecosystems across terrestrial, freshwater and marine systems. This chapter also considers potential for better coupling modelling of indirect and direct drivers of change, with potential feedback effects of changes in biodiversity and ecosystems on socioeconomic futures, through integrated assessment models (IAMs). 5
10
15
Chapter 4 deals with “modelling impacts of drivers on biodiversity and ecosystem properties and processes”. It explores existing and emerging approaches (both correlative and process-based) to modelling impacts of a broad range of direct drivers on: biodiversity across multiple levels (e.g. population, species, community) and dimensions (e.g. composition, structure, function) of biological organisation; and ecosystem properties and processes (e.g. biomass, primary production). Chapter 5 focuses on “modelling consequences of change in biodiversity and ecosystems for nature’s benefits to people”. It explores challenges associated with translating modelled biophysical changes in biodiversity and ecosystem properties and processes into expected consequences for benefits to people, by incorporating consideration of relevant values that people place on, or derive from, nature. It emphasises the importance of recognising that different decision-making processes may require a focus on different types of material and non-material values, including various ecosystem goods and services. Approaches to modelling consequences of changes in biodiversity and ecosystems for different types of benefits to people are reviewed and evaluated.
20 The remaining chapters of the report explore, in greater depth, three particularly important crosscutting challenges facing the ongoing development and application of scenario analysis and modelling from an IPBES perspective (Figure 1.6). 25
30
35
40
45
Chapter 6 articulates the need for better “linking and harmonising scenarios and models across scales and domains” and proposes practical strategies and solutions for achieving this in both the short and longer term. These include approaches to: more closely linking scenarios and models across different scales of assessment and decision-making; fostering harmonisation of scenarios and models between different regions; achieving closer coupling of models focusing on different dimensions or levels of biodiversity, or on different ecosystem properties and processes; and further coupling this modelling with models across other environmental, social and economic domains. Chapter 7 addresses the challenge of “building capacity for developing, interpreting and using scenarios and models” by proposing practical strategies that account for regional and cultural diversity in perspectives on, and capacity for, scenario analysis and modelling. These include approaches to: improving regional and national access to standard data-sets and projections that are appropriately prepared and served globally; improving access to, and useability of, software tools for scenario analysis, modelling and decision-support; developing methods for better incorporating local data and knowledge; and developing effective strategies for mainstreaming scenarios and models into participatory assessment and decision-making processes across scales and across different policy, planning and management contexts. Chapter 8 adopts an even more forward-looking, and bigger-picture, perspective in addressing the challenge of “improving the rigour and usefulness of scenarios and models through ongoing evaluation and refinement”. It lays out a comprehensive vision and strategy for taking scenario Page 28 of 35
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
analysis and modelling of biodiversity and ecosystem services to a whole new level of rigour, credibility and utility by: more closely linking scenario analysis and modelling to parallel initiatives in biodiversity/ecosystem data acquisition and change observation (monitoring) at national, regional and global scales, thereby establishing a rigorous foundation for ongoing model evaluation and calibration; and advancing the fundamental science underpinning development and application of scenarios and models through carefully prioritised research activities.
1.4 Overall messages and vision 10
NOTE: This content is preliminary only. It is copied, largely unchanged, from a stand-alone document “Deliverable 3c – indicative overall message” prepared before the first author meeting in October 2015. The section will be thoroughly revised and completed for the Second Order Draft, based on inputs from all other chapters, and working in close consultation with the CLAs of these chapters.
20
Modelling offers an effective means of explicitly describing, and quantifying, interactions between major elements of the IPBES Conceptual Framework. Linking elements of this framework through modelling allows observed, or projected, changes in one element to be used to estimate, or project, resulting changes in other elements – including effects of changes in “indirect drivers” on “direct drivers”; impacts of changes in direct drivers on “nature” (biodiversity and ecosystems); and consequences of changes in biodiversity and ecosystems for “nature’s benefits to people”, and “quality of life”.
25
Scenario analysis adds further value by using this modelling of interactions as the foundation for exploring potential impacts and consequences of possible futures, including plausible socioeconomic scenarios and/or alternative configurations of policy or management interventions affecting nature and nature’s benefits.
15
30
35
40
Modelling and scenario analysis can inform activities across the entire policy cycle – from problem identification and agenda setting (through status-and-trend assessment and scenario-based risk analysis) to supporting decision-making in policy development and implementation (through analysis of the expected consequences of alternative policy or management interventions) – see separate draft document “Role of scenarios and models in assessment and decision-making”. A diverse range of activities can benefit from these methodologies Modelling and scenario analysis can therefore contribute significantly to a wide range of activities within the IPBES Work Programme. Initially the most prominent opportunity is in relation to IPBES’s planned regional/sub-regional assessments (and subsequently the global assessment) of biodiversity and ecosystem services. Modelling can contribute to these assessments in three main ways, by: • adding value to the assessment of status and trends (past to present), in nature and nature’s benefits through model-based filling of gaps in observations underpinning key indicators, and through integration of information on multiple elements of the conceptual framework into composite indicators; • assessing future risks to nature, and nature’s benefits, through modelling of the impacts and consequences of plausible socio-economic scenarios; and
Page 29 of 35
First Order Draft
•
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
assessing how effectively these risks might be addressed through different types of highlevel policy intervention.
The potential contribution of scenario analysis and modelling to achieving the overarching goal of IPBES extends well beyond regional and global assessments. Such assessments will play an important role in problem identification and agenda setting – by appraising the need for action, and by shedding light on the likely efficacy of broad types of action. But to directly support subsequent decision-making in policy formulation and implementation, scenario analysis and modelling need to be embedded and undertaken within individual decision-making processes across a wide range of institutional/governmental contexts and scales. This is because such processes – especially at lower (more local) levels of decision-making – require consideration of detailed options for policy and/or management intervention that are rarely known in advance, but instead arise dynamically from interactions and negotiations within the process itself. To extend the benefits of scenario analysis and modelling to the broadest possible range of decision-making processes, IPBES will therefore need to complement its own use of these approaches in regional and global assessments with promotion and facilitation of their uptake by other processes, through IPBES activities such as the Taskforce on Capacity Building, and the Expert Group Developing a Catalogue of Policy Tools and Methodologies. Scenario analysis and modelling must be tailored carefully to the needs of different assessment and decision-making processes Given this breadth of potential applications, it is vital that ongoing development and use of scenario analysis and modelling caters effectively for diversity in the characteristics and needs of different assessment and decision-making processes. For example: • Assessment and decision-making processes can span a wide range of spatial scales – global, regional, sub-regional and local – and the spatial extent and resolution of scenarios and models must therefore be tailored carefully to the needs of any given process, as must the temporal scale of such analyses. • Processes often vary markedly in the focus placed on different values associated with nature, or nature’s benefits to people (e.g. different levels or dimensions of biodiversity and ecosystem properties and processes, or different types of ecosystem goods and services). Many of these activities – particularly those based in domains not focused primarily, or exclusively, on biodiversity and ecosystem services – will also involve consideration of a broader range of environmental, social and economic values. The values addressed by scenarios and models should therefore, again, be tailored carefully to the specific needs of different processes. • The appropriateness of different methodologies for scenario analysis and modelling – ranging, for example, from highly mathematical spatially-explicit techniques through to more qualitative participatory approaches – can also vary markedly between different assessment and decision-making contexts. Recognition of diversity in the characteristics and needs of different assessment and decision-making processes, and therefore the importance of tailoring scenario analysis and modelling to best serve the needs of any given application, is a major cross-cutting theme throughout our assessment.
45 Page 30 of 35
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
Considerable progress in methodological development has already been made As will be detailed in Chapters 2 to 5, considerable progress has already been made in the development of methodologies for building scenarios and models of indirect and direct drivers, modelling impacts of these drivers on biodiversity and ecosystems and consequences for nature’s benefits, and using these models to inform decision-making. Some of these methodologies are sufficiently well developed to be applicable in forthcoming IPBES regional/subregional (and subsequently global) assessments of biodiversity and ecosystem services, and in decision-making processes more broadly if properly promoted and supported through relevant IPBES activities (e.g. Taskforce on Capacity Building, Expert Group Developing a Catalogue of Policy Tools and Methodologies). Significant challenges remain, but these can be conquered with appropriate planning and effort Across all four topics addressed in Chapters 2 to 5, significant gaps and weaknesses still remain in currently available methodologies and therefore, as detailed in these chapters, much further work is needed to ensure that scenario analysis and modelling can effectively serve the needs of assessment and decision-making into the future. The remaining chapters of the report (Chapters 6, 7 and 8) explore, in greater depth, particularly important challenges facing the ongoing development and application of scenario analysis and modelling from an IPBES perspective.
Page 31 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
References
5
10
15
20
25
30
35
40
45
Abt Associates (2012) Development of probabilistic socioe-economic emissions scenarios. Final report prepared for US Environmental Protection Agency. Addison, P.F.E., Rumpff, L., Bau, S.S., Carey, J.M., Chee, Y.E., Jarrad, F.C. et al. (2013). Practical solutions for making models indispensable in conservation decision-making. Diversity and Distributions, 19, 490502. Alkemade R., van Oorschot M., Miles L., et al., (2009) GLOBIO3: A framework to investigate options for reducing global terrestrial biodiversity loss. Ecosystems 12: 374-390. Andrew, M.E., Wulder, M.A. & Nelson, T.A. (2014). Potential contributions of remote sensing to ecosystem service assessments. Progress in Physical Geography, 38, 328-353. Arhonditsis, G., Karydis, M. & Tsirtsis, G. (2002). Integration of mathematical modeling and multicriteria methods in assessing environmental change in developing areas: A case study of a coastal system. Journal of Coastal Research, 18, 698-711. Bateman, I., et al. (2014). Economic analysis for ecosystem assessments. in K.N.Ninan.ed. Valuing Ecosystem Services-Methodological Issues and Case Studies, Edward Elgar, Cheltenham and Northampton. 23-77. Bateman, I.J., Harwood, A.R., Mace, G.M., Watson, R.T., Abson, D.J., Andrews, B. et al. (2013). Bringing Ecosystem Services into Economic Decision-Making: Land Use in the United Kingdom. Science, 341, 45-50. Börner, K., et al. (2012). An introduction to modeling science: Basic model types, key definitions, and a general framework for the comparison of process models, pp. 3-20. In: Andrea Scharnhorst, A., Börner, K., and van den Besselaar, P. (Eds.). Models of Science Dynamics: Encounters Between Complexity Theory and Information Sciences. Springer, Berlin Heidelberg. Butchart, S.H.M., Walpole, M., Collen, B., van Strien, A., Scharlemann, J.P.W., Almond, R.E.A. et al. (2010). Global Biodiversity: Indicators of Recent Declines. Science, 328, 1164-1168. Cardinale, B.J., Duffy, J.E., Gonzalez, A., Hooper, D.U., Perrings, C., Venail, P. et al. (2012). Biodiversity loss and its impact on humanity. Nature, 486, 59-67. CBD (2014a) http://www.cbd.int/doc/press/2014/pr-2014-10-17-cop-12-en.pdf CBD (2014b) http://www.cbd.int/doc/press/2014/pr-2010-10-16-commodities-en.pdf Cheaib, A., Badeau, V., Boe, J., Chuine, I., Delire, C., Dufrene, E. et al. (2012). Climate change impacts on tree ranges: model intercomparison facilitates understanding and quantification of uncertainty. Ecol. Lett., 15, 533-544. Cook, C.N., Inayatullah, S., Burgman, M.A., Sutherland, W.J. & Wintle, B.A. (2014). Strategic foresight: how planning for the unpredictable can improve environmental decision-making. Trends in Ecology & Evolution, 29, 531-541. Ding, H.L. & Nunes, P. (2014). Modeling the links between biodiversity, ecosystem services and human wellbeing in the context of climate change: Results from an econometric analysis of the European forest ecosystems. Ecological Economics, 97, 60-73. Dreborg, K.H. (1996). Essence of backcasting. Futures, 28, 813-828. EEA. (2009). Looking back on looking forward: a review of evaluative scenario literature. EEA Technical Report. No.3/2009. European Environment Agency. Copenhagen. Elith, J. and Leathwick, J.R. (2009) Species distribution models: ecological explanation and prediction across space and time. Annual Review of Ecology Evolution and Systematics 40, 677-697 Feeley, K.J. and Silman, M.R. (2010) Land-use and climate change effects on population size and extinction risk of Andean plants. Global Change Biology 16, 3215-3222 Feld, C.K., Sousa, J.P., da Silva, P.M. & Dawson, T.P. (2010). Indicators for biodiversity and ecosystem services: towards an improved framework for ecosystems assessment. Biodiversity and Conservation, 19, 28952919. Ferrier, S. (2011). Extracting more value from biodiversity change observations through integrated modeling. Bioscience, 61, 96-97. Page 32 of 35
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
FRB Foundation pour la Recherche en Biodiversité (2013) "Scénarios de la biodiversité : un état des lieux des publications scientifiques françaises". Fulton, E.A. (2010) Approaches to end-to-end ecosystem models. Journal of Marine Systems 81, 171-183 Fulton, E.A., Smith, A.D.M., Smith, D.C. & Johnson, P. (2014). An Integrated Approach Is Needed for Ecosystem Based Fisheries Management: Insights from Ecosystem-Level Management Strategy Evaluation. Plos One, 9, 16. Gordon, A., et al. (2012) The use of dynamic landscape metapopulation models for forest management: a case study of the red-backed salamander. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 42, 1091-1106 Gritti, E.S., Gaucherel, C.d., Crespo-Perez, M.-V. & Chuine, I. (2013). How Can Model Comparison Help Improving Species Distribution Models? PLoS One, 8, e68823. Haines-Young R. (2011). Exploring ecosystem services issues across diverse knowledge domain using Bayesian Belief Networks. Progress in Physical Geography. 35, 681-699 Halpern, B.S., Longo, C., Hardy, D., McLeod, K.L., Samhouri, J.F., Katona, S.K. et al. (2012). An index to assess the health and benefits of the global ocean. Nature, 488, 615-+. Harfoot, M., Tittensor, D.P., Newbold, T., McInerny, G., Smith, M.J. & Scharlemann, J.P.W. (2014a). Integrated assessment models for ecologists: the present and the future. Glob. Ecol. Biogeogr., 23, 124-143. Harfoot, M.B.J., et al. (2014b) Emergent Global Patterns of Ecosystem Structure and Function from a Mechanistic General Ecosystem Model. Plos Biology 12, 24 IPCC (2014). IPPC data distribution centre. Available at http://www.ipccdata.org/guidelines/pages/definitions.html; Last accessed 3-12-2014. IPCC. (2014a). Field et al. WG2. "Climate Change 2014: Impacts, Adaptation, and Vulnerability." IPCC Switzerland IPCC. (2014b). Edenhofer et al. WG3. IPCC Switzerland Kukkala, A.S. & Moilanen, A. (2013). Core concepts of spatial prioritisation in systematic conservation planning. Biological Reviews, 88, 443-464. Leadley P., Pereira H.M., Alkemade R., Fernandez-Manjarrés J.F., Proença V., Scharlemann J.P.W., Walpole M.J. (2010) "Biodiversity Scenarios: Projections of 21st century change in biodiversity and associated ecosystem services". Secretariat of the Convention on Biological Diversity. Technical Series No. 50. Leadley, P. W., et al. (2014). "Progress towards the Aichi Biodiversity Targets: an assessment of biodiversity trends, policy scenarios and key actions." Secretariat of the Convention on Biological Diversity. Technical Series No. 78. Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, D.R. et al. (2009). Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Front. Ecol. Environ., 7, 4-11. PBL (2012) Roads from Rio+20: pathways to achieve global sustainability goals by 2050. PBL Netherlands Environmental Assessment Agency. Pereira, H.M., Ferrier, S., Walters, M., Geller, G.N., Jongman, R.H.G., Scholes, R.J. et al. (2013). Essential Biodiversity Variables. Science, 339, 277-278. Pereira, H.M., Leadley, P.W., Proenca, V., Alkemade, R., Scharlemann, J.P.W., Fernandez-Manjarres, J.F. et al. (2010). Scenarios for global biodiversity in the 21st Century. Science, 330, 1496-1501. Peterson, G.D., Cumming, G.S. & Carpenter, S.R. (2003). Scenario planning: a tool for conservation in an uncertain world. Conservation Biology, 17, 358-366. Priess, J.A. and Hauck, J. (2014) Integrative Scenario Development. Ecology and Society 19, 14 Reymondin, L. et al. (2014). Potential impact of road projects on habitat loss and greenhouse gas emissions in Guyana from 2012 to 2022. IDB Technical Note 632. Inter-American Development Bank. Ritchey, T. (2012) Outline for a Morphology of Modelling Methods. Contribution to a General Theory of Modelling Acta Morphologica Generalis 1, 1
Page 33 of 35
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Sandker, M., Campbell, B.M., Ruiz-Perez, M., Sayer, J.A., Cowling, R., Kassa, H. et al. (2010). The Role of Participatory Modeling in Landscape Approaches to Reconcile Conservation and Development. Ecology and Society, 15, 16. sCBD (2014). Global Biodiversity Outlook 4. SCBD GBO2 - Secretariat of the Convention on Biological Diversity (2006) "Global Biodiversity Outlook 2." Montreal, 81 + vii pages SCBD GBO3 - Secretariat of the Convention on Biological Diversity (2010) "Global Biodiversity Outlook 3." Montréal, 94 pages. SCBD GBO4 - Secretariat of the Convention on Biological Diversity (2014) "Global Biodiversity Outlook 4." Montréal, 155 pages. Spangenberg et al. (2012). Scenarios for investigating risks to biodiversity. Global Ecology and Biogeography.21.5-18.: Sparks, T.H., Butchart, S.H.M., Balmford, A., Bennun, L., Stanwell-Smith, D., Walpole, M. et al. (2011). Linked indicator sets for addressing biodiversity loss. Oryx, 45, 411-419. Stehfest E., van Vuuren D., Kram T. et al. (2014) Integrated Assessment of Global Environmental Change with IMAGE 3.0. Model description and policy applications, The Hague: PBL Netherlands Environmental Assessment Agency. Tallis H., Ricketts T., and Guerry A. (eds.) (2014). InVEST 2.0 Beta User’s Guide: Integrated Valuation of Ecosystem Services and Tradeoffs/ Available at: http://www.naturalcapitalproject.org/pubs/InVEST_2.0beta_Users_Guide.pdf, accessed 05-12-2014. Tallis, H., Mooney, H., Andelman, S., Balvanera, P., Cramer, W., Karp, D. et al. (2012). A Global System for Monitoring Ecosystem Service Change. Bioscience, 62, 977-986. Tett, P., Gowen, R.J., Painting, S.J., Elliott, M., Forster, R., Mills, D.K. et al. (2013). Framework for understanding marine ecosystem health. Marine Ecology Progress Series, 494, 1-27. Tittensor, D.P. et al. (2014) A mid-term analysis of progress toward international biodiversity targets. Science 346: 241-244. Trisurat, Y. 2013. Ecological Assessment: Assessing Condition and Trend of Ecosystem Service of Thadee watershed, Nakhon Si Thammarat Province (in Thai with English abstract). Final Report submitted to the ECO-BEST Project. Faculty of Forestry, Kasetsart University, Bangkok. UK NEA (2011). "The UK National Ecosystem Assessment: synthesis of the key findings." UNEP-WCMC, Cambridge, UK. UNEP GEO4: Global Environmental Outlook 4. (2007). United Nations Environment Program, Nairobi, Kenya. UNEP GEO5: Global Environmental Outlook 5. (2012). United Nations Environment Program, Nairobi, Kenya. Vackar, D., ten Brink, B., Loh, J., Baillie, J.E.M. & Reyers, B. (2012). Review of multispecies indices for monitoring human impacts on biodiversity. Ecological Indicators, 17, 58-67. van der Heijden, K. et al. (2002). The Sixth Sense: Accelerating Organisational Learning with Scenarios, Wiley & Sons, Chichester. Van Notten, P. (2005). Scenario development: a typology of approaches (Chapter 4). Available at: https://www1.oecd.org/Site/Schoolingfortomorrowknowledgebase/futuresthinking/Scenarios/37246 431.pdf. Last accessed December 1, 2014 van Oijen, M., Reyer, C., Bohn, F.J., Cameron, D.R., Deckmyn, G., Flechsig, M. et al. (2013). Bayesian calibration, comparison and averaging of six forest models, using data from Scots pine stands across Europe. For. Ecol. Manage., 289, 255-268. van Vuuren, D.P., Kok, M.T.J., Girod, B., Lucas, P.L. & de Vries, B. (2012). Scenarios in Global Environmental Assessments: Key characteristics and lessons for future use. Global Environmental Change-Human and Policy Dimensions, 22, 884-895. Walz, A., Lardelli, C., Behrendt, H., Gret-Regamey, A., Lundstrom, C., Kytzia, S. et al. (2007). Participatory scenario analysis for integrated regional modelling. Landscape and Urban Planning, 81, 114-131.
Page 34 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Williams, B.K. & Johnson, F.A. (2013). Confronting dynamics and uncertainty in optimal decision making for conservation. Environmental Research Letters, 8, 16.
Page 35 of 35
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
2 Using scenarios and models to inform decision-making in diverse policy, planning and management contexts 5
Coordinating Lead Authors: Lilibeth Acosta and Brendan Wintle Lead Authors: Zsófia Benedek, Purna Chhetri , Sheila JJ Heymans, Aliye Ceren Onur, Rosario Lilian Painter, Andriamandimbisoa Razafimpahanana and Kikuko Shoyama
Key messages: 10
15
20
25
30
35
40
Section 2.1 Decisions that impact on biodiversity and ecosystem services are numerous and diverse. A relatively small proportion of such decisions are explicitly considered ‘environmental’ decisions. Settings in which formal decision support using models and scenarios can play a part arise in policy agenda setting, policy development, planning, and management. The application of decision support approaches provides the opportunity to utilize biodiversity and ecosystem service scenarios and models in characterising cause-effect pathways and exploring the consequences of policy, planning and management options. The most appropriate decision support protocol and model of biodiversity or ecosystem services to apply in any given decision depends on the decision context. We develop a policy, planning and management decision context typology based on a set of attributes including spatial and temporal scale, political and cultural context, governance arrangements, the number of objectives considered. Section 2.2 A huge number of decision support protocols now exist that have been used in a wide variety of decision contexts, utilizing a wide variety of biodiversity and ecosystem service models and scenarios. On reviewing a large number of decision processes, we find that the bulk of documented applications of formal decision support approaches are undertaken at local-national scales. We find few documented examples of formal decision support being applied at regional (multi-national) scale and none at the global scale. We note that modelling and scenario analysis has been used effectively in setting policy agendas at regional and global scales, though formal, structured decision approaches to planning and implementing effective policy options are auspiciously absent. We also find that a key ingredient for successful application of structured decision support, models and scenarios is the dedication and continuity of involvement of decision support facilitators and modellers. A primary impediment to widespread and productive use of models and scenarios in policy, planning and management is a general lack of decision support, modelling and scenario analysis skills relative to the number of policy, planning and management processes. This may arise through a lack of appreciation among high-level decisions makers of the potential value of structured decisions approaches supported by scenarios and models in all three phases of the policy cycle. Relative to these problems, lack of data and models for biodiversity and ecosystem services is less of a problem.
Section 2.3 Page 1 of 60
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
Significant progress has been achieved in understanding impacts and feedbacks of state-variables across various organizational levels (i.e. genetic, species, ecosystem and landscape) using models and scenarios that are dynamic over space and time. However, most of the knowledge gained through data collection and modelling has not been prioritized with consideration to how it will be used in policy making, planning and management. The type, spatial and temporal extent and resolution, precision and accuracy of many BES models and scenarios do not match decision time frames, crossscale linkages among decision makers and stakeholders, and other jurisdictional needs of many decision makers. The knowledge gain necessary to support better environmental decisions can only be discovered by carefully considering the decision context in advance of data collection and modelling; preferably through the use of formal decision support paradigms such as those reviewed in 2.2. Section 2.4 Environmental problems and the process of finding technical and management solutions to these are challenged by stochastic, scientific and decision uncertainties with various levels of complexity and reducibility. These uncertainties, no matter how large or small, can have negative social, economic and ecological implications and thus need to be identified and addressed in policy, planning and management. There are now many formal methods for dealing with uncertainty. Technical approaches such as stochastic dynamic programming, formal implementations of adaptive management and robustness analyses provide solutions to a subset of problems that can be characterised mathematically. For decision contexts in which uncertainties are multidimensional, socially complex and dynamic, and have low reducibility structured deliberative methods that allow feedback and learning among decision-makers and stakeholders may be more appropriate. In many such instances, hybrid technical and deliberative approaches have been shown to provide outcomes that are both robust and socially acceptable.
Introduction 30
35
40
45
The process by which many decisions are made is unstructured. The most common method of organisational decision-making is through open dialogue in a committee setting. This may be entirely adequate for the many problems that involve small consequences and low complexity. However, even where detailed information and analyses are marshalled to support the meeting, unstructured conversation is prey to the frailties of ‘groupthink’, deference to authority, and a bias towards retaining the status quo (Burgman 2005). Meetings typically exceed the cognitive limits of the human brain. Psychologists have clearly demonstrated that our minds are incapable of processing more than about seven things at any one time. A committee discussion typically involves dozens of things, including issues, alternatives, pros, cons, objectives and criteria (Forman and Selly 2001). To the extent that they capture sound logic, formal decision protocols have advantages over unaided decision-making. Apart from buffering against cognitive limitations and negative group dynamics, a documented and traceable protocol will encourage decision-makers to be clear about judgments and assumptions (Bedford and Cooke 2001). Within structured decision making approaches, models and scenarios can play several important roles including; (i) transparently representing assumptions about cause-effect pathways that link policies and actions to outcomes, (ii) helping to reduce complexity by synthesising, analysing and representing multiple sources of information and evidence in a way that is most appropriate for the decision at hand, and (iii) helping to explore and identify Page 2 of 60
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
unforeseen consequences of policies and actions. This chapter aims to inform the IPBES plenary, and a broader audience of non-specialist scientists, policy makers, planners and managers about the possibilities and opportunities for using models and scenarios in decision making. We begin by introducing key concepts and some widely applied decision support protocols that have been used to support decisions in a policy, planning and management contexts. We highlight the role, or potential role, of biodiversity and ecosystem services modelling and scenario approaches within each of the decision paradigms reviewed and discuss some of the key challenges to widespread application of models and scenarios in real-world decision making. This chapter sets the policy and decision making scene for the three other chapters of this deliverable that provide more detail on model and scenario approaches (Chapters 3, 4, and 5). This chapter also provides the foundation for Chapter 7 to highlight capacity building needs in the area of model-supported decision analysis, and for Chapter 8 which highlights future developments which will undoubtedly include increased used of models and scenarios in decision making and the challenges of making that happen. Finally, this chapter sets the scene for Deliverable 4c which will develop an online catalogue of policy support tools and methodologies relevant to IPBES related activities. In the remainder of 2.1 we provide a broad outline of decision making contexts, providing a typology of decision contexts and identifying key axes that can be used to characterise any given decision. In 2.2 we review some of the most widely applied decision support approaches where scenarios and models have been used, providing some case studies to highlight some strengths and weaknesses for any given decision context. In 2.3 we discuss principles for aligning decision approaches, models and scenarios with the needs of decision makers. In 2.4 we discuss the particular challenge of dealing with model and scenario uncertainty in decision making and in 2.5 we conclude by sketching out a path toward increasing and improving the use of models and scenarios in policy, planning and management decision making.
2.1 Overview of policy, planning and management contexts in which scenarios and models support decision-making Almost every policy, plan and action in every sector from health to manufacturing, at every spatial and organizational scale from the individual, to the globe impacts in some way on biodiversity and ecosystem services. The number and types of decisions made appear to defy classification; and are practically infinite. We can only describe or characterise such an immense array of decisions in the coarsest of terms. The bulk of decisions or choices made on a daily basis that impact most on biodiversity and ecosystem services are seldom described or even conceived of as ‘environmental’ decisions. Almost all of them are undertaken by people outside the ‘environment sector’ with little or no consultation with environmental professionals. It is our goal to broadly categorize decisions that impact on biodiversity and ecosystem services according to a few key axes. The point of this exercise is to try and reduce some of the complexity and confusion about the range of decision protocols that helps support decisions that impact on biodiversity and ecosystem services and the contexts in which they might be most fruitfully applied. Having mapped out the decision space and scoped some potentially useful decision support approaches, we then hope to draw a link between the decision contexts, the relevant tools and the role that models and scenarios can play. Page 3 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
2.1.1 Policy, planning and management 5
10
15
20
25
30
35
40
Within the relatively small scope of decisions that are made with specific regard to biodiversity and ecosystem services, three broad types of structured decision making context can be identified; policy development and implementation, planning, and management. A policy is a principle, usually stemming from an articulation of a group or individual’s values that guide actions. For example, a primary school may have an anti-bullying policy that makes clear that bullying is not acceptable to the school and may set out procedures for dealing with accusations or occurrences of bullying. Governments and large public or private organizations commonly design policies to encourage economic prosperity or wellbeing that have implications for biodiversity and ecosystem services. Equally, policies can be developed specifically to address biodiversity or ecosystem service issues or problems that themselves may have impacts on other values such as opportunities for wealth generation via agriculture or natural resource exploitation. Consequently, there is significant attention given to understanding the broader implications of any policy prior to, and after its development and implementation. Formal decision protocols, supported by biodiversity and ecosystem service models or scenarios have been used to support policy development and implementation. A key component of policy development is policy ‘agenda’ setting (Fig. 2.1.1). This is an area of policy development in which scenarios and models have had a key role (e.g. MEA 2005, Alkemade et al. 2009, Leadley et al. 2014). Planning is the process of prioritizing and scheduling actions in order to achieve individual or organizational goals. For example, land-use planning can be undertaken to identify which activities will be allowed or encouraged in particular parts of the landscape in order to achieve landscape-level objectives for a range of criteria such as agricultural productivity, tourism service provision, and biodiversity conservation (e.g. www.agriculture.gov.au/forestry/policies/rfa). Formal decision protocols and models of biodiversity and ecosystem services have been used to support land-use planning at multiple planning scales (SAPM 2009). Management decision making occurs in such a wide variety of management contexts, at every spatial and temporal scale that a formal definition of management decision making seems trivial. In complex instances of management decision making that involve highly uncertain benefits and costs due to complex ecological or social system dynamics, and multiple criteria for measuring success, formal decision protocols can be extremely useful for ensuring that management is effective and efficient in meeting objectives for biodiversity, ecosystem service, and other criteria (Runge et al. 2011a). Policy makers, planners and managers draw on a vast array of theoretical frameworks and practical guidance. For example, a plethora of published frameworks for adaptive policy, adaptive planning and adaptive management exist that describe very similar steps and approaches for dealing with uncertainty and complexity in policy development and implementation, planning and management (Fig. 2.1.1). In most representations of policy, planning, and management cycles, three broad phases consistently appear; (i) formulation, (ii) implementation, and (iii) evaluation. Individual contexts and world views generate a multitude of variations and subdivisions of these three phases which allow more specific advice to be provided about how each of the key steps should proceed. Contextdependant variations on decision support protocols, and the models and scenarios used to support them are critical in tailoring the method to the need, as we will see in the following sections.
Page 4 of 60
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
Figure 2.1.1 A theoretical framework for policy formulation, implementation and evaluation (adapted from Howlett et al. 2009) identifying the activities most likely to utilize models and scenarios. The policy cycle is frequently described as iterative and similar to adaptive planning and management (McFadden et al. 2011, Walters 1986).
2.1.2 Attributes that define the decision context 10
15
20
25
30
Policy, planning and management decisions may be one-off interventions in response to a simple problem to which there is high certainty about the most appropriate solution. However, most environmental decisions are characterized by high uncertainty due to a lack of knowledge about cause and effect relationships (Ludwig et al. 2001). The degree and type of uncertainty inherent in a particular decision problem determines the sorts of analytical and decisions support approaches that can be applied (Peterson et al. 2003, Regan et al. 2005) and partly motivate the need for models and scenarios. We deal with the role and implication of uncertainty in decision making, modelling and scenarios in section 2.4.1. Decisions that impact on biodiversity and ecosystem services take place across many spatial and temporal scales, jurisdictions and administrative contexts; from the largest most complex systems involving many actors, cultural perspectives, values and services, to problems involving relatively few actors and species, and services at the local level (Table 2.1.2). The scale of the policies, plans and actions will vary depending on the relevant drivers of change, both direct and indirect. Matching the response to the scale of the problem and drivers and ensuring that multiple responses do not create conflict can be a huge challenge. Biodiversity and ecosystem services have specific spatial and temporal distributions that overlap with human management units or jurisdictions in complex ways. Various stakeholders have rights, obligations and interests at a variety of spatial scales. Global responses to ecosystem problems are warranted when those problems potentially affect all people and ecosystems of the world, in other words common pool resources. Multilateral, regional and bilateral agreements require consensus by a group of nations but implementation often requires action within national boundaries. Specific national policies also exist independently of agreements with other nations because of specific national priorities, highlighting the problem of policies and plans that conflict across scales. The scale at which human and biotic processes operate influences the sorts of decision approaches, models and scenarios that are relevant to a particular decision. Page 5 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Spatial scale partly determines who will be represented in a decision problem and whose interests are considered. Table 2.1.2 Variables that define a decision context and how they vary.
Axes Political scale Cultural context (value/knowledge system) Geography/ecology Flows across landscape Temporal dynamics Decision process - Objectives - Stakeholders - Temporal scope
From Local Homogenous Single species Single ecosystem Short term Participatory (multiagent) Single objective Single Once-off
To Regional Diverse Multi species Linked ecosystems Long term Top down (singleagent) Multi objective Multiple Sequential
5
10
15
20
25
30
Many environmental problems are characterised by high social and cultural complexity, leading to divergent views about the appropriateness of policies, plans or actions and criteria for what constitutes a good outcome. Differences in capacities and power determine the effectiveness of stakeholder representation and the acceptability of decision outcomes. Large and wealthy organizations, including companies and national governments may have greater resources and better access to information than other stakeholders, leading to a greater influence over the decision process. Assessing the impacts on livelihoods of policies, plans and management options may require culturally specific, local level understanding to properly evaluate costs and benefits to all stakeholders (Runge et al. 2011a, Nordström et al. 2010, Rowland et al. 2014). Cultural norms, values, practices, ideologies and customs shape people’s understanding of their needs, rights, roles, possibilities and hence on their actions, including engagement in policy development and implementation (Borrini-Feyerabend et al. 2004). All stakeholders use their beliefs as the basis for determining the range of options they will consider and the criteria by which they will measure outcomes. The importance of taking into account multiple belief systems during policy formulation is being increasingly recognized especially in areas where indigenous people have consolidated their property and representation rights (Tauli-Corpuz et al. 2010, UNDRIP 2007, Runge et al. 2011a). Some ‘local’ decisions take place within a particular ecosystem or geographic domain that can be considered for the purposes of the decision process discrete and sufficiently buffered from the ecological processes playing out in other systems, so as to simplify the characterisation of biodiversity and ecosystem service values and dynamics. However, many land-use planning and policy processes play out over multiple ecosystems that are connected by complex flows of biotic and abiotic resources, and which are subject to multiple different types of ecological and social dynamics that may play out of multiple temporal scales. For example, some integrated catchment management strategies must consider simultaneously terrestrial, river, estuarine and near-shore ocean ecosystems, each with unique economic drivers and pressures such as agriculture, aquaculture, and fishing (e.g. Brodie et al. 2012).
Page 6 of 60
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
The governance system under which decisions are made, and the degree to which power over a given decision is shared among actors contribute significantly to the nature of the decision approach taken and the types of decision support, models and scenarios likely to be relevant. For example, socalled ‘top-down’, ‘single-actor’ decision problems may be amenable to application of economic optimization approaches, while more ‘participatory’, ‘multi-actor’ decision structures may be much less amenable to such approaches. Sequential decision processes provide the opportunity to value the role of learning and to establish formal programs of ‘continuous improvement’, often invoking ideas embodied in adaptive management (Walters 1986). However, with this opportunity comes complexity. Many reasons have been proposed for the auspicious lack of working examples of adaptive management (AM) in broad scale, multi-objective decision problems (Walters 2007, Wintle and Lindenmayer 2008). The complexity associated with achieving a working adaptive management process in all but the simplest of resource management problems may be partly to blame. In the remainder of this chapter we will explore how the attributes of decision context influence the choice of decision analysis and support approach that is most appropriate and how this, in turn influences the types of models and scenarios that are used.
2.2 Lessons learnt from application of established methodological paradigms and frameworks of scenarios and models 20
25
30
35
There are many established paradigms and approaches to support environmental policy, planning and management decisions ranging from specific technical tools within a very specific domain of application such as mathematical optimization approaches, through to broad frameworks such as ‘structured decision making’ (sensu Gregory et al. 2012) that set out prescriptions for dealing with most challenges confronting practitioners charged with managing a complex decision problem. In many instances, the application of more general approaches to decision support includes application of some specific tools such as models and scenarios. For example, it is very common for a ‘structured decision making’ exercise to include some multi-criteria analysis (sensu Hajkowicz et al. 2000, Hajkowicz 2008), perhaps supported by some economic cost-benefit analysis. Here we provide a nonexhaustive overview of the main classes of environmental decision support approaches in a rough order of most specific to most general. We start with a description of the most commonly documented approaches, including a case study of the application of each and a general assessment of strengths and weaknesses. We then generalize these approaches into a coarse typology based on the attributes of decision problems described in the previous section. We conclude with a discussion about which approaches seem most amenable to the use of models and scenarios and how models and scenarios might be better integrated with existing decision making approaches.
2.2.1 Single objective risk analysis methods 40
Risk analysis models encourage decision-making on the basis of expected consequences. That is, the calculation of risk as the product of likelihood and consequence is essentially an estimate of expected (dis)utility (Savage 1954). While consideration of adverse consequences alone will often suggest the desirability of avoidance or mitigation measures, conditioning estimates of consequence with assessment of likelihood may imply that such measures are not warranted. If estimates of likelihood and consequence are unbiased, then decisions based on risk should lead to more effective allocation of resources (Arrow and Lind 1976). Page 7 of 60
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
The real world challenges of decision making are seldom so simple. Consequences are not restricted to impacts that can naturally or readily be described in single criterion (e.g. monetary) terms. Multiple values imply multiple objectives each requiring estimates of expected consequence. Uncertainty about consequences and likelihoods brings into play complex risk preferences that must be considered. All formally considered decisions involve alternatives and cause-and-effect predictions of expected consequence. When predictions are made over multiple objectives, an additional element is required to resolve the decision problem: the articulation of preferences or trade-offs reflecting the relative importance of the different objectives (Howard 2007).
10
Most single attribute risk management tools do not directly treat trade-offs among competing objectives. A subset of these tools may be helpful in understanding cause-and-effect for estimating expected consequences for individual options or objectives, but on their own, they will generally be inadequate for making most real-world decisions which tend to involve trade-offs.
15
2.2.2 Consequence tables
20
25
Most environmental policy, planning and management decisions involve trade-offs (Keeney 2002). Consequence tables are the first of the multi-objective decision support tools described here to deal explicitly with trade-offs. There are three core elements to any multi-objective decision problem; alternatives, expected consequences, and trade-offs. These elements are compactly reported in a consequence table. An example is shown below (Table 2.2.2), where alternatives comprise six hypothetical candidate options for reducing impacts on a near-shore reef system resulting from nutrient outflow from an agricultural catchment. The table can be populated with qualitative or quantitative estimates of expected consequence. Experts and non-expert stakeholders alike are notoriously deficient in their capacity to make internally consistent probabilistic judgments (Hastie and Dawes 2010). Modelling tools that assist in the coherent treatment of probabilities include fault tree analysis, event tree analysis, Markov analysis, Monte Carlo simulation and Bayes nets. For example, Jellinek et al. (2014) developed a Bayes net to explore the relative improvement in vegetation condition resulting from a range of investments in woodland management including reducing stock grazing and direct restoration activities.
30 Table 2.2.2 The example below uses coarse verbal (negative) impact descriptors typically seen in a qualitative risk matrix approach. Trade-offs involve consideration of the performance of each alternative against each objective.
35
Example objectives
do nothing A1
A2
A3
A4
A5
A6
Biodiversity - fish Biodiversity - coral Economic impact Costs to implement Recreational fishing Tourism
High Extreme Low Low High High
High High Medium Low Medium High
Medium Medium Medium Low Medium Medium
Medium Medium Medium High Medium Medium
Low Low High Extreme Medium Low
Low Low Extreme High Low Low
High High Low Medium Medium High
The preparation of a consequence table itself offers substantial insulation against the pitfalls of unaided decision-making. But unless the decision problem can be meaningfully simplified to two or three objectives and two or three alternatives, the cognitive and emotional demands on decisionPage 8 of 60
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
makers and stakeholders can lead to poor outcomes. In many instances, a consequence table can be simplified through identification of the strictly non-dominated set of alternatives, and consideration of practically dominated alternatives and redundant objectives. An alternative is strictly dominated if, in comparison with any other single alternative, it performs worse on at least one objective and no better on any other objective. Driscoll et al. (2010) illustrate identification of the non-dominated set in a hypothetical trade-off between asset protection and biodiversity conservation in the context of wildfire management.
2.2.3 Benefit-cost analysis 10
15
If all expected consequences can be assigned a monetary value then benefit-cost analysis is applicable. Selection of the alternative with the highest benefit-cost ratio has a strong basis in public policy and welfare economics. However, the monetisation of non-market impacts is difficult. Where revealed preferences are deemed inadequate or absent, robust techniques for stated preference are available (Bennett and Blamey 2001), but the time and resources required to apply these methods are substantial. In any case, stakeholders are unlikely to feel comfortable with monetisation of all objectives, especially those dealing with social and environmental outcomes. In addition, the impact of discounting over time becomes more difficult when dealing with future scenarios.
2.2.4 Programming methods for decision support 20
25
30
35
40
There are potentially thousands of alternative options in most realistic planning and management decision problems. Various mathematical programming techniques from the field of Operations Research are available to help identify better (or best) candidates from a large set. Linear Programming (LP) and Stochastic Dynamic Programming (SDP) employ algorithms designed to optimise some objective function under specified constraints (Chankong and Haimes 2008). In LP, a static linear relationship (or near-linear) relationship between actions and expected consequences is required. This may be inappropriate in many ecosystems, where outcomes for objectives are dynamic and non-linear in relation to actions or sets of actions. With detailed understanding of cause-and-effect, SDP can accommodate non-linear, dynamic outcomes associated with the stochastic risk (e.g. risks associated with wildfires) superimposed on the deterministic influence of management actions (e.g. fuel reduction burning in high fire risk places). SDP recognises that what might be considered a desirable action depends on the state of the system (Minas et al. 2012, Richards et al. 1999). For example, low fuel loads and a forest age structure skewed towards regrowth may imply lesser need for burning compared to circumstances where fuel loads are high and old growth is relatively well represented. The capacity to capture greater realism in SDP is attractive, but computational overheads and the requirement for sophisticated causal understanding mean that most applications are substantially simplified. Goal programming (GP) avoids the naive binary logic of a step function that is commonly employed in setting objectives for LP and SDP. GP requires specification of a performance aspiration for each objective. The underlying algorithm searches among the candidates for the alternative having the minimum multi-dimensional distance to the goal set (Chankong and Haimes 2008). Conceptually, the method could be used profitably by a single decision-maker. In a multi-stakeholder setting, GP is
Page 9 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
open to abuse because stakeholders will tend to manipulate outcomes through articulation of insincere positions on what might be considered an appropriate goal for each objective.
2.2.5 Multi Criteria Decision Analysis (MCDA) 5
10
Maguire (2004) cites two interacting flaws commonly encountered in risk-based decision support: (a) incoherent treatment of the essential connections between social values and the scientific knowledge necessary to predict the likely impacts of management actions, and (b) relying on expert judgment about risk framed in qualitative and value-laden terms, inadvertently mixing the expert’s judgment about what is likely to happen with personal or political preferences. The family of techniques under the banner of MCDA seek to avoid these flaws through explicit separation of the task of causal judgment from the task of articulating value judgments or trade-offs (Ananda and Herath 2009).
2.2.5.1 Multi-attribute value theory (MAVT) 15
The technical description of MAVT is provided by Bedford and Cooke (2001) and Keeney (2007). The task of MAVT is to find a simple expression for the decision-maker’s value function over two or more n
relevant attributes (i.e. objectives and associated criteria): v(x1,…,xn) = ∑ w i v i (x i ) i =1
20
25
30
35
40
where the wi are the weights and the vi are value functions for any single attribute. Weighting of the individual value functions can be done formally by the method of indifferences, akin to the underpinnings of stated preference techniques used in evaluation of non-market impacts in benefitcost analysis (Bennett and Blamey 2001). There are many shortcut methods for eliciting weights (Hajkowicz et al. 2000). Of these, the swing weight method has been shown to be one of the more effective, both in terms of its efficiency and its insulation against abuse (Fischer 1995). Whatever method is used in their elicitation, the interpretation of the weights is critical. Methods that do not explicitly deal with indifferences are prey to abuse. Users are inclined to specify weights that reflect the relative importance of the attributes, irrespective of the units or the range of consequences relevant to the decision context. But the weights have units because the underlying attribute scales have units. Changing the units or range of an attribute must lead to a change in the weights. For the additive value model to be valid the attributes need to be mutually preferentially independent. In practice, the assumption of preferential independence is reasonable if the set of objectives is consistent with the following properties (Keeney 2007): • Complete – all of the important consequences of alternatives in a decision context can be adequately described in terms of the set of fundamental objectives. • Non-redundant – the fundamental objectives should not include overlapping concerns. • Concise – the number of objectives should be minimal. • Specific – each objective should be specific enough so that consequences of concern are clear and attributes can readily be selected or defined. • Understandable – any interested individual knows what is meant by the objectives. Where objectives satisfy these properties there is a strong case for use of simple weighted summation. While the analyst needs to be careful to ensure preferential independence, the mechanics of MAVT are straight-forward. Arithmetic operations are simple and easy to implement in a spreadsheet. Strictly speaking, MAVT is applicable where there is no uncertainty in the estimation Page 10 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
of consequences or where decision-makers and stakeholders can be assumed to be risk-neutral. These assumptions are unrealistic in the context of the SFMS project.
2.2.5.2 Analytic Hierarchy Process (AHP) 5
10
15
20
25
30
AHP is commonly encountered in MCDA applications in the natural resource management literature (Mendoza and Martins 2006). It is essentially a variant of MAVT designed to minimise the elicitation burden on experts and decision-makers. Most applications employ the same additive value model described above for MAVT. Using a nine point preference scale and matrix computations to translate ordinal judgments into cardinal judgments, (a) marginal value functions and (b) weights, are derived through pairwise comparisons of alternatives and objectives, respectively (Saaty 1980). A variety of software packages are available, although for simple problems the calculations can be done in a spreadsheet. AHP’s strength in minimising elicitation burden is also its weakness. It’s possible to obtain marginal value functions without any explicit estimation of consequences. For decision problems involving self-evident cause-and effect relationships this may be acceptable. This may fall down when consequences of alternative options involve difficult probabilistic judgments that are likely to be logically challenging (Hastie and Dawes 2010). AHP has also been criticised on theoretical grounds because it allows rank reversal upon introduction of a new alternative - a violation of decision theory’s independence of irrelevant alternatives axiom (von Neumann and Morgenstern 1944). The modified AHP (mAHP) is free of this problem. It uses standard MAVT techniques to obtain marginal value functions, and limits the use of pairwise comparisons to the derivation of weights. Moffett and Sarkat (2006) advocate use of mAHP because of the relative ease of obtaining weights. But like direct weighting, weights obtained through pairwise comparisons via mAHP result in poor capture of stakeholder preferences. In general respondents tend to assign weights according to the perceived importance of objectives, irrespective of the consequences associated with the specific alternatives being considered. Any weighting technique that fails to promote normative interpretation of weights through explicit consideration of the range of consequences is inadequate (Steele et al. 2009).
2.2.5.3 Outranking
35
40
Outranking techniques stem from the French school of MCDA, which places less emphasis on normative understanding of how decisions should be made based on axioms of rationality (von Neumann and Morgenstern 1944) and greater emphasis on behavioural models of decision-making (Roy 1973). Outranking techniques typically involve sequential elimination of alternatives (Chankong and Haimes 2008). Weights are assigned to each objective according to their perceived importance, without consideration of the range of consequences associated with alternatives. For each pair of alternatives a concordance index and a discordance index are constructed. The concordance index coarsely characterises the strength of the argument that one alternative is better than another based on the weighted sum of objectives for which it dominates the other. The discordance index reports the strength of the argument against eliminating the (weakly) dominated alternative. Decision-
Page 11 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
makers work through a consequence table iteratively, adjusting critical thresholds for concordance and discordance until a satisfactory choice is made.
5
10
There are numerous techniques and software packages that fall under the banner of outranking (e.g. ELECTRE, PROMETHEE, GAIA; see Figueira et al. 2005 for details). The techniques vary according to how expected consequences are characterised. If a consequence table is populated using qualitative ordinal descriptors of impact (as in section 2.3) ELECTRE can informally assist stakeholders progress trade-offs and difficult decisions involving more than a handful of objectives and alternatives. While other outranking techniques can be used where consequence estimates are quantitative or semiquantitative, there is little argument for doing so, because in these circumstances MAVT offers a much firmer normative basis for decision-making.
2.2.5.4 Multi-attribute utility theory (MAUT) 15
20
25
The formal description of MAUT developed by von Neumann and Morgenstern (1944) was (and remains) a high point in the theory of decision-making. It is also a wholly impractical approach to typical multi-objective, multi-stakeholder problems. Many of the developments and refinements of MCDA since the 1950s are essentially pragmatic short-cuts for MAUT. MAUT can be used when a consequence table is populated by statistical distributions describing probabilistic uncertainty in the performance of each alternative against each objective. These circumstances are rare indeed, especially in natural resource management. Aside from difficulties in obtaining detailed probabilistic causal judgments, there are distinctly onerous demands on decisionmakers and stakeholders in the elicitation of trade-offs under MAUT. Populating a consequence table with probabilistic outcomes clearly defines a strong role for models and scenarios. In practice, only the most committed and indefatigable participants in a group decision-making settings are capable of formally addressing trade-offs using MAUT. Box 2.2.5 MCDA case study – the use of a web-based MCDA system in participatory environmental decision making in Finland Mustajoki et al. (2004) describe the use of MCDA to planning for multiple uses of the Paijanne Lake, Finland’s second largest lake. The lake has been regulated since 1964, with the original objectives being to increase hydropower production and to decrease agricultural flood damages. The lake has extensive recreational housing developments along the shores and there are tens of thousands of recreational users and fishermen on the lake. There has been growing public interest to reconsider the regulation policy to better take into account the increased recreational use and current high environmental awareness. Problems currently recognized on the lake include the low water levels during spring, changes in the littoral zone vegetation and negative impacts of the regulation on the reproduction of fish stocks. An extensive multi-disciplinary research project was carried out in 1995–1999 to re-evaluate the regulation policy of the lake. The aims of the project were to assess the ecological, economic and social impacts of the regulation. Stakeholder opinions were sought about the current regulation and its development, comparison of new regulation policy options, and recommendations to diminish the harmful impacts of the regulation. An open and participatory planning process was considered necessary to gain public support for the project and to find consensus on a new regulation strategy. A steering group consisting of 18 representatives of different stakeholders was set by Ministry of Agriculture and Forestry, the permit holder of the regulation license. Additionally, four working groups were established to Page 12 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
improve the communication between the water resource authorities, local stakeholders, experts on regulation, and researchers. To inform the public, a local press conference was arranged after almost every steering group meeting.
2.2.6 Adaptive Management 5
10
15
20
Adaptive management is a framework by which prudent management actions can be selected in the face of uncertainty and via which uncertainties can be reduced through learning by doing (Walters 1986). Although originally devised for the purpose of fisheries management (Walters and Hilborn 1976), it has been applied to a variety of environmental management issues such as setting waterfowl hunting regulation (Johnson et al. 1997), translocation of threatened species (Rout et al. 2009), learning about forest management options (Wintle and Lindenamayer 2008), pest management (Shea et al. 2002), wildlife disease management (McDonald-Madden et al. 2010b), and vegetation restoration (McCarthy & Possingham 2007). The process of adaptive management requires that stakeholders clearly articulate their objectives for management, so that the actions selected are those that will best achieve those objectives over the specified time frame. Uncertainty can be characterised as multiple hypotheses/scenarios or models that link actions to outcomes. Regular monitoring is necessary to verify that actions are implemented as intended, to allow rapid response to new knowledge and changing conditions. Monitoring also offers the opportunity to systematically update knowledge, reassess hypotheses and adjust the action strategy accordingly. The role of monitoring in actively and passively learning is central to the concept of adaptive management (Runge 2011, Figure 2.2.6). The other steps common to many of the decision frameworks discussed in this chapter include problem framing, objective elicitation, alternatives (options) development, consequence evaluation, and decision-making, and implementation.
Set-up phase
Proble
Elicit
Evaluat
Iterative Learn phase
Develo Decide
Evaluat e
Imple
Monit
25 Figure 2.2.6 An adaptive management framework for threatened and endangered species management (Adapted from Runge 2011 and Lahoz-Monfort et al. 2014).
Box 2.2.6 Adaptive management case study – Management of North American Mallard ducks Nichols and Williams (2006) summarise an adaptive management program that has been working for 10 years to support the management (hunting regulations) of mid-continent Mallard ducks in Page 13 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
North America. Management objectives are to maximize cumulative harvest over a long time period (including the devaluation of harvest when predicted population size falls below the North American Waterfowl Management Plan goal threshold of 8.8 million breeding mallards). Management actions include four regulatory packages specifying daily bag limits and season lengths for each of the four major North American flyways. Four models (scenarios) of system response to harvest management are included in the model set. These models reflect two different hypotheses about the effect of hunting mortality on annual duck survival (compensatory mortality reflecting minimal effects of hunting and additive mortality reflecting maximal effects of hunting mortality), and two hypotheses about the strength of density-dependent relationships defining reproductive rates (weakly and strongly density-dependent). At the initiation of this management process in 1995, all four models (representing all possible combinations of these four hypotheses) were given equal credibility weights of 0.25, indicating no greater faith in the predictions of one model than in those of any other. A complex monitoring program is used to estimate breeding population size and number of wetlands in Prairie Canada (an important environmental covariate), rates of survival and harvest, and preseason age ratio. In each spring, the new estimate of population size is compared against predictions made the previous spring corresponding to each of the four models. These comparisons are combined with the model weights from the previous year to update the weights. Learning thus occurs when weights become large for some models, giving them more credibility and thus more influence in the decision process, and small for others. The decision about which set of harvest regulations to implement depends on system state, as defined by estimated numbers of ducks and ponds.
2.2.7 Scenario planning 5
10
15
20
Scenarios are plausible and internally consistent, but not necessarily probable, futures (Schwartz 1996). Börjeson et al. (2006) refer to scenario planning as a tool for exploring possible, probable and/or preferable futures. They divide scenarios into three main categories; these are predictive which tries to respond the question “what will happen?”, explorative which tries to respond “what can happen?” and normative scenarios “How can a specific target be reached?”. Unlike forecasting, the focus of scenario planning is not to assess the probability of future events, rather, it explores possible futures that may arise under different conditions, and what those different futures might mean for current decisions (Schoemaker 1995). Assumptions about future events or trends are questioned, and uncertainties are made explicit (Bohensky et al. 2006). Scenario planning typically takes place in a workshop setting. Participants explore current trends, drivers of change, key uncertainties, and how these factors might interact to influence the future (Schoemaker 1993). To do so, they draw on both qualitative and quantitative information, including datasets (WCS & Bio-Era 2007), spatially explicit data (Santelmann et al. 2004), and expert/stakeholder judgement (Schoemaker 1993). Based on this information, a set of plausible future scenarios is developed. Participants then consider a range of policy or response options, and assess how robust those options are to the different scenarios developed. One of the early applications of scenario planning was to navigate an oil crisis in the 1970s. Shell Oil identified a plausible scenario – considered unlikely and rather radical – where a coalition of oil exporting countries limited production and drove prices up (Peterson et al. 2003). The company adjusted its business practices to buffer itself against this scenario, by making its shipping and refining processes more efficient. In recent years, there have been many applications of scenario Page 14 of 60
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
analysis on a landscape scale (Steinitz et al. 2003, Baker et al. 2004, Berger and Bolte 2004, Hulse et al. 2004, Santelmann et al. 2004, Shearer 2005, Walz et al. 2007). Many studies have defined landscape scale scenarios using qualitative techniques based on participation of stakeholder groups (Hulse et al. 2004; Patel et al. 2007). Others have combined participatory approaches with quantitative scenario modelling in the analysis of landscape futures (Walz et al. 2007). Quantitative modelling techniques have included spatial multi-criteria analysis (Pettit and Pullar 2004, Berger 2006), agent-based modelling (Happe et al. 2006), actor-based modelling (Bolte et al. 2004), and integrated assessment and modelling (Carmichael et al. 2004, van Ittersum et al. 2008, Wei et al. 2009). Some studies (Liu et al. 2007, Meyer and Grabaum 2008) have found optimisation and scenario analysis to be a valuable combination for selecting land use and management alternatives under uncertainty. A larger application of scenario analysis for environmental planning can be seen in the ‘Millennium Ecosystem Assessment’ (MEA, 2005), which explored the impact of changing ecosystems on human well-being. This project was ambitious. Three of four detailed scenarios examined in the MEA suggest that “significant changes in policies, institutions, and practices could mitigate some but not all of the negative consequences of growing pressures on ecosystems, but the changes required are substantial and are not currently under way” (MEA, 2005). Given the breadth, magnitude and coarseness of changes suggested in the report, the assessment was not easily translated into finer scale policy and on-ground decision making. While the MEA findings were successfully integrated into the global programs of the Convention on Biological Diversity, the Ramsar Convention on Wetlands, and the Convention to Combat Desertification, recommendations were less relevant and effective at a local scale, which is where the most tangible changes happen (Tallis and Kareiva 2006). Strengths and weaknesses of Scenario Planning – Developing plausible scenarios helps us take the long view in a world of great uncertainty (Schwartz 1991, Huntley et al. 2010). Scenarios are narratives of the future defined around a set of unpredictable drivers, intended to expand insight by identifying unexpected but important possible directions and outcomes. Scenarios have a timeline over which meaningful change is possible. Scenarios help to develop the means to work towards preferred futures (Huntley et al. 2010, van der Heijden 2005). They are used in long-range planning and the development of robust plans and encompass a broad span of future possibilities so the future can be met with some degree of confidence. No scenario is ever seen as absolute, as the probability of any scenario being realized is inconceivably small (Stone and Redmer 2006). A challenge of scenario planning is to determine the real needs of corporate leaders and managers. Often, they may not know what they need to know, or may not know how to describe the information that they really want. A value of scenario planning is that leaders can make mistakes and learn from mistakes without risking important and costly failures in real life (Stone and Redmer 2006). They can make these mistakes in a pleasant, unthreatening, gamelike environment, while responding to a wide variety of concretely presented scenarios based on facts. Scenario development often happens in intervention-style processes that may not be tailored to diverse stakeholder demands, or compatible with the need to develop longer-term adaptation pathways that deal with evolving multidimensional challenges because they focus too much on single actions (Berkes and Folke 2002, Wilkinson and Eidinow 2008, Vervoot et al. 2014).
45 Page 15 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Box 2.2.7 Case study – Scenario Planning in the Hudson River Estuary Watershed How a scenario planning exercise typically unfolds in practice is best explored with an example. One such example is The Nature Conservancy’s (TNC) 2008 effort to help communities within the Hudson River Estuary Watershed, USA, prepare for the impacts of climate change (see also Cook et al. 2014b for further analysis). In a series of workshops over the course of 18 months, more than 160 stakeholders were consulted, including railroad executives, utility companies, the insurance industry, emergency and health groups, planners and conservation leaders. They identified and discussed important drivers (e.g., land use trends, the political climate) and key uncertainties around those drivers (e.g., will there be strong “top-down” political support for climate change adaptation?). By manipulating these uncertainties and trends, they created four plausible scenarios. Scenarios were described using suggestive titles (e.g., Stagflation Rules) and narrative details, e.g. “the early years of the scenario witness low to negative economic growth, falling real estate values and little new development in the region...” (p. 6). Different elements of each scenario were specified, for example, the projections for the price of gas under the Procrastination Blues scenario were “decline from $3.80 to $2.05 from 2008-2011, then rise rapidly back to $5.00/gal by 2016...”. The feasibility of different policies or response options (e.g. changing the requirements for new storm water permits) could then be evaluated, in terms of both the likelihood that they would be adopted in each scenario and how they would perform in each scenario. The ‘top performing’ options were those that scored relatively highly across the four scenarios. The TNC project provides a good example of scenario planning because it was well defined on many fronts. As a starting point, focussing on the Hudson River Estuary Watershed provided clear geographic scope—but beyond this—the drivers explored were also well defined and easily monitored (e.g. the price of gas), meaning that trends within different scenarios could be explicitly and realistically quantified. Moreover, the different response options that were evaluated were specific enough to be implemented on the ground, for example, developing emergency actions plans with community involvement.
2.2.8 Structured decision making (SDM) 5
10
15
Like MCDA and AH approaches reviewed in section 2.2.5, Structured Decision Making (SDM; Gregory et al. 2012) derives from multi-attribute utility theory (MAUT; Raiffa 1968). However, SDM also draws heavily on more recent developments in Decision Analysis (Raiffa 1968, Keeney 1982, Hammond et al. 1999) and psychology (Kahneman and Tversky 2000). SDM is an organized approach to identifying and evaluating creative options and making choices in complex decision situations. Gregory et al. (2012) define SDM as “the collaborative and facilitated application of multiple objective decision making and group deliberation methods”. SDM is designed to deliver insight to decision makers about how well their objectives may be satisfied by potential alternative courses of action. It helps find acceptable solutions across groups, and clarifies divergent values which may underpin irreducible trade-offs. SDM is a very general approach to decision support (Fig. 2.2.8), which can conceivably be applied to any environmental decision problem at any scale and any level of social and institutional complexity. However, it is the value of SDM in situations in which there are conflicting values and conflicting views about the consequences of various courses of action due to uncertainty that differentiate it from the simpler analytical (or ‘normative’) approaches (Gregory et al. 2012). The attributes of SDM that distinguish it from the more technocratic approaches such as MCA is the emphasis placed on the understanding and dealing with difficult group dynamics through Page 16 of 60
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
a collaborative approach to clarifying objectives, exploring cause and effect relationships and dealing with contentious trade-offs. To some extent, the application of SDM formalizes or prescribes an approach to dealing with the ‘human’ elements, including judgement biases, group dynamics, and risk preferences, in decision making. The more technical support tools such as MCDA may be used in an SDM process where they add value or clarity to the process, but the SDM process is not centred on the use of such tool.
Figure 2.2.8 Six basic steps in Structured Decision Making (Adapted from Gregory et al. 2012).
10
15
20
25
30
There are six basic steps identified in the structure decision making framework (Fig. 2.2.8, Gregory et al. 2012). Clarifying or scoping the decision context involves identifying what the decision is about, which decision(s) will be made, by whom, and when. The spatial and temporal scale over which the decision applies is a key component of clarifying the decision context. Defining objectives and performance measures is a big focus of the SDM approach, which defines what matters in the decision context and how these things will be measured. Objectives and performance measures drive the search for management and policy options and provide the basis on which they will be compared. The use of objective hierarchies appears to be characteristic of SDM, possibly due to the strong focus on collaboration and encouraging participants to explore, and hopefully better understand each other’s values. Developing decision alternatives is a creative, deliberative process that aims to tailor candidate actions (or actions sets) in a way that serves the defined objectives. It is quite common that certain actions most suit the objectives of a particular stakeholder. Evaluating the performance of a particular stakeholders preferred actions against the criteria of other stakeholders is a key part of understanding the consequences of each alternative. A basic tool used widely in SDM is the ‘consequence table’ which sets out the expected outcome of each action for each performance measure relating to an objective. The process of estimating consequences of actions for objectives is a key place in which models and scenarios can play a role in SDM. Models and scenarios can help in the exploration of expected outcomes arising from courses of action and the uncertainty about those expected outcomes. Evaluating trade-offs and selecting favoured options then proceeds by considering which options provide reasonable outcomes across all of the objectives considered. Proponents of SDM are generally eager to point out that the evaluation of trade-offs involves “valuebased judgements about which reasonable people may disagree” (Gregory et al. 2012). Finally, implementation and monitoring of outcomes enables some post-hoc evaluation of outcomes for the Page 17 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
purposes of reporting and learning (McDonald-Madden et al. 2010b); providing opportunity for the SDM process to be adaptive (sensu Walters 1986).
5
10
15
Strengths and weaknesses of SDM – Two key strengths of SDM emerge from many of the reported applications. These include the clear separation of facts from values that is at the heart of the approach (Maguire 2004), and the way in which SDM helps with partition and therefore simplify the technical and social complexity that commonly hinders most real-world decision problems. One of the developers of SDM theory and practice describes it as “... the formal use of common sense for decision problems that are too complex for informal use of common sense” (Keeney 1982). This quote highlights the point that there is nothing mysterious or even particularly new about any aspect of SDM, other than the way in which it brings together may key concepts from decision theory toward a workable protocol for deliberations. A weakness of SDM is that guidance on how to undertake any given step within the SDM ‘cycle’ tends to be minimal and vague. The key text on SDM for environmental applications (Gregory et al. 2012) emphasizes that the use of SDM is something of an ‘art’. Knowing which specific tools to employ in any given decision context at each stage of the SDM process clearly requires significant experience. This means that SDM cannot simply be used ‘off-the-shelf’ by relatively inexperienced decision analysts. This may be just as well given that poor decisions can have large consequences.
20 Box 2.2.8 SDM case study: Non-native fish management in the Glen Canyon dam, USA Runge et al. (2011b) describe a structured decision-making project run by the U.S. Geological Survey concerning control of non-native fish below Glen Canyon Dam in the states of Utah and Arizona in the USA. They created a forum to allow agencies and Tribes to articulate their values, develop and evaluate a broad set of potential non-native fish control alternatives, and to define individual preferences on how to manage the trade-offs inherent in this managing the problem. Two face-to-face workshops were used to discuss objectives and represent the range of concerns of the relevant agencies and Tribes. A set of non-native fish control alternatives was developed. Between the two workshops, four assessment teams worked to evaluate the control alternatives against an array of objectives. At the second workshop, the results of the assessment teams were presented. Multi-criteria decision analysis were used to examine the trade-offs inherent in the problem, and allowed the participating agencies and Tribes to express their individual judgments about how those trade-offs should best be managed in selecting a preferred alternative. An effort was made to understand the consequences of the control options for each group’s objectives. In general, the objectives reflected desired future conditions over 30 years. Multi-criteria decision analysis methods allowed the evaluation of alternatives against objectives, with the values of individual agencies and tribes deliberately preserved. Trout removal strategies in particular parts of the catchment, with a variety of permutations in deference to cultural values, were identified as top-ranking portfolios for all agencies and Tribes, based on cultural measures and the probability of keeping the endangered humpback chub population above a desired threshold. Sport fishery and wilderness recreation objectives were better supported by the top-ranking portfolio than other options. The preference for the preferred removal portfolios was robust to variation in the objective weights and to uncertainty about the population underlying dynamics, over the ranges of uncertainty examined. A ‘value of information’‖ analysis (sensu Runge et al. 2011b) led to an adaptive strategy that includes three Page 18 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
possible long-term management actions, and seeks to reduce uncertainty about the degree to which trout limit chub populations, and the effectiveness of particular removal strategies in reducing trout emigration to where the largest population of humpback chub exist. In the face of uncertainty about the effectiveness of the preferred removal strategy, a case might be made for including flow manipulations in an adaptive strategy.
2.2.9 Management strategy evaluation (MSE) 5
10
15
20
Management strategy evaluation (MSE, sometimes also termed as “Management Procedure Approach”, “Harvest Strategy Evaluation” or “Operating Management Procedures”) uses simulation models within an adaptive framework (sensu Walters 1986). The objective of the approach is to assess the consequences of alternative management strategies in a virtual world, taking multiple and often conflicting objectives into account (Butterworth 2007, Bunnefeld et al. 2011). Thus, MSE can be used to reveal the trade-offs in performance across a range of management objectives (Holland 2010). MSE does not prescribes an optimal strategy, instead, it provides the decision maker with the information on which (given their own objectives, preferences, and attitudes to risk) a rational decision should be based. The conceptual framework and the subsystems modelled by MSE are shown in Fig. 2.2.9; the modelling steps are discussed based on Rademeyer et al. 2007. First, an ‘operating model’ (or preferably, a set of candidate models) is created to address all of the key biological processes, tradeoffs and uncertainties to which an ideal management procedure would be robust (usually one model is chosen as a reference model). These operating models (most typically population dynamics models) are used to compute how the resource responds to alternative scenarios (different future levels of catch or effort). Then performance of each model is integrated over all the considered scenarios. Likelihood of the occurrence of each scenario is regarded as a relative weight given to the output statistics. The final management strategy (procedure) ideally is chosen based on clear, a priori defined objectives.
Page 19 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 2.2.9 The MSE framework (Adapted from Adam et al. 2013, p. 5.)
5
10
15
20
25
The method is typically used in the marine context to identify fishery rebuilding strategies and ongoing harvest strategies for setting and adjusting the total allowable catch, but there is a possibility for application for conservation purposes as well (Winship et al. 2013). Terrestrial application has also been promoted lately (e.g. Bunnefeld et al. 2011, Edwards et al. 2014). Strength and weaknesses of MSE – A core strength of the approach is its robustness to natural variation, and to uncertainty and error, both in stock assessments and implementation of management controls (Punt and Donovan 2007, Holland 2010): instead of using a single model to find an optimal solution, several candidate models are considered in a virtual world to evaluate alternate hypotheses (Rademeyer et al. 2007). MSE promotes consultation (Bunnefeld et al. 2011) whereby managers and other stakeholders can provide input into the candidate models and scenarios (e.g. Nuno et al. 2014). As the framework requires clear objectives to do the evaluations against, participants have to be clear about their objectives. This strength is increasingly exploited as the number of examples is growing where individual stakeholders are provided certain flexibility (Plagányi et al. 2007) or indigenous interests are regarded (Plagányi et al. 2013) in the management of socio-economic systems. The benefit of transparency for the industry is that is allows adjustment of the capacities well in advance in the light of future expectations, thus job security and more stable economic environment is provided. Although MSE is suggested to be the best practice in fisheries management, relatively few fisheries worldwide have been or are managed on this basis (de Moor et al. 2011). The reason behind include complexity and computation-intensive analysis as well as the lengthy development time, and the rigidity of framework (flexibility is needed to take the socio-economic reality of the moment into account). In addition, regular, planned reviews are needed, especially when new monitoring data are Page 20 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
available (Butterworth 2007). Similar to other paradigms, few guidelines are available how to construct, evaluate, implement, and present the results of an MSE framework (Rademeyer et al. 2007). Box 2.2.9 MSE case study – Joint management of fisheries in South Africa Plagányi et al. (2007) reports about the management of South African sardine and anchovy fisheries. The two species have to be managed jointly as anchovy harvest is necessarily accompanied by the bycatch of juvenile sardine; however the latter is more valuable when adult; thus resulting in a trade-off. In the first joint management plan in 1994, TACs (total allowable catches) were calculated based on abundance estimates from recruitment hydroacoustic surveys (in May) and spawning biomass (in November). The total allowable bycatch (TAB) of sardine was based on the anchovy TAC, but the latter was not affected by the TAC or TAB of sardine. However, the constraint posed by the sardine TAB proved to be too strict, thus the management plan was updated in 1999 to allow a more flexible sardine TAB to be set, depending on the relative recruitments estimations of the two species at any point in time. A trade-off curve was used in the selection of management goals to show explicitly the inverse relationship between the projected anchovy catch, with its associated juvenile sardine bycatch, and the directed (adult) sardine catch. Since the 2002 update of the management strategy TAC calculations have been based on an additional 3 years of data and also, the anchovy TAC has been to be adjusted during the season to allow further catches late in the year when the relative sardine bycatch is very low. The simulation testing included implementation uncertainty as the sardine TAB has not always been filled in practice. Furthermore, individual rights-holders in the fishery were to select their own anchovy– sardine trade-off, instead of adapting to a fixed, universal value. However, unexpected record levels of both species recruitment called for the adjustment of the new management strategy (in 2004), to exploit the (short-term) record biomass levels. Also, the new strategy allows for the increase of the sardine TAB (and anchovy TAC) set at the start based on the May recruitment survey, if sardine abundance is high. Basically, recruitments estimations are based on well-defined age-structured population models (for equations with the most recent calculations see de Moor, 2014). The key model inputs include November survey estimates of adult fish, May data about juvenile recruitment; catchment data are also regarded. The directed sardine TAC, an initial anchovy TAC and an initial sardine TAB for a given year are calculated based on the November results (for anchovy a historical average of 19841999 is taken as a baseline to which the current data are compared). Results are tested by simulation to ensure robustness in terms of expected catches and uncertainties about the resource dynamics. Thus initial harvest control rules are made, of which the lower limit is -15% (directed sardine TAC) and -25% (anchovy TAC) of the previous year (to ensure economic stability). Upper limits are based on processing capacities. These initial rules are adjusted during the year taken the May data into account (De Oliveira and Butterworth 2004.).
5 2.2.10 Strategic Environmental Assessment (SEA)
10
Strategic Environmental Assessment (SEA) is the systematic environmental assessment of policies, plans and programmes (Thérivel and Paridario 2013); an evidence based instrument that adds scientific rigour to policy making via suitable assessment methods and techniques (Fisher 2007). The primary aims and objectives of conducting SEA (UNEP 2000) include; (i) supporting informed and Page 21 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
integrated decision making by identifying the environmental effects of proposed actions, (ii) considering alternatives and specifying appropriate mitigation measures; and (iii) contributing to environmentally sustainable development, by providing early warnings of cumulative effects and risks (Du et al. 2012).
5
10
15
20
25
SEA is now becoming a more frequently used tool with regulations and guidelines for SEA being proposed in many countries worldwide. For example, in the EU the SEA Directive requires an environmental assessment for plans and programmes at national, regional and local levels of jurisdiction. Increasingly, developing countries are introducing legislation or regulations to undertake SEA – sometimes in EIA (Environmental Impact Assessment) laws (e.g. China, Belize, Ethiopia) and sometimes in natural resource or sectors laws and regulations (e.g. South Africa, Dominican Republic). In Australia ‘strategic assessments’ aim to analyse the cumulative impacts of multiple stressors on species listed as threatened under the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act). The Convention on Biological Diversity (Article 6b and Article 14) encourages the use of SEA in its implementation (without making it a specific requirement). The Paris Declaration calls for the development of common approaches to environmental assessment generally, and to SEA specifically (www.oecd.org/dac). EIA legislation often requires an SEA-type approach. However, SEA and EIA are used at different levels of the decision-making hierarchy. The former addresses policies, plans and programmes; the latter focuses on the lowest hierarchical level (projects). Thus, SEA applies to different geographic and time scales and different levels of detail at strategic and project tiers and it is often conducted before a corresponding EIA. Table 2.2.10 summarizes the changing focus of SEA, depending on how far away from the project level it is applied. Whereas at lower tiers, SEA is likely to be based on a more rigorous EIA-based approach (such as field surveys, overlay mapping and multi-criteria analysis), at higher tiers it is likely to be more flexible (and possibly non-EIA based, like in case of forecasting, backcasting and visioning, see Zhang et al. 2004, Liu et al. 2006, Wang et al. 2007, Du et al. 2012). Generally speaking, quantification is more difficult at higher tiers that come with a greater degree of uncertainty (for positive examples see Fischer 2002).
30 Table 2.2.10 The changing focus of SEA from lower tiers to higher tiers.
Page 22 of 60
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
Strength and weaknesses SEA – The primary strengths of SEA include a relatively good integration of environment and development objectives, provision for the role of science-based evidence to support informed decisions, capacity to identify and generate new options, potential to build public engagement and improved transparency, increased chance of early problem identification, relative ease of trans-boundary co-operation, and clarity around institutional responsibilities (e.g. division of responsibility between local government, line departments, state/provincial and national/central governments). On the downside, SEA generally covers a large area and a large number of alternatives, which may make collection and analysis of data complex, operating as the broad scale over a greater range of impacts and values may make SEA less ‘certain’ than any individual EIA, the production or disclosure of (often confidential) data needed for public participation can face opposition from a variety of private and public stakeholders; though the imperative to publicise likely impacts that might otherwise remain hidden can also be seen as a strength. There are no hard rules about the nature of public consultation under SEA, which opens the method up to minimal or ‘token’ consultation. Lack of expertise and specialist skills among the general public can lead to power differentials in the process where some stakeholders are well resources, informed and organized. In most administrations under severe human resource and financial constraints, SEA may be seen as a large administrative burden and impossible to properly manage, audit and enforce.
20 Box 2.2.10 SEA case study - Integrated assessment of biodiversity impacts arising from the PerthPeel housing corridor plan, Australia. In 2011, the Australian Government and Western Australian Government formally agreed to undertake a comprehensive strategic assessment of the Perth and Peel Regions of Western Australia in accordance with strategic approvals provisions under the EPBC Act. The strategic assessment was designed to enable a 'big-picture' approach to environment and heritage protection that provides certainty in the long term by determining; (i) areas to be protected from development, (ii) areas suitable for development, (iii) the type of development that will be allowed, and (iv) the conditions under which development may proceed. The primary types of development being considered under the SA were residential, industrial and infrastructure development (roads and rail), as well as expansion of mining and forestry activities over a 30 year planning horizon in a region covering 8,200 km2 (www.environment.gov.au/node/18607). The assessment process commenced with the generation of maps outlining development options. These options were assessed for their likely impacts on known occurrences and predicted suitable habitat of species listed as threatened under the EPBC Act and relevant state biodiversity legislation. State Government planning and environmental agencies were responsible for developing the strategic plan taking into consideration projected human population growth in the region. A three step biodiversity assessment approach was employed (Fig. 2.2.10) to help agencies assess the potential impacts of development plans on biodiversity across the region. An iterative process was employed to analyse impacts and refine options, ultimately leading to identification of a lowest impact biodiversity scenario that satisfied development objectives. High value biodiversity areas were characterised using the spatial prioritization software ‘Zonation’ (Moilanen et al. 2005), based on correlative species distribution models for 61 threatened species and point data for a further 135 species for which there were insufficient records to build distribution models. The strengths of this case study included the development of multiple options which were Page 23 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
assessed in an iterative manner, the integration of state-of-the-art biodiversity modelling with a broad scale decision process, and a relatively generous amount of time dedicated to assessing biodiversity impacts of development options prior to decision making.
Figure 2.2.10 SEA analysis framework for assessing the biodiversity impacts of planning options in the Perth-Peel planning region of south-western Australia.
2.2.11 Integrated Territorial Planning (ITP) 5
10
15
Policies and strategic plans containing specific objectives may or may not have spatial components that are implemented through zoning, permits and regulations. However, land use plans are always linked to specific objectives that may be included within policies or strategic plans and therefore follow a similar process of formulation, implementation and evaluation. However, land use planning also occurs informally at the level of individual landowners, farmers and communities (Amler et al. 1999). It is entirely compatible with the Ecosystem Approach, the primary framework for action under the Convention on Biological Diversity (CBD), which is a strategy for the integrated management of land, water and living resources that promotes conservation and sustainable use in an equitable way (Shepherd 2004). The integration of territorial plans needs to occur horizontally between neighboring and sometimes overlapping jurisdictions and vertically from the individual land use plot to the national and supranational levels, in order to promote common interests or reconcile objectives that could be negatively impacted. Integrated territorial planning must respond to jurisdictions that are recognized by specific national legislation and that are hierarchically organized, such as national, subnational, Page 24 of 60
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
protected area, private and collective communal land (Amler et al. 1999). Additionally multi stakeholder platforms can be established to facilitate spatial planning across areas that do not respond specifically to jurisdictions, such as watersheds, individual ecosystems or areas of influence of development projects. In this context the strong links across the scales need to be considered in the analysis of land or marine land management (Lambin et al. 2005, Ballinger et al. 2010). Integrated Coastal Zone Management (ICZM) and Integrated Watershed Management (IWM) are applications of territorial planning in specific contexts that are implemented through cross-jurisdictional agreements between representative state, grass roots or private stakeholders (Alves et al. 2011, Ballinger et al. 2010).
10
15
The recent development of Geographic Information Systems (GIS) provides a technical framework that characterizes and quantifies territorial attributes and allows analysis of spatial attributes with spatial statistics (Greenberg et al. 2002). Multivariate spatial analysis of residuals used in GIS are particularly well suited to support integrated land use planning, because of the diversity, as well as uneven and spatially specific distribution of both human influences and interests, as well as ecosystem components (Woolmer et al. 2008). The interpretation of landscape pattern indices needs to be based on stochastic models that handle landscape heterogeneity and where spatial parameters are estimated from observed data (Fortin et al. 2003). However, the causality between landscape metrics and ecological processes is frequently uncertain to a certain degree.
20
25
Maintaining landscape connectivity is a key issue for biodiversity conservation, species population survival and ecosystem stability and integrity (Taylor et al. 1993, Crist et al. 2005, Ricotta et al. 2000, Noss and Daly 2006). The need for incorporating connectivity criteria in integral territorial planning is backed by theoretical frameworks such as island biogeography, metapopulation models and population genetics (Crooks and Sanjayan 2006); and requires explicit spatial information about conservation objectives and minimum spatial requirements. Tools like Marxan (Ball et al. 2009) enable mapping ecological needs and distribution of focal species, through the spatialization of demographic and species distribution data that can then be overlapped with human territorial needs using a variety of methodologies (Sanderson et al. 2002).
30
35
40
Strength and weaknesses of ITP – The strengths of ITP lies in its visual products, or thematic maps, that can be used, across cultures and technical capacities, to bridge knowledge systems by presenting both technical information and local knowledge and values. This strength represents an opportunity for addressing the monitoring challenge mentioned above through participatory methodologies. The critical technical challenges remaining include downscaling hydro climatic models to a level relevant to land managers (Sachindra et al. 2014), and scaling up the effects of local land use changes (Feaster et al. 2014, Nobre 2014). Additionally in order to increase political support and resources for territorial management models are required to inform the livelihood costs and benefits of environmental services under different development scenarios. Finally, in order to evaluate the effectiveness of the territorial plans cost effective and informative monitoring systems must be established and integrated across scales. This is particularly challenging at the local scale where financial and technical resources are scarce.
45 Page 25 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Box 2.2.11 ITP case study: Tacana land management is key to curb deforestation The Tacana indigenous land is found in Northwestern Bolivia along the road from San Buenaventura to Alto Madidi (north of Ixiamas) and the eastern border of Madidi National Park and Natural Area of Integrated Management. GIS and spatial statistical analysis were used to analyze the correlation between geographical conditions and loss of forest cover during a historical period (2005-2010) as then projected to 2021. Determinant factors were also included in the analyses to permit the comparison of deforestation in areas with and without land management: land ownership, land management and improvement of road infrastructure. Three possible deforestation scenarios were modeled: 1) Base Scenario to 2021 in the absence of land management efforts, using historical deforestation rates (2005-2010) in the area outside of the Tacana Indigenous Territory; 2) Improved Road Scenario to 2021, using historical deforestation rates (2005-2010) along the Yucumo-Rurrenabaque road; and 3) Land Management Scenario to 2021 using historical deforestation rates (2005-2010) within the Tacana Indigenous Territory. The lowest percentage of deforestation, of only 0.5% per year, occurs within the Tacana Indigenous Territory, in a scenario with land management even in areas within the indigenous territory that are found along the road from San Buenaventura to Ixiamas. The highest rate of forest loss corresponded to the area along the road between Yucumo and Rurrenabaque, with 3.7% per year, while the strip between San Buenaventura and Alto Madidi, in areas of private property and of migrant farmers, presented a rate of 2.3%. Using these historical deforestation rates, the study projected deforestation between 2010 and 2021. The scenario modeling concludes that Tacana land management will avoid forest loss over 230,842 hectares between 2010 and 2021. In addition, a spatial analysis of the areas that would be deforested in the absence of Tacana land management shows that it would avoid impacts on Madidi National Park and Natural Area of Integrated Management, along the foothills of the Andean piedmont. This protected area is essential for conservation of biodiversity and the headwaters of over a hundred streams that provide water to the entire Iturralde province. Tacana land management also protects deforestation of connectivity corridors between Madidi protected area and the Tacana Indigenous Territory. These corridors are critical for maintaining wildlife populations that are important for endangered wide-ranging species and for management of subsistence hunting by the indigenous communities. Deforestation is also avoided in important areas for erosion control and along the course of main rivers and streams.
2.2.12 Decision Delphi technique
5
10
The Delphi technique was developed by the RAND Corporation in the late 1960's as a forecasting methodology (Gordon & Helmer 1964, Linstone & Turoff 1975). Soon after, it was adapted as a decision tool (Rauch 1979). Rauch (1979) defines three relevant types of Delphi; Classical Delphi, Policy Delphi and Decision Delphi. The focus of classical Delphi is on forecasting and elicitation; describing the future, while the latter two are focussed on mediating outcomes that influence the future. Classical Delphi may play a role in agenda setting (sensu Chapter 1 of this report), while Policy and Decision Delphi are particularly appropriate when decision-making is required in a political or emotional environment, or when decisions affect strong factions with opposing preferences. Decision Delphi can be used formally or informally to exploit the benefits of group decision making while attempting to insulate against its limitations (e.g deference to authority and groupthink). Example applications of Delphi as a decision tool include allocation of national level health funding in the USA (Hall et al. 1992) and setting priorities for the IT industry in Taiwan (Madu et al. 1991). Page 26 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Delphi can work as an informal, subjective decision support model when the decisions are based on opinion, and can be converted to a formal model, when the data is more knowledge-based. While the specific approach to the application of Decision Delphi varies according to context, a few characteristic steps can be identified:
5
10
15
20
25
30
1. 2. 3. 4. 5. 6.
Pick a facilitation leader. Select a panel of experts. Identify a criteria list from the panel. The panel ranks the criteria. Calculate the mean and deviation of criteria – rank and explore reasons for deviations. Re-rank the criteria - Repeat the ranking process among the panellists until the results stabilize. The following steps are particular to the use of Delphi in a decision context, such as in ranking environmental project options. 7. Identify project constraints (e.g. budgets, legislation) and preferences 8. Rank projects by constraint and preference - Each panellist ranks the projects first by the hard constraints, for example, up to the maximum available budget. 9. Spread preference points (limited to some arbitrary number) among the project list. Analyze the results and feedback to panel. 10. Find the median ranking for each project and distribute the projects into ‘preference quartiles’. Produce a table of ranked projects, with preference points. Projects between the 25th and 75th quartile may be considered to have consensus (depending on the degree of agreement desired); projects in the outer-quartiles should be discussed. Discuss reasons for the large observed differences. 11. Re-rank the projects until rankings stabilize. Box 2.2.12 Decision Delphi case study: Choosing focal species for conservation in an urbanizing environment Hess and King (2002) describe a process for selecting focal species, using a suburbanizing region in North Carolina in the United States as an example. They identified focal species to be used for conservation planning in the region using a three-part Delphi survey, administered to a panel of experts. The panel identified six landscape types and nine associated focal species. They report that administering the Delphi survey was more labor-intensive, and took longer, than anticipated.
35
2.2.13 Lessons learnt – what works well and why... what doesn’t
40
45
On considering the applications of the (sample of) decision approaches described in the previous sections, a defining feature of the successful decision processes appears to be the level of commitment and involvement of the decision analyst or facilitator throughout the length of the decision process. In the previous sections we have documented successful examples of decision processes that ranged from highly participatory, deliberative, non-technical exercises (e.g. the Tacana land management exercise; 2.2.11), or more at the technical end of the spectrum (e.g. the Paijanne Lake MCDA exercise in Finland; 2.2.5), or a combination of the two (e.g. non-native fish control in Glen Canyon; 2.2.8) and all had very strong commitment and support from decision analysts, modellers, and/or facilitators. These people might be considered ‘champions’ of their given Page 27 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
decision strategy, and like champions of change, they are essential for successful use of models and scenarios in formal decision processes (Guisan et al. 2013).
5
10
15
20
25
30
Nonetheless, there is a mismatch between the preponderance of academic and theoretical studies around modelling, scenario development, and formal decision approaches, and the relatively small number of documented case studies that present successful application of models and scenarios in decision making in the environment sector, especially at the broader regional and global scales. It is hard to imagine that the relatively small number of documented successful examples of models and decision making is due solely to a lack of champions. By comparison to biodiversity and ecosystem services, examples of successful application of formal decision approaches such as MCDA and scenario planning (often using models and scenarios) in other sectors such as manufacturing, business, and the military abound. There appears to be some particular impediment to wider application of such approaches in biodiversity and ecosystem service policy, planning and management. It may relate to the complexity of such systems, a general lack of trust in data and measurement methods, or a willingness to invest the time and financial resources into making it work well. There is currently a large gap between models, scenarios and the decision applications that they are theoretically built to support. The following sections aim to provide some clues about how we might bridge that gap.
2.2.14 A taxonomy of decision support paradigms and approaches for using models and scenarios in decision making. Considering the approaches and case studies described above leads to some generalizations about the sorts of decision approaches that lend themselves to application in particular decision making contexts. While some aspects of this relationship between decision context and methods are selfevident; for example, the use of MCDA in decisions involving multiple stakeholders or decision makers, other patterns emerge which may be less obvious a priori. For example, it appears that for the most part, sequential decision problems tend primarily to address single-objective problems, while large-scale, multi-objective problems tend not to be handled as adaptive management problems. This may be simply that regional-scale, multi-stakeholder decisions problems tend to be once-off decisions with no plan or program for future changes, or because the inherent complexity of such decisions precludes analysing them as sequential decision problems, even if they are so in reality.
Page 28 of 60
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
Figure 2.2.14 Three key axes of decision context, with the dashed arrows indicating increasing complexity from a single (once-off) decision made by a single group with a single objective at a local scale, to a sequential decision made by a group of decision makers with multiple (usually competing) objectives at regional/global scales. Numbered circles indicate individual applications of a given decision support method, undertaken in different parts of the decision space. For example, circle 1 represents a study (Joseph et al. 2009) in which a single organization (NZ DoC) used a single objective criteria (maximize increase in species persistence/$) at the national scale regarding. Circle 5 identifies a conservation planning exercise, undertaken by the Malagasy governments, with the single objective of identifying the areas of Madagascar that would most efficiently increase the representativeness of the Madagascan reserve system (Kremen et al. 2008). There was no explicit consideration of sequentially increasing the reserve system or the multiple competing social or cultural objectives in the structured part of the reserve design process, though these considerations would likely have played out in the less structured political process. In contrast, study 2 reports on a decision process in which multiple cultural groups with multiple (incommensurable) objectives participated in a decision about the control of non-native fish species in the Glen Canyon dam in southern USA (Runge et al. 2011a). Study 2 was described as a ‘structured decision making’ exercise (sensu Gregory et al. 2012), supported by MCDA with swing-weighting to help identify dominated options.
A key observation across the hundreds of documents reviewed here is that most documented applications of formal decision support using the methods reviewed above occur at national, subnational and finer scales. We note that at regional and global scales, there is a rapidly growing number of applications of modelling and scenario analysis in policy agenda setting (e.g. Pereira et al. 2010, Alkemade et al. 2009, Leadley et al. 2010, Leadley et al. 2014, SCBD GBO4 2014), but this has yet to translate into a commensurate rise in the application of formal decision support approaches for biodiversity and ecosystem service policy implementation at those scales. The potential for application of formal, structured approaches to policy development and implementation at the broad scales seems great, but there are clearly strong political, cultural and practical impediments that must be overcome.
30
2.2.15 The role of scenarios and models in the decision-making approaches reviewed Page 29 of 60
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
It is apparent that scenarios and models can inform decision making by helping to interpret biological data with the aim of assessing status and trends in nature, and by projecting potential changes under different plausible future conditions. Scenarios can be used to identify alternative policy and management interventions that appear likely to result in acceptable long-run outcomes under a broad range of plausible futures (Peterson et al. 2003). Model-based sensitivity analysis can help to identify key uncertainties that impact most on beliefs about the apparent best course of action, thus setting a research agenda to resolve important uncertainties (Wintle et al. 2011). Modelling and sensitivity analysis can help identify surrogate measures (or indicators) that are highly observable and easily measured and which correlate strongly with fundamental performance measures.
10
15
20
25
30
35
40
Documented examples of the use of models and scenarios in real-world policy, planning and management decision-making tend to be at the less complex end of the decision context spectrum (Fig. 2.2.14). Some of the best documented examples of models being used in biodiversity management include the role of population dynamics models in sustainable harvest quota setting in fisheries and waterfowl management (e.g. Johnson et al. 1999), which are relatively simple singleobjective problems. In contrast, it is rare to see models and scenarios used in complex, multiobjective trade-off problems (but see Runge et al. 2011). This is something of a paradox, given that models and scenarios should be at their most useful when complexity and uncertainty are high. Different decision/policy contexts are more or less amenable to the use of scenarios and models. Governments of nation states engage in global or regional processes to define common principles and objectives concerning social, economic, political, institutional, and environmental components of sustainable development. In theory, global the design of policy initiatives should be particularly well suited to the use of scenarios and models because of the complexity and uncertainty surrounding outcomes that cannot be easily processed by the human brain, the access to specialized technical support and large data sets through multilateral agencies, and the need to balance and reach agreements across different interests, cultures and priorities. However, at a global level, beyond the policy agenda setting role for models and scenarios noted above, there is precious little documented evidence of models and scenarios being used within a structure process to develop and implement plausible, socially acceptable, sustainable and efficient policy at that scale. Regional multilateral platforms are being used more widely to develop policies around environmental issues such as watershed management, coastal fisheries, pollution, and also around critical drivers of change such as economic and infrastructure integration. However, the documented use of models and scenarios in these contexts is rare, compared with finer scale applications of models supporting local resource management decisions. This is partially because the number of jurisdictions decreases with increasing scale and hence the number of policy platforms is also smaller. Nevertheless, examples are available for CITES where simple scenarios of trade processes have been used to visualize and compare implementation of the processes involved in the control of the trading of CITES listed species in order to clarify issues related to the implementation of electronic permit systems (CoP13, Bangkok, 2004). Additionally, perhaps the most relevant to IPBEs are the models of some CBD indicators developed to inform development of post 2010 targets (Mace et al. 2013, Leadley et al. 2010).
Page 30 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
2.3 Aligning the design of scenarios and models with the needs of decision-makers
5
10
15
20
2.3.1 Modelling the right thing - Matching state variables to fundamental objectives Biodiversity can be generally classified into three major attributes at four organizational levels (Fig. 2.3.1). The organisational levels correspond to genetic or individual, species or population, ecosystem or community and landscape, while the attributes which determine and constitute biodiversity refer to composition, structure and function (Noss 1990). Compositional biodiversity has to do with the identity and variety of elements in a collection, structural biodiversity is the physical organization or pattern of a system, and functional biodiversity involves ecological and evolutionary processes (Franklin 1988). These major attributes of biodiversity and ecosystem services are further discussed in Chapter 3. Following the hierarchy concept, assessments and monitoring of biodiversity and ecosystem services should be at multiple levels of organizations because no single level of organization (e.g., gene, population, community) is fundamental (Noss 1990, Zacharias and Roff 2000). The appropriate state variables (or indicators) for models and scenarios depend on management objectives and the decision context in which models and scenarios are being used. Depending on the decision at hand, anything from coarse indices of biodiversity (Alkemade et al. 2009) through to a detailed information about the genetic variation of a single species in a local area may be needed to help support the decision. It is these choices, predicated on the needs of the decision maker and stakeholder, that determine the complexity of and uncertainty in models and scenarios that will be used to underpin their decisions. In general, that complexity increase as decision makers demand information about multiple attributes at multiple levels of organization (Fig. 2.3.1).
25
Figure 2.3.1 State-variables based on hierarchical complexity of attribute and organizational levels
Page 31 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
2.3.1.1 Individual level
5
10
15
20
25
30
35
40
45
At the individual level, models deal with genetic component, structure and process (or function). Genetic diversity is crucial for the long-term survival and evolution of species, and even more so for populations and species that are threatened or endangered (Espindola et al. 2014). The new strategic plan of the United Nations Convention on Biological Diversity that was adopted in 2011 explicitly addresses the conservation of genetic diversity, but implementation gaps in existing conservation policy to genetic biodiversity remains difficult to overcome (Wennerstörm et al. 2013). Information that can be acquired via genetic monitoring is often unavailable (Anderson and Karel 2009). Collections of national data, which are important for providing sound science to base policies and frameworks (e.g. CBD, CITES) have major gaps with respect to aquatic genetic variation below the species level, while small, local databases that do include knowledge on intra-specific variation are scattered, not easily accessed and limited in scope (FAO 2013). According to FAO, global policies and laws for fish and aquaculture do not address issues below the species level. More recently, new genetic-based concepts (e.g. population genetics, community genetics) recognise the importance of linking genetics not only along parallel attributes, but also across all organizational levels. Population genetic approaches have emerged as particularly effective means of monitoring populations for conservation and management objectives because acquisition and analysis of genetic data are often more feasible and reliable than data acquired via traditional genetic monitoring approaches (Schwartz et al. 2006). Population genetic data are being used more commonly in the identification of management units in natural populations (Anderson and Karel 2009). Community genetics, a relatively new science, are also proposed for conservation management. For example, plant genotype identity and genotypic diversity are found to be linked to community and ecosystemlevel processes, revealing that genetic variation can have extended effects beyond an individual’s phenotype (Hersch-Green et al. 2011). Moreover, the effects of genetic variation can have large impacts on direct and indirect species interactions, associated biodiversity and ecosystem function, and species adaptation to a changing environment (Bailey et al. 2012, Wennerstörm et al. 2013). The community genetics approach to management and conservation is likely to be productive in three major and interactive areas including climate change, exotic invasions, and habitat restoration (Bailey et al. 2012). Although too little is known regarding the impacts of genetic diversity on ecological structure, the study of Whitlock (2014) confirms that careful conservation management for the maintenance of genetic diversity (managing for ‘genetic health’) can lead to greater sustainability in populations and communities that are the focus of conservation efforts.
2.3.1.2. Species/Population level Species based approach has been applied to conserve a targeted ecosystem. The conservation of threatened species is more likely to obtain consensus from multi-stakeholders from global to local scale. But the conservation of processes and habitats of species can be supported by ecosystem and landscape approach (Franklin 1993). Prioritization of conservation effort and actions can be enhanced by extinction risk estimates of specific species compared across a range of competing policy, planning or management options (e.g. Akçakaya et al. 2004, Wintle et al. 2005, Wintle et al. 2011). Population viability analyses (PVA) are computer-based simulation models which project changes in initial population abundance over a time period and account for processes such as inbreeding depression, density dependence and demographic stochasticity. Individual-based models, age-structured and stage-structured demographic models are the main PVA approaches have been Page 32 of 60
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
developed to support conservation assessment and conservation decision making (Akçakaya and Gulve 2000). Estimating extinction risk is highly uncertain because no proven methods or reliable data exist for verifying extinctions and extinction risk in natural populations depends on many stochastic factors that affect individuals and populations (Melbourne and Hastings 2008). Arguments have been made in favour of other ways to summarise the results of population simulation models that are more robust than extinction risk estimation when using them to rank policy, planning and management options (McCarthy & Thompson 2001). The minimum viable population (MVP) concept aims to describe the population size that is required for a high probability of survival of a population over a given period and has been used in the IUCN’s Red List criteria of small and range-restricted populations (Trail et al. 2007). The utility of MVP for conservation planning has been assessed by long-term population data (Flather et al. 2011), and while the concept remains difficult to support empirically, it is a useful heuristic and communication device for conservation.
15 2.3.1.3 Community/Ecosystem level
20
25
30
35
40
The concept of ecosystems services has become an important model for linking the functioning of ecosystems (biodiversity) to human welfare benefits. Understanding this link is critical in decisionmaking contexts. Identifying critical ecosystem services and modelling its trade-offs and flows from supply to demand are thus important foci in decision processes (Burkhard et al. 2013, Fisher et al. 2009, Reeth, 2013). At the ecosystem level, indicators of change are now crucial for the EU Marine Strategy Framework Directive’s indicators of Good Environmental Status. To operationalize the ecosystem approach, the European Union has identified a range of desirable aspects of the ecosystem through the Marine Strategy Framework Directive (MSFD) 2008/56/EC (EU-COM, 2008). Good Environmental Status (GES) involves protecting the marine environment, preventing its deterioration, and restoring it where practical, while using marine resources sustainably (Jennings and Rice, 2011). Ecosystem Services are monitored through a list of indicators and reference levels that are suggested nationally (Shepard et al. 2014). Some indicators (e.g. Greenstreet et al. 2012, Rombouts et al. 2013) are strongly oriented towards communities such as the demersal (bottom dwelling) fish assemblage and rely on bottom-trawl survey data (Dickey-Collas et al. 2014), while others are based more on size based approaches of the whole food web (Blanchard et al. 2011) and still others on trophic food web models (Nicholson et al. 2012; Shin et al. 2010a and 2010b). These indicators are now well studied in large projects such as Indiseas (www.indiseas.org). Similar to the MSFD, the Convention on Biological Diversity (CBD) set new targets to halt biodiversity loss through the development of national biodiversity strategies, targets and action plans (Nicholson et al. 2012). Thus global biodiversity indicators are now proposed for communities such as African mammals assessed using the Red List Index, an index of extinction risk for species of plants and animals (Nicholson et al. 2012). Nicholson et al. (2012) uses biodiversity indicators to evaluate the impacts of policies in an indicator policy cycle and to test the ability of indicators to represent biodiversity trends.
2.3.1.4. Landscape level 45
Landscape planning (including planning protected areas) is an effective strategy to reduce global biodiversity loss (Possingham et al. 2006). However, there are large gaps in species representation in Page 33 of 60
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
current protected areas at a global scale (Rodrigues et al. 2004). Systematic conservation planning using decision-support tools such as Marxan (Ball et al. 2009) and Zonation (Moilanen et al. 2005) have been applied to assist the selection and design of new protected areas to complement existing reserves in terms of species representation (Margules and Pressey 2000). Species distribution models, which relate species occurrences to a set of environmental predictors, provide information on the spatial distribution of species and/or suitable bio-climatic environments for the species within a region of interest (Elith and Leathwick 2009, Kearney 2006). These models are commonly utilized in systematic spatial prioritization approaches which are now widely applied in the design of efficient land and marine protected (e.g. Leslie et al. 2003; Delavenne et al. 2011), terrestrial conservation planning (Kremen et al. 2008, Smith et al. 2008, Lessmann et al. 2014).
2.3.2 Modelling at the right scale – matching the spatial and temporal resolution of models and scenarios to the problem at hand 15
20
25
30
The ecological hierarchy concept refers not only to multiple levels of organization but also to multiple spatial and temporal scales. A parallel hierarchy concept on social system exist and it emphasizes that environmental management scales also cut across both space and time. This concept refers to cross-scale linkages, as defined by institutional interactions both horizontally or across space and vertically or across levels of organization (Berkes and Folke 2002, Berkes 2007). Thus, both ecological and social systems have hierarchical organization, with each subsystem nested in a larger subsystem. The benefits from the ecosystem services and objectives for ecosystem management differ for each institution and in their spatial jurisdictions. Because jurisdictional boundaries rarely coincide with ecosystem boundaries, cross-scale institutions that are in tune with the scales at which ecosystems function are needed to avoid a mismatch in scale between ecological and social systems (Berkes 2002). The mismatch between these systems is perhaps the archetypal scale problem (Cash et al. 2007). Mismatch in scale occurs because, according to Fremier et al. (2013), various lags in time and space exist between the production and consumption of different ecosystem services (Fig. 2.3.2). Collapsing fisheries, transboundary pollution problems, vulnerability to repeated extreme events, and the inability to address human-induced disease outbreaks are examples of devastating outcomes of not considering multiple scales in environmental policy, management, and assessment (MEA 2005, Cash et al. 2007).
2.3.2.1 Spatial and temporal scales
35
40
Temporal scale can be thought of as divided into different “time frames” related to rates, durations, or frequencies (Cash et al. 2006). Spatial scales can refer to both geographical and jurisdictional scales that are clearly bounded and organized political units (i.e. towns, counties, states or provinces, nations, regions) (Cash et al. 2006). Dynamics in spatial and temporal scales contribute to complexity and stochasticity in environment. Ecological dynamics is always stochastic at small scales, but radical changes in the scope and aims of ecology to cover ecosystem, landscape and even region and global over the past decades reflect in part the need to address pressing societal issues of global environmental change (Chave 2013). An important principle in ecosystem management is that analyzing and managing natural resources at different geographic scales is necessary to account for the functions, interactions, and emergent properties within not only ecological but also social systems (Cheng and Daniels 2003). Ecosystem models and scenarios that addresses jurisdictional spatial scales can match the constitutional and legal decision-making structure. However, many Page 34 of 60
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
environmental issues often cut across jurisdictional scales (e.g. cross-border watershed, global climate, marine resources). To understand an ecological system, it is important to study it at the appropriate scale, and develop models that bridge across scales (Chave 2013). According to Chave (2013), one of the great successes in ecosystem science in the last two decades has been to wed empirical data and models, and to move towards predictive models of the biosphere. Global scale quantitative models and long-term scenarios are now widely used to understand anthropogenic impacts on ecosystem services and biodiversity. However, information on time and distance lags in ecosystem-service provisioning between the service producer and consumer, which is crucial for effective management and proper valuation of ecosystem services (Fig. 2.3.2), is not yet captured in regional or global models. The issues of spatial and temporal scales in models and scenarios of biodiversity and ecosystem services are further discussed in chapters 3 and 6.
Figure 2.3.2 Effective management of ecosystem services requires an understanding of the lags between production and consumption, particularly across well-connected landscape features, such as within river– riparian systems. The gray line illustrates the increasing importance of management or payment for ecosystem services schemes and of matching the scale of the services with that of the organization. (Source: Fremier et al. 2013)
20 2.3.2.2. Social cross-scale linkages
25
30
Cross-scale linkages are interactions among actors or group of actors (communities, NGOs, governments, donor agencies, private companies, and scientists) between geographical and jurisdictional scales, who may have different priorities, objectives, and interests. Cross-scale linkages are temporally dynamic because the actors and the constitutions and motivations binding them change over time. Not only ecosystems and ecosystem services can change but also the value humans attribute to them (Bastian et al. 2013). The strength and direction of linkages may change over time as a consequence of those interactions or influence of other variables (Cash et al. 2007). An understanding of cross-scale linkages, or direct interactions through networks to provide information or tangible resources related to the management system, is important in managing multiple use resources (Adger et al. 2006). Cross-scale institutional linkages are the norm and even universal in natural resource management (Berkes 2002, Adger et al. 2006). Many ecosystem services Page 35 of 60
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
and functions debated and negotiated at the global policy levels like carbon sequestration, genetic biological resources, and water resources are considered as public or common goods. So far little is known about the cross-scale nature of resource management systems and the dominant mechanisms of cross-scale interactions (Adger et al. 2006, Cash et al. 2007). Research has just begun to explore the importance of cross-scale linkages for local level common pool resource management (Antinori and Garcia-Lopez 2008). There is so far no approaches to model social cross-scale linkages at higher levels of organization.
2.3.3 Choosing the right models and scenarios for the user – balancing accuracy, precision, complexity, interpretation, communication To capture the objectives (needs) of the decision makers and stakeholders at multiple scales, determining appropriate state-variables ‘Factors (system attribute)’ or ‘criteria’ that include both relevant properties of the socio-ecological system of interest (Possingham 2001) is critical in modelling biodiversity and ecosystem services. Fundamental objective is high-level outcomes that a planning process aims to accomplish through implemented actions (Shields et al. 2002). Thus, there is a need to distinguish high-level goals from lower-level criteria and indicators. High-level goals are an integrated part of the definition of any given planning problem and should be shaped primarily by societal or cultural values (Rapport et al. 1998, Norton and Steinemann 2002). Just as the landscape’s composition, structure and process (Fig. 2.3.1) affect the type and rate of ecosystem-service provisioning, social heterogeneity (i.e. position, location, property size, economic status, access to information, etc. affects motivations for conservation) is an important driving force in determining the adoption of conservation incentives for ecosystem services (Fig. 2.3.2) (Fremier et al. 2013). Thus, models and scenarios that integrate knowledge on both ecological and social landscapes characterizing the service provisions is crucial for policy, planning and management. This implies that the spatial and temporal scales of biodiversity and ecosystem services models and scenarios have to be improved to better match not only the jurisdictional duties and decision time frames but also the cross-scale linkages among decision makers and their stakeholders. Bottom-up research and management approaches that enable cross-scale linkages include Ecosystem-based Management (EBM), adaptive management, Participatory Rural Appraisal (PRA), and Participatory Action Research (PAR) (Berkes 2002). Common to these bottom-up approaches is deliberation among decision makers and stakeholders. Deliberation is important in dealing with problems of multiple and competing objectives and with competing understandings of human– ecosystem interactions (Berkes 2007). Moreover, these approaches are useful in incorporating global objective (e.g., climate change adaptation) into ecosystem management at lower scale. Objective hierarchies approach in SDM can be used to identify fundamental objectives and means to achieve the goal. There are examples of application of hierarchies approach in ecosystem management (Reichert et al. 2007, Reichert et al. 2013, Langhans et al. 2014, Bino et al. 2013). Mapping of ecosystem services is a useful tool for illustrating and quantifying the spatial mismatch between ecosystem services delivery and demand that can then be used for communication and to support decision-making (Crossman et al. 2013). Specific ecosystem unit that is modelled in terms of ecological phenomena often has mismatch with management scales, sectors, practices as some cases suggested e.g., watershed management, Page 36 of 60
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
habitat conservation planning. Hierarchy theory has been proposed as a way to understand scale effects on ecological patterns and processes (e.g., Simon 1962, Allen and Starr 1982, Urban et al. 1987). This theory provides a conceptual framework in which the role of scale is well-defined, and states that our understanding of a phenomenon depends on referencing the next higher and lower levels of resolution. A phenomena or an organism is thought to be bounded by processes generally operating at larger scales above it, and to impose bounds on processes and organisms at the level below it (Tang et al. 1997). Recognizing and resolving scale mismatches is an important aspect of building resilience in social ecological systems, and solution to scale mismatches usually require institutional changes. Especially long-term solutions to scale mismatch problems will depend on social learning and the development of flexible institutions that can adjust and reorganize in response to changes in ecosystems (Cumming et al. 2006). Downscaling long term global socioeconomic scenarios at city scale is useful at the urban area level, to help local decision makers develop these local adaptation and mitigation policies (Viguie 2014). Guidelines and models have been developed to help understand how information can be translated across scales (King 1991, Rastetter et al. 1992, Pacala and Deutschman 1995). However, models do have limitations and data required to improve high-level and cross-scale models are often not available. Recommendations on improving data for developing scenarios and modelling biodiversity and ecosystem services are provided in Chapter 8. Scaling up models that predict individual or stand-level responses to predict responses at the landscape level may be complicated by the presence of processes that act at higher levels of organizations not captured by individual or stand level models (Tang et al. 1997). Uncertainties that accompany model limitations should be identified and communicated.
2.4 Characterizing and dealing with uncertainty in scenarios and models in decision-making 25
2.4.1 Implications of uncertainty in scenarios and models on decision-making processes - fragile decisions – dangers of ignoring 30
35
40
Models can be used for simulation purposes (or develop scenarios) in order to obtain a better understanding of complex systems or for prediction/forecasting to assist managers with assessing the utility of proposed management actions or the response of the system to other types of perturbations (Maier and Ascough 2006). However, without adequate attention to the role and implications of uncertainty in models and scenarios, the outcome may be of limited value and could result in incorrect policy decisions, with all the attendant consequences (Petersen et al. 2013). From the management point of view, uncertainty is the lack of exact knowledge, regardless of what is the cause of this deficiency (Refsgaard et al. 2007). Each decision or set of decisions has associated gains or losses which are usually dependent on several random factors and thus highly uncertain (Fenton and Neil, 2012). Uncertainties are the greatest threats to project success and political will to approve and implement a project withers if there is too much uncertainty (Loftin 2014). The implications of uncertain outcomes can be catastrophic, as was seen for the collapse of the Newfoundland cod stocks in the early 1990s due to significant overfishing. The collapse occurred because the stock size was uncertain because of the type of data used to estimate the stocks (Walters and Maguire 1996) (Box 2.4.1). According to Waedekker et al. (2008), a responsible communication of uncertainty information leads to a deeper understanding and increased awareness of the phenomenon of Page 37 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
uncertainty and its policy implications. And this may result in a more responsible, accountable, transparent and effective use of intrinsically uncertain science in decision-making. Box 2.4.1 Case study on North Atlantic cod collapse The North Atlantic cod stock supported one of the largest and most economically important fisheries in the world for almost 500 years (Hutchings and Reynolds 2004). The implications of getting it wrong has been felt for the past 20 years with the cod stocks being designated “critically endangered” by IUCN (Hutchings 2000). The implication of not taking uncertainty into consideration when setting quotas for this stock had local, provincial, national and international repercussions, but the most important were on the small coastal communities of Newfoundland, affecting tens of thousands of Canadian fishers, and the overall Newfoundland economy (Budreau and McBean 2007). Uncertainties affect not only the predicted result of a model–data synthesis process, but also the predicted best estimate (Raupach et al. 2005) and the consequences of not making uncertainties explicit can be far reaching.
5
10
15
20
25
30
2.4.2 Sources of uncertainty in using scenarios and models for decisionmaking – known and unknown unknowns Sources of uncertainty are defined and categorized in different ways in various literature. Despite increasing recognition of the potential effects of uncertainty, there has been little agreement on a commonly shared terminology or a generic typology of uncertainties in environmental decisions (Mosadeghi et al. 2012). There are three dimensions most commonly used to distinguish typology and terminology of uncertainty (Mosadeghi et al. 2012, Petersen et al. 2013): (1) location of uncertainty, where the uncertainty manifests itself within the model structure and which include context, data, model, expert judgement and outputs; (2) level of uncertainty, where the uncertainty manifests itself along the spectrum between deterministic knowledge and total ignorance and which include statistical uncertainty, scenario uncertainty, and recognised ignorance; (3) nature of uncertainty, whether the uncertainty is due to the imperfection of our knowledge (i.e. epistemic uncertainty) or due to the inherent variability of the phenomena being described (i.e. ontic uncertainty). Petersen et al. (2013) found two more dimensions of uncertainty: the qualification of the knowledge base, and the amount of ‘value-ladenness’ and subjectiveness. Uncertainty is difficult to classify into only one of the categories (Regan et al. 2002). However, in the context of the decision-making process, it is useful to categorize the sources according to reducibility of or degree of management uncertainty. Uncertainties exist because the environment is complex and stochastic. These uncertainties are most difficult to avoid and manage. Models and scenarios can improve understanding of the environment by reducing or clarifying complexity. Models themselves cannot reduce uncertainty per se, but they can be used to characterise it and help understand how to reduce or deal with it. There are inherent uncertainties in models and scenarios because they are developed based on incomplete knowledge and are thus only simple abstraction and partial representation of an entire system. These uncertainties can be reduced through improvement of scientific knowledge and managed with application of appropriate decisions tools (Regan et al. 2005). Uncertainty about what people value and how they respond to risk also needs to be dealt with in decision-making. These uncertainties can be reduced with appropriate communication and deliberation strategies. These three sources of uncertainty conform to the broad categories of
Page 38 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
uncertainty in the environmental decision context as defined by Balint et al. (2011): stochastic, scientific, and administrative (decision or implementation).
2.4.2.1. Stochastic uncertainty 5
10
15
Environmental problems, which are affected by stochastic variables, are strongly affected by uncertainty because they are regarded as complex, characterized by a lot of dynamic processes and interactions and non-linear feedback processes leading to threshold effects and surprises (Sigel et al. 2007). Complexity from ecological perspective generally implies the existence of multiple levels of interconnected, dynamic relationships among large numbers of interactive agents (Balint et al. 2011). Complexity makes it difficult to forecast the effect of potential solutions and demands large amount of knowledge and information to find solutions for environmental problems (Sigel et al. 2007). Balint et al. (2011) define stochastic uncertainty as events that are largely random, unpredictable, and uncontrollable. For other authors, stochastic uncertainty goes beyond ecosystem. Stochastic uncertainty is the inherent uncertainty or randomness of nature, human behaviour and social, economic, and cultural dynamics (Mosadeghi et al. 2012). No matter how precise the data collection and for how long historical data time series exist, there will always be some uncertainty related to the chaotic nature of natural phenomena (Refsgaard et al. 2007). Thus, inherent indeterminacy and/or unpredictability, randomness and chaotic behaviour are variability-related uncertainty, which is typically not reducible through more research (Petersen et al. 2013).
20
25
Many natural and anthropogenic drivers of biodiversity and ecosystem services are sources of stochastic uncertainty including, for example, invasive species, pollution, climate change, etc. (see chapter 2). Climate change and extreme events are perhaps one of the most important dynamic processes that contributes to the complexity and stochasticity of the environment. Global climate models are thus applied to develop alternative scenarios for ecosystem changes. Chapters 2, 3 and 4 discuss examples of scenario analysis and model applications to assess impacts of climate change and its associated uncertainties on biodiversity and ecosystem services.
2.4.2.2. Scientific uncertainty 30
35
40
45
Models and scenarios are influenced by theories or concepts, data and methods, all of which are possible sources of scientific uncertainties depending on knowledge and information. The sources uncertainties, which Regan et al. (2002) categorized as linguistic and epistemic, cut across the various stages of developing models and scenarios. Linguistic uncertainty, which is associated with problems of natural language and scientific vocabulary, include issues of vagueness, context dependence, ambiguity, indeterminacy of theoretical terms, and underspecificity. Context dependence is closely related to context uncertainty, which concerns the framing and delineation of the problem, including the choices that determine what is considered inside or outside system boundaries: ‘delineation of the system and its environment’ (Petersen et al. 2013). Considering context uncertainty in decisionmaking could help to avoid problems arising from incorrect problem framing (Dunn 2001 as cited by Mosadeghi et al. 2012). Context includes uncertainty about the external economic, environmental, political, social, and technological situation that forms the context for the problem being examined (Walker et al. 2003; Refsgaard et al. 2007; Warmink et al. 2010). Epistemic uncertainty, associated with knowledge of the state of a system, includes measurement error, systematic error, natural variation, inherent randomness, model uncertainty, and subjective Page 39 of 60
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
judgment. Most of the epistemic uncertainties discussed by Regan et al. (2002) are related to data generation and model application. Predictive models are generally subject to input, model and parameter uncertainty (e.g. Loucks and Lynn, 1966; Burges and Lettenmaier, 1975; Vicens et al. 1975 as cited in Maier and Ascough 2006). Traditionally, the focus of environmental decision-making process has been on uncertainty in data and environmental models (Maier and Ascough 2006). Uncertainty in data includes both observations or measurements and prior estimates for model quantities (Raupach et al. 2005). Parameters (or prior estimates) used as model inputs are also an important source of data uncertainty. They are generally obtained directly from measured data or indirectly from measured input-output data by calibration (Maier and Ascough 2006). Regan et al. (2002) identified two sources of model uncertainty: first, when only variables and processes that are regarded as relevant and prominent for the purpose at hand are featured in the model; and second, in the way constructs are used to represent observed processes. Both influence the way the model is structured and are thus also referred to as model structure uncertainty (Wintle et al. 2003; Mosadeghi et al. 2012Conroy et al. 2011). Finally, there are things about which we currently know nothing that will impact on biodiversity and ecosystem services state variables in the future that cannot be parameterized in a model or characterised as bounds around a predictions of expected value (Wintle et al. 2010). ‘Unknown unknowns’ go beyond the conventional interpretation of Knightian uncertainty in which the probabilities (or probability distributions) of a set of known contingencies are unknown (Knight 1921), to encompass the case in which both the contingencies and (by definition their probabilities) are unknown (Taleb 2007). These uncertainties are considered by some to be the most important uncertainties in our lives (Taleb 2007). Various forms of scientific uncertainty in modelling biodiversity and ecosystem services are further discussed in chapter 3.
40
To model complex environmental problems, scientists are increasingly adopting multiple modelling approaches or integrated assessment models (Burnham and Anderson 2002). Maier and Ascough (2006) describe this as an increase in model complexity in order to better represent uncertainty about environmental and socio-environmental systems. Better identification and quantification of sources of uncertainty allow management or policy options to be evaluated with in terms of both expected outcomes and robustness to uncertainty (Wintle et al. 2011). A relatively recent development is the combination of species distribution models, land-use change predictions, and dynamic population models to predict the relative and combined impacts of climate change, land-use change, and altered disturbance regimes on species’ extinction risk (Colinsk et al. 2013). Because each model introduces its own source of uncertainty through different parameters and assumptions, Colinsk et al. (2013) emphasised the implications of compounded uncertainty for environmental management. Uncertainty also mounts when empirical data are not available to support integration of different models and modellers’ only option are to use expert judgement. Regan et al. (2002) refer to this as subjective judgment, an epistemic uncertainty due to insufficient empirical data to make reliable statements about parameter values and judgment of an expert based on observations and experience is used in place of empirical data. Chapter 6 provides detailed discussion on uncertainty in linking and harmonizing models and scenarios of complex environments.
45
There are several ways to describe, manage and reduce epistemic and linguistic uncertainties including, statistical techniques, probability distributions, fuzzy sets, multidimensional measures (Regan et al. 2002). Because epistemic uncertainty is due to imperfect knowledge, Mosadeghi et al. (2012) mentioned that this may be reduced by more precise data collection, additional monitoring
25
30
35
Page 40 of 60
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
and, using longer time series in modelling natural systems. According to Conroy et al. (2011), structural uncertainty is often represented by alternative models of system dynamics, each with associated measures of relative credibility. The reduction of structural uncertainty can also be achieved through the collection of data which is used as evidence in support of one model structure over another. This is a key objective of decisions tools like adaptive management (Walters 1986, Williams et al. 2002).
2.4.2.3. Decision uncertainty 10
15
20
If models and scenarios, which are used to assist at various stages of the environmental decisionmaking process, are to provide effective decision support, the uncertainties associated with all aspects of the decision-making process need to be taken into account explicitly (Maier and Ascough 2006). According to Maier and Ascough (2006), sources of uncertainties should cover not only data and models but also “human” inputs in decision-making process. These include knowledge, experience and expertise of modeller (as discussed above), as well as the more important influence of political “clout” and perceived importance of stakeholder(s), knowledge, values and attitudes of stakeholders, strength of argument presented by stakeholders, values and attitudes of decisionmakers, and current political “climate”. For example, on the one hand, values and attitudes of decision-makers can influence the environmental problem that is addressed, assessment criteria that are used, and alternative solutions that are considered and ultimately selected, and on the other hand, the values and attitudes of stakeholder groups can influence the choice, screening and assessment process of potential alternatives (Maier and Ascough 2006). These human factors affect the interpretation and valuation of uncertainties, whether inherent in the environmental system or generated from models and scenarios, and can further create uncertainties in decision-making process.
25
30
35
40
These factors cut across the different categories of human related uncertainties which were identified by Granberg et al. (2008), including cognitive, strategic, institutional and normative uncertainties. Cognitive uncertainty means that the causal relations are not firmly established and that actors disagree on the roots of a problem and its potential solutions (Giddens 1991, cited in Granberg et al. 2008). Strategic uncertainty means that it is not possible to foresee what strategic action the actors involved in the problem-solving efforts will take. Approaching the issue from different interests in a situation of cognitive uncertainty often result in diverging and conflicting strategies that, in turn, can lead to stagnation and deadlock in problem-solving efforts (van Bueren et al. 2003 as cited in Granberg et al. 2008). Institutional uncertainty arises from the fact that decisions are often made in different places and at different levels (Lidskog et al. 2005 as cited in Granberg et al. 2008). The institutional setting for decision-making is highly fragmented, resulting in difficulties to coordinate decision-making and actions. In addition, many environmental problems are cross-cutting and wide in scope, and cannot be handled by a solitary actor on a single societal level (Eckerberg and Joas 2004 as cited in Granberg et al. 2008). Normative uncertainty means that there is no set of shared values and norms that could guide the choice among societal goals such as material prosperity, human health, and biodiversity (cf. Dower 2007 as cited in Granberg et al. 2008). Even if there exists a shared understanding of the issue at stake, and an institutional setting that fosters collective decision-making, it may be hard to develop effective regulation and deliver concerted action because there is no guidance on how to prioritise between different objectives.
Page 41 of 60
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
Integrated models for assessing environmental management that allow for heterogeneity in behaviour of decision makers have recently been developed, but they do not consider uncertainty in human behaviour or in the various model components. Maier and Ascough (2006) suggested that developing frameworks that enable the uncertainties associated with human inputs to be accounted for explicitly is one of the upcoming challenges. This includes the development of uncertainty analysis methods that are able to cater for subjective and non-quantitative factors (e.g. van der Sluijs et al. 2005), human decision-making processes (which may be influenced by political and other external factors), and uncertainties associated with the model development process itself (e.g. Refsgaard et al. 2006). Chapter 6 discusses the relationship between decision stakes and system uncertainty.
2.4.3 Technical approaches to dealing with uncertainty in decision making 15
20
25
30
35
40
A multitude of technical, mostly mathematical, methods exist for dealing with uncertainty in choice problems. Stochastic Dynamic Programming can provide exact optimal solutions to choice problems when uncertainty about future states and the benefits arising from candidate actions can be described using probability distributions. SDP can be thought of as a generalisation of linear programming that copes with uncertainty about cause-effect relationships or future states. SDP has been used widely in determining optimal catch quotas in complex wildlife harvesting and fisheries adaptive management approaches (e.g. Johnson et al. 1997) and a limited number of land management and conservation applications (e.g. Richards et al. 1999, McDonald-Madden et al. 2010b). Partially observable Markov decision processes (POMDP) are a further generalization of SDP that allow for the case in which the state of the system cannot be perfectly observed, for example when the spatial distribution of a threatened species cannot be assumed to be known (e.g. Chades et al. 2008). A range of other methods under the general heading of robust optimization (sensu Ben-Tal and Nemirovski 2002) appear to show potential but are yet to see widespread application in biodiversity or ecosystem service decision making. Info-gap decision theory (Ben-Haim 2003) is an approach proposed to address decision uncertainty when plausible probability distributions cannot be defined. Info-gap has been used widely in the conservation academic literature (e.g. Regan et al. 2005, Moilanen et al. 2006, Wintle et al. 2011), though it is yet to see practical application in conservation as it has in other disciplines (Ben-Haim 2010). An advantage of the technical decision support approaches to dealing with uncertainty is that the role of modelling is clearly defined. Without a system model describing the dynamics of the system, there is no possibility of undertaking the uncertainty analysis. Use of uncertainty analysis methods demands that the analyst is explicit about the uncertainties impacting on a particular decision. Even if exact optimal solutions do not meet manager and stakeholder expectations because all of their concerns can seldom be incorporated, application of these formal uncertainty analyses does highlight which uncertainties are most important and which ones are inconsequential in the decision context. This can provide a strong motivation and guidance for investing in reduction of critical uncertainties. The primary impediment to the use of these models are the high technical expertise demand and limitations on the number of state variables (and states) that can be handled in practice.
2.4.4 Communicating uncertainty to enhance utility of scenarios and models in policy, planning and management Page 42 of 60
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
Various approaches to the communication and presentation of uncertainty have been developed, but not all are easy to understand by non-technical audiences, and they can also unexpectedly lead to misinterpretation (Waedekker et al. 2008). Van der Jeur et al. (2010) mentioned available uncertainty guidance documents, which range from being very generic and broad to more specific guidance documents. A recent and relevant document for communicating uncertainty is the “Guide for Uncertainty Communication” jointly developed by the PBL Netherlands Environmental Assessment and Utrecht University (Peterson et al. 2013). This guidance document suggests three sets of questions that researchers should answer when communicating about uncertainty in written reports and presentations. These questions relates to audience and relevance, distribution of information and presentation of information. In relation to the questions on audience and relevance, Petersen et al. (2013) explained that the basis of successful communication of uncertainty is differentiating among different audiences like policymakers, scientists, business sector, and general public. This is because different uncertainties are relevant to different people, in different situations, and in different stages of a policy cycle (Waedekker et al. 2008). Moreover, the basis for an efficient communication of final forecast products lies also on the assessment of how users perceive and understand uncertainty, and tend to act in the face of uncertain information (Ramos et al. 2010). According to Waedekker et al. (2008), communication of uncertainties aimed at policymakers and other parties involved in policymaking is important because uncertainties can influence the policy strategy. Uncertainty information was seen as important to put issues on the agenda, to prioritise them, and to phase the policy process. Moreover, policy relevance of uncertainty communication is influenced by factors like the place of an issue in the policy cycle, and its novelty, topicality and controversiality. Waedekker et al. (2008) concludes that, while policymakers strongly called for information on the implications of uncertainty, it does not mean that scientists should tell policymakers what to do, but that they should provide them with useful insights, to help them make their decisions. A systematic approach for distributing uncertainty information to audiences is the “Progressive Disclosure of Information (PDI)”, which classifies medium of communication into outer and inner layer (Petersen et al. 2013, Waedekker et al. 2008). The outer layers are non-technical information (e.g. press release, summary) which integrates uncertainties into the message and gives emphasis on context, implications and consequences, while the inner layers are detailed technical information (e.g. appendices, background report) which discusses uncertainties separately and gives emphasis on types, sources and the extent of uncertainty. Information disclosure is considered concrete operationalization of transparency in the global environmental domain and applied in various international, inter-governmental, private and NGO “governance-by-disclosure” initiatives (Gupta 2008). A very relevant example at the international level is the Aarhus Convention that aims to enhance access to information, public participation in decision-making and access to justice in environmental matters (UNECE 1998). In the private sector, public disclosure of corporate environmental information is an informal policy tools to overcome the ineffectiveness of regulative and market-based approaches to protect the environment (Anbumozhi et al. 2011). The PDI approach draws on these initiatives to organize distribution of uncertainty information according to the contents, style and degree of details and progressively disclosing layers of information from nontechnical to more specialised, according to the needs of the user (Waedekker et al. 2008).
Page 43 of 60
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
Referring to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC, 2001), Waedekker et al. (2008) discussed and compared the two forms of presenting uncertainty information – verbal and graphical. The IPCC uses words to reflect different levels of certainty (probability or confidence). While using words has the advantage that that people are better at hearing/reading, using and remembering risk information described in words rather than in numbers, it results in loss of precision, and words have different meanings for different people (Waedekker et al. 2008). However, Waedekker et al. (2008) argued that broad ranges and wordings may more accurately reflect the limited state of knowledge, which is not always obvious because experts tend to be overly precise and underestimate the uncertainty associated with their own predictions (cf. Slovic et al. 1981). By contrast graphics has the advantage of conveniently summarizing significant amounts of uncertainty (Ibrekk and Morgan, 1987; Wardekker and Van der Sluijs, 2005; Krupnick et al. 2006). But Waedekker et al. (2008) explained that most graphical expressions are not straightforward to understand and thus problematic, particularly when communicating with people who are not used to working with these expressions. They further added that ‘deep’ uncertainties (e.g. qualitative issues, such as problem framing, choice of methods, general level of knowledge and value-ladenness) cannot be easily quantified or expressed probabilistically and are hard to communicate using traditional methods, such as probability terms, uncertainty ranges, and error bars.
2.4.5 Dealing with uncertainty in scenarios and models through participatory, deliberative decision-making process According to Ramos et al. (2010), a lot still needs to be done to show the value of the process of communicating uncertainty, which should go far beyond the traditional approach of just displaying numbers, scores and good performance measures. In discussing communicating uncertainty in hydro-meteorological forecasts, they concluded that the practice of face-to-face forecast briefings, focusing on sharing how forecasters interpret, describe and perceive the model forecasted scenarios, together with continuous technical training, is essential in the communication of uncertain forecasts. Such a participatory approach is very important for environmental problems where complexity cannot be reduced through reduction in scientific (i.e. linguistic, epistemic) uncertainty. Citing van den Hove (2000), Proctor and Drechsler (2006) gives justification for participatory approaches to environmental problems because they are characterized by complexity, uncertainty, large temporal and spatial scales, and irreversibility. Moreover, wicked environmental problems demand learning throughout the decision-making process to deal with uncertainties. “Wicked problems are characterized by a high degree of scientific uncertainty and a profound lack of agreement on values” (Balint et al. 2011). According to Balint et al. (2011), in wicked problems, ecological complexity is compounded by social complexity, involving multiple, active, stakeholder groups with diverse values operating in an uncertain and shifting administrative, economic, political, and legal environments. Figure 2.4.4 situates environmental problems with wicked characteristics within the complexity and reducibility axes of uncertainty sources. While linguistic and epistemic uncertainties related to development of models and scenarios can be managed through communication, stochastic and decision-making uncertainties demand participation and/or deliberation among decision managers and stakeholders. Unlike communication which is a one-way dissemination of uncertainty information, deliberation through participatory approaches allows feedback and learning. Public Page 44 of 60
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
deliberation, which can be facilitated through workshops, town hall meetings, video conferences, interactive websites, etc., is important on three levels (Renn 2006, Balint et al. 2011): first, deliberative processes are necessary to define the role and relative importance of scientifically or systematically derived knowledge versus local experience and more idiosyncratic knowledge; second, deliberation is needed to find the most appropriate way to deal with uncertainty and to set efficient and fair trade-offs between potential overprotection and underprotection in the face of uncertain outcomes; and third, deliberation is needed to address the wider concerns of the affected groups and the public at large. Referring to wicked problems, Ascough et al. (2008) reasserted the importance of developing innovative methods for quantifying the uncertainty associated with human input by noting that human attitudes, beliefs, and behavior provide a large area beyond scientific and technical uncertainty in the ultimate solution to environmental problems. Wicked problems Stochastic Decision Reducibility Low uncertainty uncertainty of Epistemic Linguistic uncertainty High uncertainty uncertainty
Approach Deliberation High (e.g. SPA, CAMA) Complexity Communication/Deliberation Low (e.g. PDI, DMCE)
Figure 2.4.4 Approaches to dealing with uncertainty at different levels of complexity and reducibility.
15
20
25
30
35
40
There are only a few participatory approaches that allow deliberation to deal with the complexity of both nature- (stochastic) and human-related (decision-making) uncertainties in ecosystem management. The most frequently and widely applied approaches include, for example, including Scenario Planning Approach (SPA), Collaborative Adaptive Management Approach (CAMA), and Deliberative Multi-Criteria Evaluation (DMCE). SPA offers a framework for developing more resilient conservation policies when faced with uncontrollable, irreducible uncertainty (Peterson et al. 2003), i.e. stochastic uncertainty. It is a strategic foresight technique which is distinctive because it addresses uncertainty and not risk (Tapinos 2012). Scenario planning is a participatory approach that demands deliberation of scenarios between managers and stakeholders. Through scenarios, scientists and decision makers can collectively embrace uncertainty, prepare for a range of potential futures, and turn would-be crises into opportunities for positive change (Bohensky et al. 2006). The family of methods broadly known as ‘strategic foresight’ (Cooke et al. 2014a) show promise as a flexible set of deliberative tool to assist decision makers and stakeholders envisage and value a range of futures without being overly blinkered by the past. Horizon scanning is an example of a strategic foresight approach that has become an important agenda setting tool in conservation (Sutherland et al. 2009). To deal with scientific uncertainties, planners and decision makers often propose an adaptive management approach (Balint et al. 2011). Balint et al. (2011) explained that conventional forms of adaptive management incorporate scientific methods into the decision process, but the difficulties in reaching satisfactory decisions in many complex environmental dilemmas result not only from scientific uncertainty but also from conflicting social values and divergent levels of risk tolerance. The trend in application of CAMA, which integrates scientific adaptive management and stakeholder participation, is a manifestation of the increasing complexity and ‘wickedness’ of resource management problems (Scarlett 2013). Scarlett (2013) explained that CAMA presents prospects of Page 45 of 60
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
enhancing mutual learning among scientists, stakeholders, and decision makers, thereby enhancing the perceived credibility, relevance, and legitimacy of the science deployed to inform decisions. In a recent study (Caves et al. 2013), the integration of scenario planning into CAMA has been recommended to help managers and stakeholders understand how multiple external drivers may interact with ecosystem responses, and identify robust actions and monitoring efforts needed to respond to and detect the rapid and uncontrollable changes, e.g. climate shifts. One of the advantages of multi-criteria decision analysis approach in group decisions is the capacity for calling attention to similarities or potential areas of conflict between stakeholders with different views, which results in a more complete understanding of the values held by others (Linkov et al. 2004). Many models and scenarios used for multi-criteria analysis are based on optimization theories that are applicable for analysis of environmental problems which are less complex and reducible uncertainties (i.e. epistemic and linguistic). Optimal management is appropriate for cases with controllable and low uncertainty (Peterson et al. 2003). The traditional multi-criteria decision analysis lacks a participatory component, but DMCE offers an opportunity for allowing diverse views to enter the decision making process, for facilitating consensus-building, and for initiating a dynamic process of social learning (Rauschmayer and Wittmer 2006, as cited in Lieu et al. 2011). DMCE attempts to combine the advantages of traditional multi-criteria evaluation, providing structure and integration in complex decision problems, with the advantages of deliberation and stakeholder interaction provided by a `citizens' jury' (Proctor and Drechsler 2006). Functioning as a platform for risk communication, the DMCE also offers an opportunity for diverse views to enter the decision-making process and for the negotiation of consensuses (Lieu et al. 2011). Because DMCE, unlike SPA and CAMA, only deals with less complex environmental problems, it allows for risk analysis of alternative management options.
25
2.5 Key challenges, opportunities and strategies for better integration of scenarios and models in policy, planning and management [incomplete] 30
35
In the future, citizen-led assessments calling on scientific experts to review and synthesize information according to their own terms of reference are likely to become more common. These assessments may be based on empirical observation and participatory methods in order to facilitate incorporation of local knowledge, culture, internal norms and organizational structures. However, local concerns may arise due to processes occurring at much broader scales, so there is likely to be a demand for models and scenarios that can link broad-scale processes to local outcomes. A major challenge here is downscaling of global and regional scenarios at the level relevant for local decision makers and stakeholders (see chapter 6). But at the same time, there is a demand for making lowlevel data available and compatible for global and regional models of biodiversity and ecosystem services (see chapter 8). These top-down and bottom-up issues in scenarios and models require capacity-building of both scientists and decision makers (see chapter 7).
40 The issue of spatial scale appears to be a dominant challenge in relation to the use of models and scenarios in formal decision processes. Many encouraging examples of the application of models and scenarios within formal decision processes at local-national scales can be found in published literature. As we move to regional (multi-national) and global, a small but growing number of Page 46 of 60
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
examples exist in which models and scenario have been used to set the policy agenda by highlighting key future challenges to biodiversity and ecosystem services. However, this has not led to widespread adoption of formal decision approaches at those scales which provide a means to ensure that model and scenarios outputs are used efficiently and properly. Several possible reasons exist for the lack of compelling examples of structured decision approaches at global and regional scales. Political forces may not be completely comfortable with ‘handing over’ complex and sensitive decisions to technocrats using systems that policy makers don’t fully understand or trust, or they may be uncomfortable with the level of transparency about motivations, values and scientific facts that structured approaches bring to decision-making. This implies several key challenges. There is the challenge of educating policy makers to understand that involvement in decision processes doesn’t have to mean relinquishing power. Conveying the notion that structured approaches to decision making that judiciously utilize models and scenarios can help reduce complexity, distil true differences of opinion and values from linguistic ambiguities or confusion, increase mutual understanding of each others’ values, and reduce conflict.
15
Page 47 of 60
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
References:
5
10
15
20
25
30
35
40
45
Adam, S.M., Sharma, R., Bentley, N. (2013) Progress and Arrangements for Management Strategy Evaluation Work of Indian Ocean Skipjack Tuna. http://ipnlf.org/wp-content/uploads/2013/10/IOTC-2013WPTT_XX-MSE_SKJ_v3.pdf Last accessed: 09.12.2014. Adger, W. N., K. Brown, and E. L. Tompkins. 2005. The political economy of cross-scale networks in resource comanagement. Ecology and Society 10(2): 9. [online] URL: http://www.ecologyandsociety.org/vol10/iss2/art9/ Akçakaya H.R. and P. Sjögren-Gulve. (2000) Population viability analyses in conservation planning: an overview. Ecological Bulletins 48:9-21. Akçakaya, H.R., Radeloff, V.C., Mladenoff, D.J. & He, H.S. (2004). Integrating landscape and metapopulation modeling approaches: Viability of the Sharp-tailed Grouse in a dynamic landscape. Conservation Biology, 18, 536-527. Alkemade R., van Oorschot M., Miles L., et al., (2009) GLOBIO3: A framework to investigate options for reducing global terrestrial biodiversity loss. Ecosystems 12: 374-390. Allen, T. F. H. and T. B. Starr, (1982). Hierarchy: Perspectives for ecological complexity. Chicago: University of Chicago Press, 310 pp. Alves, F.L., Coelho, C., Coelho, C.D. and Pinto, P. (2011). Modelling Coastal Vulnerabilities – Tool for Decision Support System at Inter-municipality Level. Journal of Coastal Research, SI 64 (Proceedings of the 11th International Coastal Symposium), 966 – 970. Szczecin, Poland, ISSN 0749-0208 Amler, B., Betke D., Eger, H., Ehrich, C., Kohler A., Kutter, A., von Lossau, A., Müller, U., Seidemann, S., Steurer, R., W. Zimmermann. (1999). Land Use Planning Methods, Strategies and Tools. Working Group on Integrated Land Use Planning. Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH P.O. Box 5180, 65726 Eschborn, Germany Ananda, J and G. Herath. (2009). A critical review of multi-criteria decision making methods with special reference to forest management and planning. Ecological Economics 68 (2009) 2535–2548. Anbumozhi, V., Q. Chotichanathawong, and T. Murugesh. 2011. Information Disclosure Strategies for Green Industries. ADBI Working Paper 305. Tokyo: Asian Development Bank Institute. Available: http://www.adbi.org/working-paper/2011/08/22/4678.info.disclosure.strategies.green.industries Anderson, J.D. and W.J. Karel, (2009) A Genetic Assessment of Current Management Strategies for Spotted Seatrout in Texas, Marine and Coastal Fisheries: Dynamics, Management, and Ecosystem Science 1:121– 132. Antinoria, C. and G.A. Garcia-Lopez. 2008. Cross-Scale Linkages in Common-Pool Resource Management: The Evolution of Forest Associations in the Mexican Forest Commons, 12th IASC 2008 Biennial Conference, University of Gloucester, Cheltenham, England, U.K. Arrow, K.J. and Lind, R.C. (1970). Uncertainty and the evaluation of public investment decisions. American Economic Review, 60: 364 – 378. Ascough II J.C., H.R. Maier, J.K. Ravalico and M.W. Strudley, 2008. Ecological Modelling 219: 383–399, doi:10.1016/j.ecolmodel.2008.07.015 Bailey, J.K., M.A. Genung, J. O’Reilly-Wapstra, B. Potts, J. Rowntree, J. A. Schweitzer and T.G. Whitham (2012) New frontiers in community and ecosystem genetics for theory, conservation, and management, New Phytologist (2012) 193: 24–26. Baker, J.P., Hulse, D.W., Gregory, S.V., White, D., Van Sickle, J., Berger, P.A., Dole, D., Schumaker, N.H. (2004) Alternative futures for the Willamette river basin. Oregon. Ecol. Appl. 14, 313-324. (in Bryan et al. , 2011) Balint, P.J., Ronald E. Stewart, Anand Desai, and Lawrence C. Walters, 2011. Wicked environmental problems: managing uncertainty and conflict, Island Press: Washington, DC 20009, ISBN-13: 978-1-59726-474-7. Ball, I.R., H.P. Possingham, and M. Watts. 2009. Marxan and relatives: Software for spatial conservation prioritisation. Chapter 14: Pages 185-195 in Spatial conservation prioritisation: Quantitative methods
Page 48 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
and computational tools. Eds Moilanen, A., K.A. Wilson, and H.P. Possingham. Oxford University Press, Oxford, UK. Ballinger, R. C., Cummins, V. and Smith, H. D. (2010). Implementing ICZM: the experience of Northwest Europe. Ocean & Coastal Management 53(12), pp. 725-726. (10.1016/j.ocecoaman.2010.10.022) Bastian, O., Syrbe, R.U., Rosenberg, M., Rahe, D., Grunewald, K. (2013). The five pillar EPPS framework for quantifying, mapping and managing ecosystem services, Ecosystem Services 4, 15–24. Bedford, T. and Cooke, R. (2001). Probabilistic risk analysis. Foundations and methods. Cambridge University Press. Ben-Haim, Y. (2006). Info-gap Decision Theory. 2nd edn. Elsevier Ltd., Oxford, UK. Ben-Haim, Y. (2010) Info-Gap Economics: An Operational Introduction, Palgrave-Macmillan. Bennett, J. and Blamey, R. (2001). The choice modelling approach to environmental valuation. Edward Elgar, Cheltenham. Ben-Tal, A. & Nemirovski, A. (2002). Robust optimization - methodology and applications. Mathematical Programming, Series B, 92, 453-480. Berger, P. A. (2006). Generating agricultural landscapes for alternative futures analysis: A multiple attribute decision‐making model. Transactions in GIS, 10(1), 103-120. Berger, P.A., Bolte, J.P. (2004) Evaluating the impact of policy options on agricultural landscapes: an alternativefutures approach, Ecol. Appl. 14, 342-354. (in Bryan et al. , 2011) Berkes, F. 2002. Cross-scale institutional linkages for commons management: perspectives from the bottom up. Pages 293-321 in E. Ostrom, T. Dietz, N. Dolšak, P.C. Stern, S. Stonich, and E.U. Weber, editors. The Drama of the Commons. National Academy, Washington, D.C., USA. Berkes, F. 2007. Community-based conservation in a globalized world, PNAS 104(39): 15188–15193. www.pnas.org/cgi/doi/10.1073/pnas.0702098104 Berkes, F., Folke, C. (2002) Back to the future: ecosystem dynamics and local knowledge. In: Gunderson, L.H., Holling, C.S. (Eds.), Panarchy: Understanding Transformations in Systems of Humans and Nature. Island Press, Washington, DC. (in Vervoot et al., 2014) Bino, G, Jenkins, K., Kingsford, R.T. (2013) Adaptive management of Ramsar wetlands, National Climate Change Adaptation Research Facility, Gold Coast, 222 pp. Blanchard, J.L., Law, R., Castle, M.D., and Jennings, S. (2011). Coupled energy pathways and the resilience of size-structured food webs. Theoretical Ecology, 4, 289–300. Bohensky, E.L., Reyers, B., Van Jaarsveld, A.S. (2006). Future ecosystem services in a southern African River Basin: A scenario planning approach to uncertainty. Conservation Biology, 20(4), 1051–1061. Bolte, J.P., Hulse, D.W., Gregory, S.V., Smith, C. (2004) Modeling biocomplexity e actors, landscapes and alternative futures. Environ. Modell. Softw. 22, 570-579. (in Bryan et al. , 2011) Börjeson, L., Höjer, M., Dreborg, K.H., Ekvall, T., Finnveden, G. (2006) Scenario types and techniques: Towards a user’s guide, Futures, 38, 723–739. Borrini-Feyerabend, G., M. Pimbert, M. T. Farvar, A. Kothari and Y. Renard, Sharing Power. 2004. Learning by doing in co-management of natural resources throughout the world, IIED and IUCN/CEESP/ CMWG, Cenesta, Tehran. Brodie, J.E., et al., 2012. Terrestrial pollutant runoff to the Great Barrier Reef: An update of issues, priorities and management responses. Mar. Pollut. Bull., doi:10.1016/j.marpolbul.2011.12.012 Budreau, D. and McBean, G. 2007, Climate change, adaptive capacity and policy direction in the Canadian North: Can we learn anything from the collapse of the east coast cod fishery? Mitig Adapt Strat Glob Change 12:1305–1320 Bunnefeld, N., Hoshino, E., Milner-Gulland, E.J. (2011) Management strategy evaluation: a powerful tool for conservation?. Trends in ecology & evolution, 26(9), 441-447. Burgman, M.A. (2005). Risks and decisions for conservation and environmental management. Cambridge University Press, Cambridge. Burkhard, B., Crossman, N., Nedkov, S., Petz, K., Alkemade, R., (2013) Mapping and modelling ecosystem services for science, policy and practice. Ecosystem Services, 4, 1-3. Page 49 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Butterworth, D. S. (2007) Why a management procedure approach? Some positives and negatives. ICES Journal of Marine Science, 64, 613–617 Carmichael, J., Tansey, J., Robinson, J. ( 2004) An integrated assessment modelling tool. Glob. Environ. Change, 14, 171-183 . (in Bryan et al. , 2011) Cash, D. W., W. Adger, F. Berkes, P. Garden, L. Lebel, P. Olsson, L. Pritchard, and O. Young. 2006. Scale and cross-scale dynamics: governance and information in a multilevel world. Ecology and Society 11(2): 8. [online] URL: http://www.ecologyandsociety.org/vol11/iss2/art8/ Caves, J. K., G. S. Bodner, K. Simms, L. A. Fisher, and T. Robertson. 2013. Integrating collaboration, adaptive management, and scenario-planning: experiences at Las Cienegas National Conservation Area. Ecology and Society 18(3): 43. http://dx.doi.org/10.5751/ES-05749-180343 Chades, I., McDonald-Madden, E., McCarthy, M.A., Linkie, M., Wintle, B.A. & Possingham, H.P. (2008). Save, survey or surrender: optimal management of a cryptic threatened species. Proceedings of the National Academy of Science, 105, 13936–13940. Chankong, V. and Haimes, Y.Y. (2008). Multiobjective decision making. Theory and methodology. Dover, New York. Chave, J. 2013. The problem of pattern and scale in ecology: what have we learned in 20 years? Ecology Letters, 16: 4–16 doi: 10.1111/ele.12048. Cheng, A.S. and S.E. Daniels. 2003. Examining the Interaction Between Geographic Scale and Ways of Knowing in Ecosystem Management: A Case Study of Place-Based Collaborative Planning, Forest Science 49(6): 841–854. Conlisk, E., Alexandra D. Syphard, Janet Franklin, Lorraine Flint, Alan Flint and Helen Regan, 2013. Uncertainty in assessing the impacts of global change with coupled dynamic species distribution and population models, Global Change Biology 19: 858–869, doi: 10.1111/gcb.12090. Conroy, M. J., Michael C. Runge, James D. Nichols, Kirk W. Stodola, Robert J. Cooper, 2011. Conservation in the face of climate change: The roles of alternative models, monitoring, and adaptation in confronting and reducing uncertainty, Special Issue Article: Adaptive management for biodiversity conservation in an uncertain world, Biological Conservation 144: 1204–1213, doi:10.1016/j.biocon.2010.10.019. Cook, C.N., Inayatullah, S., Burgman, M.A., Sutherland, W.J., Wintle, B.A. (2014a) Strategic foresight: How planning for the unpredictable can improve environmental decision-making. Trends in Ecology and Evolution, 29(9): 531-541. Cook, C.N., Wintle, B.A., Aldrich, S.C., Wintle, B.C. (2014b) Using strategic foresight to assess conservation opportunity. Conservation Biology, 28(6): 1474-1483. Crist, M. R., Wilmer, B. O., Aplet, G. H. (2005). Assessing the value of roadless areas in a conservation reserve strategy: biodiversity and landscape connectivity in the northern Rockies. Journal of applied ecology, 42(1), 181-191. Crooks, K. R., and Sanjayan, M. eds. (2006) Connectivity Conservation. Cambridge University PRess. Cambridge. Crossman, N.D., Burkhard, B., Nedkov, S., Willemen, L., Petz, K., Palomo, I., Drakou, E. G., Martín-Lopez, B., McPhearson, T., Boyanova, K., Alkemade, R., Egoh, B., Dunbar, Martha B., Maes, J., (2013) A blueprint for mapping and modelling ecosystem services, Ecosystem Services 4, 4-14. Cumming, G. S., Cumming, D. H. M., Redman, C. L., (2006) Scale mismatches in social-ecological systems: causes, consequences, and solutions. Ecology and Society 11 de Moor, C. L., Butterworth, D. S., and De Oliveira, J. A. A. (2011) Is the management procedure approach equipped to handle short-lived pelagic species with their boom and bust dynamics? The case of the South African fishery for sardine and anchovy. ICES Journal of Marine Science, 68: 2075–2085. de Moor, C.L. (2014) Final anchovy TAC and sardine TAB for 2014, using interim OMP-13v3. DAFF Branch Fisheries document: FISHERIES/2014/JUN/SWG-PEL/36: 4pp. De Oliveira, J. A. A., Butterworth, D. S. (2004) Developing and refining a joint management procedure for the multispecies South African pelagic fishery. ICES Journal of Marine Science, 61: 1432-1442.
Page 50 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Delavenne, J., Metcalfe, K., Smith, R. J., Vaz, S., Martin, C. S., Dupuis, L., Coppin, F., and Carpentier, A., 2011. Systematic conservation planning in the eastern English Channel: comparing the Marxan and Zonation decision-support tools. – ICES Journal of Marine Science, doi:10.1093/icesjms/fsr180. Dickey-Collas, M.; Engelhard, G. H. ; Rindorf, Anna; Raab, K.; Smout, S.; Aarts, G. ; Deurs, Mikael van; Brunel, T.; Hoff, Ayoe; Lauerburg, R.A.M. ; Garthe, S.; Andersen, Ken Haste; Scott, F.; van Kooten, T.; Beare, D.; Peck, M.A. (2014) Ecosystem-based management objectives for the North Sea : riding the forage fish rollercoaster. ICES Journal of Marine Science, 71(1), 128-142. Driscoll, Don A., Lindenmayer, David B., Bennett, Andrew F., Bode, Michael, Bradstock, Ross A., Cary, Geoffrey J., Clarke, Michael F., Dexter, Nick, Fensham, Rod, Friend, Gordon, Gill, Malcolm, James, Stewart, Kay, Geoff, Keith, David A., MacGregor, Christopher, Russell-Smith, Jeremy, Salt, David, Watson, James E.M., Williams, Richard J. and York, Alan, (2010). Fire management for biodiversity conservation: key research questions and our capacity to answer them, Biological conservation, vol. 143, no. 9, pp. 1928-1939. Du, J., Yang, Y., Xu, L., Zhang, S., Yang, F. (2012) Research on the alternatives in a strategic environmental assessment based on the extension theory. Environmental monitoring and assessment, 184(9), 58075819. Edwards, C.T., Bunnefeld, N., Balme, G.A., Milner-Gulland, E. J. (2014) Data-poor management of African lion hunting using a relative index of abundance. Proceedings of the National Academy of Sciences, 111(1), 539-543. Elith, J., Leathwick, J. (2009) The contribution of species distribution modelling to conservation prioritization. Pp. 70–93 in A. Moilanen, K. A. Wilson, H. P. Possingham, eds. Spatial conservation prioritization. Quantitative methods and computational tools. Oxford Univ. Press, Oxford, U.K. Espindola, S., A.D. Cuaron, O.E. Gaggiotti, E. Vazquez-Dominguez (2014) High genetic structure of the Cozumel Harvest mice, a critically endangered island endemic: conservation implications, Conserv Genet (2014) 15:1393–1402. FAO (2013) scoping policy analysis for aquatic genetic resources, CGRFA-14/13/Inf.24, Geneva. Feaster, T.D., Benedict, S.T., Clark, J.M., Bradley, P.M., Conrads, P.A., 2014. Scaling up watershed model parameters—flow and load simulations of the Edisto River Basin, South Carolina, 2007–09: U.S. Geological Survey Scientific Investigations Report 2014-5104, 34 p. Fenton, N. and Neil, M., 2012. Risk Assessment and Decision Analysis with Bayesian Networks. CRC Press. Figueira, J. Greco, S. and Ehrogott, M. (2005). Multiple criteria decision analysis: State of the art surveys. Springer, New York. Fischer, G. W. (1995) Range sensitivity of attribute weights in multiattribute value models. Organizational Behavior and Human Decision Processes, 62(3), 252-266. Fischer, T. B. (2002). Strategic environmental assessment in transport and land use planning. Earthscan, London. Fischer, T.B. 2007. Theory and Practice of Strategic Environmental Assessment – towards a more systematic approach, Earthscan, London (also published in Chinese by Science Press in 2008). Fisher, B., Turner, R. K., Morling, P. (2009) Defining and classifying ecosystem services for decision making, Ecological Economics 68, 643-653. Flather CH, Hayward GD, Beissinger SR, Stephens PA (2011) Minimum viable populations: Is there a ‘magic number’ for conservation practitioners? Trends in Ecology and Evolution 26:307-316. Forman, E.H. and Selly, M.A. (2001). Decision by objectives. How to convince others that you are right. World Scientific Publishing, Singapore. Fortin, M. J., Boots, B., Csillag, F., & Remmel, T. K. (2003). On the role of spatial stochastic models in understanding landscape indices in ecology. Oikos, 102(1), 203-212. Franklin JF (1993) Preserving biodiversity: Species, ecosystems, or landscapes? Ecological Applications 3:202205. Franklin, J.F. 1988. ‘Structural and functional diversity in temperate forests’, in Wilson, E.O. (Ed.), Biodi6ersity. National Academy Press, Washington, D.C. pp. 166–175. Page 51 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Fremier, Alexander K., Fabrice A. J. DeClerck, Nilsa A. Bosque-Pérez, Natalia Estrada Carmona, Renée Hill, Ta ylor Joyal, Levi Keesecker, P. Z ion Klos, Alejandra Martínez-Salinas, Ryan Niemeyer, Andre Sanfiorenzo, Kristen Welsh, and J. D . Wulfhorst. 2013. Understanding Spatiotemporal Lags in Ecosystem Services to Improve Incentives, BioScience 63(6): 472-482. García-Nieto, A.P.,García-Llorente, M., Iniesta-Arandia, I. ,Martín-López, B.,2013. Mapping forest ecosystem services: from providing units to beneficiaries. Ecosystem Services 4,126–138. Gordon, Theodore J. and Helmer, Olaf, (1964) Report on (I Long-Range Forecasting Sfrrdy, Rand Paper P-2982. Granberg, M., Rolf Lidskog and Stig Larsson, 2008. Dealing with uncertainty: a case study of controlling insect populations in natural ecosystems, Local Environment 13(7): 641–652, DOI: 10.1080/13549830802260191. Greenberg et al. 2002 - Greenberg J.D., Gergel S.E. and Turner M.G. (2002). Understanding landscape metrics II. In: Gergel S.E. and Turner M.G. (eds.). Learning landscape ecology, pp. 101-111. Springer, New York. Greenstreet, Simon P. R.; Rossberg, Axel G.; Fox, Clive J.; Le Quesne, William J. F.; Blasdale, Tom; Boulcott, Philip; Mitchell, Ian; Millar, Colin; Moffat, Colin F. (2012) Demersal fish biodiversity: species-level indicators and trends-based targets for the Marine Strategy Framework Directive. ICES Journal of Marine Science, 69(10) 1789-1801. Gregory, R., Failing, L., Harstone, M., Long, G., McDaniels, T. and Ohlson, D. (2012). Structured decision making. A practical guide to environmental management choices. Wiley-Blackwell, Chichester. Guisan, A., Tingley, R., Baumgartner, J.B., Naujokaitis-Lewis, I., Sutcliffe, P.R., Tulloch, A.I.T. et al. (2013). Predicting species distributions for conservation decisions. Ecology Letters, 16, 1424-1435. Gupta, Aarti. 2008. Transparency Under Scrutiny: Information Disclosure in Global Environmental Governance, Global Environmental Politics 8(2): 1-7. Hajkowicz S.A., McDonald G.T, Smith P.N. (2000). An evaluation of multiple objective decision support weighting techniques in natural resource management. Journal of Environmental Planning and Management, 43: 505-518. Hajkowicz, S.A. (2008). Rethinking the economist’s evaluation toolkit in light of sustainability policy. Sustainability Science, Practice and Policy, 4, 1-8. Hall, Nicholas, John Hershey, Larry Kessler, R. Craig Stotts. (1992). A Model for Making Project Funding Decisions at the National Cancer Institute. Operations Research 40; 1040-1052 Hammond, J.S., Keeney, R.L. & Raiffa, H. (1999). Smart choices: A practical guide to making better life decisions. Broadway Books, New York Happe, K., Kellermann, K., Balmann, A. (2006) Agent-based analysis of agricultural policies: an illustration of the agricultural policy simulator Agripolis, its adaptation and behaviour. Ecol. Soc. 11, U481-U507. (in Bryan et al. , 2011) Hastie, R. and Dawes, R.M. (2010). Rational choice in an uncertain world. The psychology of judgment and decision making. Sage, Los Angeles. He F. and Hubbell S.P. (2011) Species-area relationships always overestimate extinction rates from habitat loss. Nature 473:368-371. Hersch-Green, E.I., N.E. Turley and M.T. J. Johnson (2011) Community genetics: what have we accomplished and where should we be going?, Phil. Trans. R. Soc. B (2011) 366, 1453–1460, doi:10.1098/rstb.2010.0331. Hess, G.R. & King, T.J. (2002) Planning open spaces for wildlife 1. Selecting focal species using a Delphi survey approach. Landscape and Urban Planning, 58. Holland, D.S. (2010) Management strategy evaluation and management procedures: tools for rebuilding and sustaining fisheries”, OECD Food, Agriculture and Fisheries Working Papers, No. 25, OECD Publishing. doi: 10.1787/5kmd77jhvkjf-en Howard, R.A. (2007). The foundations of decision analysis revisited. In: W. Edwards, R.F. Miles Jr., D. von Winterdfeldt, D. (eds). Advances in decision analysis. From foundations to applications. Cambridge University Press, Cambridge. Page 52 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Howlett, M. Ramesh, M. & A. Perl. (2009). Studying Public Policy: Policy cycles & policy subsystems. Don Mills. Oxford University Press. Hulse, D.W., Branscomb, A., Payne, S.G. (2004) Envisioning alternatives: using citizen guidance to map future land and water use, Ecol. Appl., 14, 325-341. (in Bryan et al. , 2011) Huntley, W.L., Bock, J.G., Weingartner, M. (2010) Planning the unplannable: Scenarios on the future of space, Space Policy, 26, 25-38. Hutchings, J.A. 2000. Collapse and recovery of marine fishes. Nature (London) 406: 882-885. Hutchings, J.A., and Reynolds, J.D. 2004. Marine fish population collapses: Consequences for recovery and extinction risk. BioScience 54: 297-309. Ibrekk, H., Morgan, M.G., 1987. Graphical communication of uncertain quantities to nontechnical people. Risk Analysis 7, 519–529. Jellinek S., Rumpff R., Driscoll D.A., Parris K.M, Wintle B.A. 2014. Modelling the benefits of habitat restoration in socio-ecological systems. Biological Conservation 169: 60-67. Jennings, S. and Rice, J., 2011. Towards an ecosystem approach to fisheries in Europe: a perspective on existing progress and future directions. Fish and Fisheries 12, 125–137 Johnson, F., and K. Williams. (1999). Protocol and practice in the adaptive management of waterfowl harvests. Conservation Ecology 3(1): 8. [online] URL: http://www.consecol.org/vol3/iss1/art8/ Johnson, F.A., Moore, C.T., Kendall, W.T., Dubovsky, J.A., Caithamer, D.F., Kelley, J.R. et al. (1997). Uncertainty and the management of mallard harvests. Journal of Wildlife Management, 61, 202-216. Joseph, L.N., Maloney, R.F. & Possingham, H.P. (2009). Optimal allocation of resources among threatened species: a project prioritization protocol. Conservation Biology, 23, 328-338. Kahneman, D. & Tversky, A. (2000). Choices, values, and frames. Cambridge University Press, Cambridge. Kearney, M. (2006). Habitat, environment and niche: what are we modelling? Oikos, 115, 186-191. Keeney R.L (2007). Developing Objetives and Attributes. In W. Edwards, R.F. iles JR., & D. Von Winterfeldt (Eds.). Advances in Decision Analysis: From Foundations to Applications. Cambridge: Cambridge University Press, 104-128. Keeney, R.L. (1982). Decision analysis: an overview. Operations Research, 30, 803-838. Keeney, R.L. (2002). Common mistakes in making value trade-offs. Operations Research, 50, 935-945. King, A.W., (1991) Translating models across scales in the landscapes. In: Turner, M.G., Gardner, R.H. (Eds.). Quantitative methods in Landscape Ecology. Ecological Studies 62, Springer-Verlag New York, NY, pp. 479-517. (In Tang et.al, 1997) Knight, F.H. (1921). Risk, uncertainty, and profit. Houghton Mifflin Co. Kremen, C., A. Cameron, A. Moilanen, S.J. Phillips, C.D. Thomas, H. Beentje, J. Dransfield, B.L. Fisher, F. Glaw, T.C. Good, G.J. Harper, R.J. Hijmans, D.C. Lees, E. Louis, R.A. Nussbaum, C.J. Raxworthy, A. Razafimpahanana, G.E. Schatz, M. Vences, D.R. Vieites, P.C. Wright, and M.L. Zjhra, (2008) Aligning conservation priorities across taxa in Madagascar with high-resolution planning tools. Science, 320(5873): p. 222-226. Krupnick, A., Morgenstern, R., Batz, M., Nelson, P., Burtraw, D., Shih, J.S., McWilliams, M., 2006. Not A Sure Thing: Making Regulatory Choices under Uncertainty. Resources for the Future (RFF), Washington. Lahoz-Monfort JJ, Guillera-Arroita G and Hauser CE (2014) From planning to implementation: explaining connections between adaptive management and population models. Front. Ecol. Evol. 2:60. doi: 10.3389/fevo.2014.00060 Lambin, E., Geist, H. and Rindfuss, R.R., (2005). Land-Use and Land-Cover Change: Developing and Implementing an Agenda for Local Processes with Global Impacts, IHDP Update, pp. 1-3. Langhans, S., Reichert, P., Schuwirth, N. (2014) The method matters: a guide for indicator aggregation in ecological assessments. Ecological Indicators 45, 494–507. Leadley, P., Pereira, H.M., Alkemade, R., Fernandez-Manjarres, J.F., Proenca, V., Scharlemann, J.P.W. et al. (2010). Biodiversity scenarios: projections of 21st century change in biodiversity and associated ecosystem services. In: Secretariat of the Convention on Biological Diversity (ed. Diversity SotCoB). Page 53 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Published by the Secretariat of the Convention on Biological Diversity, Montreal, p. 1–132. Technical Series no. 50. Leadley, P.W., Krug, C.B., Alkemade, R., Pereira, H.M., Sumaila U.R., Walpole, M., Marques, A., Newbold, T., Teh, L.S.L, van Kolck, J., Bellard, C., Januchowski-Hartley, S.R. and Mumby, P.J. (2014): Progress towards the Aichi Biodiversity Targets: An Assessment of Biodiversity Trends, Policy Scenarios and Key Actions. Secretariat of the Convention on Biological Diversity, Montreal, Canada. Technical Series 78, 500 pages. Leslie, H., R. Ruckelshaus, I. R. Ball, S. Andelman and H. P. Possingham. 2003. Using siting algorithms in the design of marine reserve networks. Ecological Applications 13: S185-S198 Lessmann, J., Munoz, J., Bonaccorso, J (2014) Maximizing species conservation in continental Ecuador: a case of systematic conservation planning for biodiverse regions. Ecology and Evolution 2014; 4(12): 2410–2422. Linkov, I., A. Varghese, S. Jamil, T.P. Seager G. Kiker and T. Bridges, 2004, Multi-Criteria Decision Analysis: A Framework for Structuring Remedial Decisions at Contaminated Sites, in Linkov, I. and Ramadan, A. eds “Comparative Risk Assessment and Environmental Decision Making” Kluwer, 2004, p. 15-54. Linstone, Harold A. and Turoff, Murray, Eds., (1975) The Delphi Method: Techniques and Applications,AddisonWesley, Reading, Mass. Liu, S., Michael Hurley, Kim E. Lowell, Abu-Baker M. Siddique, Art Diggle, David C. Cook, (2011). An integrated decision-support approach in prioritizing risks of non-indigenous species in the face of high uncertainty, Ecological Economics 70: 1924–1930, doi:10.1016/j.ecolecon.2011.05.021. Liu, Y., Guo, H., Zhang, Z., Wang, L., Dai, Y., & Fan, Y. (2007). An optimization method based on scenario analysis for watershed management under uncertainty. Environmental Management, 39(5), 678-690. Liu, Z. S., Shang, J. C., Yang, D. M., Yao, W. G. (2006) Mathematical model of alternative schemes in strategic environmental assessment for mineral resources exploitation. Journal of Northeast Normal University, 1(38), 121–126. Loftin, M. K. 2014. Truths and governance for adaptive management. Ecology and Society 19(2): 21. http://dx.doi.org/10.5751/ES-06353-190221 Ludwig, D., Mangel, M. & Haddad, B. (2001). Ecology, conservation and public policy. Annual Reviews in Ecology and Systematics, 32, 481-517. Mace, G., Perrings, C., Le Prestre P.L., Cramer, W., Diaz, S., Anne Larigauderie A., Scholes, R.J. and Mooney H. A. (2013). Science to Policy Linkages for the Post-2010 Biodiversity Targets. In: Biodiversity Monitoring and Conservation: Bridging the Gap between Global Commitment and Local Action, First Edition. Edited by Ben Collen, Nathalie Pettorelli, Jonathan E.M. Baillie and Sarah M. Durant. Madu, Christian, Chu-Hua Kei, Assumpta Madu, (1991) "Setting Priorities for the IT Industry in Taiwan - A Delphi Study", Lonc Range Planning, v24, n5, pp105-118. Maguire L.A. (2004). What can decision analysis do for invasive species management? Risk Analysis, 24: 859 – 868. Maier H.R. and J.C. Ascough II, 2006. Uncertainty in Environmental Decision-Making: Issues, Challenges and Future Directions, The International Environmental Modelling & Software Society, http://www.iemss.org/iemss2006/papers/w2/pp.pdf Margules, C. R., and R. L. Pressey. 2000. Systematic conservation planning. Nature 405:243–252. McCarthy, M.A. & Possingham, H.P. (2007). Active adaptive management for conservation. Conservation Biology, 21, 956-963. McCarthy, M.A. & Thompson, C. (2001). Expected minimum population size as a measure of threat. Animal Conservation, 4, 351-355. McDonald-Madden, E., Baxter, P.W.J., Fuller, R.A., Martin, T.G., Game, E.T., Montambault, J. et al. (2010a). Monitoring does not always count. Trends in Ecology & Evolution, 25, 547-550. McDonald-Madden, E., Probert, W., Hauser, C., Runge, M., Jones, M., Moore, J. et al. (2010b). Active adaptive conservation of threatened species in the face of uncertainty. Ecological Applications, 20, 1476–1489. McFadden, J.E., Tim L. Hiller, Andrew J. Tyre, (2011). Evaluating the efficacy of adaptive management approaches: Is there a formula for success? Journal of Environmental Management 92: 1354-1359. Page 54 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
MEA (Millennium Ecosystem Assessment) (2005) Key Questions on Biodiversity in the Millennium Ecosystem Assessment, Ecosystems and Human Well-Being: Biodiversity Synthesis. Washington, DC: Island Press. Melbourne B.A. and Hastings A. (2008) Extinction risk depends strongly on factors contributing to stochasticity. Nature 454:100-103. Mendoza, G.A. and Martins, H. (2006). Multi-criteria analysis in natural resource management: a critical review of methods and new modeling paradigms. Forest Ecology and Management, 230: 1 -22. Meyer, B. C., & Grabaum, R. (2008) MULBO: Model framework for multicriteria landscape assessment and optimisation. A support system for spatial land use decisions. Landscape Research, 33(2), 155-179. Minas, J.P., Hearne, J.W. and Handmer, J.W. (2012). A review of operations research methods applicable to wildfire management. International Journal of Wildland Fire, 21: 189 – 196. Moffett, A. and Sarkat, S. (2006). Incorporating multiple criteria into the design of conservation area networks: a minireview with recommendations. Diversity and Distributions, 12: 125 – 137. Moilanen, A., Franco, A.M.A., Early, R.I., Fox, R., Wintle, B.A. & Thomas, C.D. (2005). Prioritizing multiple-use landscapes for conservation: methods for large multi-species planning problems. Proceedings of the Royal Society B, 272, 1885-1891. Mosadeghi, R., Jan Warnken, Rodger Tomlinson & Hamid Mirfenderesk, 2012. Uncertainty analysis in the application of multi-criteria decision-making methods in Australian strategic environmental decisions, Journal of Environmental Planning and Management, DOI:10.1080/09640568.2012.717886 Mubareka, S.B., Maes, J., Lavalle, C. ,De Roo, A., 2013.Estimation of water requirements by Livestock in Europe. Ecosystem Services 4,139–145. Mustajoki J., Hämäläinen R.P. and Marttunen M. (2004). Participatory multicriteria decision analysis with WebHIPRE: a case of lake regulation policy. Environmental Modelling & Software, 19, 537-547 Nichols, J.D. & Williams, B.K. (2006). Monitoring for conservation. Trends in Ecology and Evolution, 21, 668-673. Nicholson E, Collen B, Barausse A, Blanchard JL, Costelloe BT, et al., 2012. Making Robust Policy Decisions Using Global Biodiversity Indicators. PLoS ONE 7(7): e41128. doi:10.1371/journal.pone.0041128 Nobre A., D. 2014. The Future CLimate of Amazonia. Scientific Assessment Report. ARA (Articulacion Regional Amazonica). São José dos Campos SP, Brazil - October 2014 Nordström, E., Eriksson L., Öhman
K. 2010. Integrating multiple criteria decision analysis in participatory forest planning: Experience from a case study in northern Sweden. Forest Policy and Economics, 2010, vol. 12: 562-574 Norton, B., and Steinemann, 2002. A. "Environmental Values and Adaptive Management." In Searching for Sustainability, ed. Norton, B. Cambridge, U.K: Cambridge University Press, pp. 514-547. Noss, R. F., and Daly, K.M. (2006). Incorporating connectivity into broad scale planning. PP 587-619 In: Crooks, K. R., and Sanjayan, M. eds. Connectivity Conservation. Cambridge University PRess. Cambridge. Noss, R.F. (1990) Indicators for Monitoring Biodiversity: A Hierarchical Approach, Conservation Biology, Volume 4, No. 4. Nuno, A., N. Bunnefeld, and E. Milner-Gulland. (2014) Managing social–ecological systems under uncertainty: implementation in the real world. Ecology and Society 19(2): 52 Pacala, S.W. and Deutschman, D.H., (1995) Details that matter: the spatial distribution of individual trees maintains forest ecosystem function, Oikos, 74, 357-365. (In Tang et.al, 1997) Patel, M., Kok, K., Rothman, D.S. (2007) Participatory scenario construction in land use analysis: an insight into the experiences created by stakeholder involvement in the northern Mediterranean. Land Use Policy 24, 546-561. (in Bryan et al. , 2011) Pereira, H. M., Leadley, P. W., Proença, V., Alkemade, R., Scharlemann, J. P., Fernandez-Manjarrés, J. F., Walpole, M. (2010) Scenarios for global biodiversity in the 21st century. Science, 330(6010), 1496-1501. Petersen, A.C., P.H.M. Janssen, J.P. van der Sluijs, J.S. Risbey, J.R. Ravetz, J.A. Wardekker, H. Martinson Hughes, 2013. Guidance for Uncertainty Assessment and Communication, 2nd Edition, PBL, The Netherlands. Peterson, G.D., Cumming, G.S. & Carpenter, S.R. (2003). Scenario planning: a tool for conservation in an uncertain world. Conservation Biology, 17, 358-366. Page 55 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Pettit, C., Pullar, D. (2004) A way forward for land-use planning to achieve policy goals by using spatial modelling scenarios, Environ. Plan. B-Plan. Des., 31, 213-233. (in Bryan et al. , 2011) Plagányi, É. E., Rademeyer, R. A., Butterworth, D. S., Cunningham, C. L., and Johnston, S. J. (2007) Making management procedures operational — innovations implemented in South Africa. ICES Journal of Marine Science, 64, 626–632 Plagányi, É. E., van Putten, I., Hutton, T., Deng, R. A., Dennis, D., Pascoe, S., Skewes, T., Campbell, R. A. (2013). Integrating indigenous livelihood and lifestyle objectives in managing a natural resource. Proceedings of the National Academy of Sciences, 110(9), 3639-3644. Possingham, H. P., (2001) The Business of Biodiversity: Applying decision theory principles to nature conservation, The Australian Conservation Foundation. Possingham, H. P., K. A. Wilson, S. J. Andelman, and C. H. Vynne. (2006). Protected areas. Goals, limitations, and design. Pp. 507–549 in M. J. Groom, G. K. Meffe, C. R. Carroll, eds. Principles of conservation biology. 3rd ed. Sinauer Associates, Inc., Sunderland, MA. Proctor, W. and M. Drechsler, (2006). Deliberative multicriteria evaluation, Environment and Planning C: Government and Policy, 24: 169-190, DOI:10.1068/c22s Punt, A. E. and Donovan, G. P. (2007). Developing management procedures that are robust to uncertainty: lessons from the International Whaling Commission. ICES Journal of Marine Science, 64: 603–612. Rademeyer, R. A., Plagányi, É. E., and Butterworth, D. S. (2007). Tips and tricks in designing management procedures. ICES Journal of Marine Science, 64, 618–625. Raiffa, H. (1968). Decision analysis: Introductory lectures on choices under uncertainty. Addison-Wesley, Reading Ramos M.H., Thibault Mathevet, Jutta Thielen and Florian Pappenberger, (2010). Communicating uncertainty in hydro-meteorological forecasts: mission impossible? Meteorol. Appl. 17: 223–235, DOI: 10.1002/met.202. Rapport, D.J., Gaudet, C., Karr, J.R., Baron, J.S., Bohlen, C., Jackson, W., Jones, B., Naiman, R.J., Norton, B., Pollock, M.M., (1998) Evaluating landscape health: integrating societal goals and biophysical process, Journal of Environmental Management 53, 1-15. Rastetter, E.B., King, A.W., Cosby, B.J., Hornberger, G.M., O’Neill, R.V., Hobbie, J.E. (1992) Aggregating finescale ecological knowledge to model coarser-scale attributes of ecosystems, Ecol. Appl. 2, 55-70. (In Tang et.al, 1997) Rauch, W. 1979. The Decision Delphi. Technological Forecasting And Social Change 15, 159-169 Raupach, M.R ., P. J. Rayner, D. J. Barrett, R. S. Defries, M. Heimann, D . S . O Jima, S. Quegan and C.C. Schmullius. 2005. Model–data synthesis in terrestrial carbon observation: methods, data requirements and data uncertainty specifications, Global Change Biology 11: 378–397, doi: 10.1111/j.13652486.2005.00917.x Reeth W.V. (2013) Chapter 4 - Ecosystem Service Indicators: Are We Measuring What We Want to Manage?, In Ecosystem Services, edited by Sander JacobsNicolas DendonckerHans Keune, Elsevier, Boston, 41-61 Refsgaard,J.C.,van der Sluijs,J.P.,Højberg,A.L.,Vanrolleghem, P.A. 2007. Uncertainty in the environmental modelling process - A framework and guidance. Environ. Model. Softw. 22, 11, 1543-1556. Regan, H.M., Ben-Haim, Y., Langford, B., Wilson, W.G., Lundberg, P., Andelman, S.J. et al. (2005). Robust decision making under severe uncertainty for conservation decision making. Ecological Applications, 15, 1471-1477. Regan, H.M., Colyvan, M. and Burgman, M.A. (2002). A taxonomy and treatment of uncertainty for ecology and conservation biology. Ecological Applications, 12: 618 – 628. Reichert, P., Borsuk, M., Hostmann, M., Schweizer, S., Spörri, C., Tockner, K., Truffer, B. (2007) Concepts of decision support for river rehabilitation, Environmental Modelling & Software 22, 188-201. Reichert, P., Schuwirth, N., Langhans, S. (2013) Constructing, evaluating and visualizing value and utility functions for decision support, Environmental Modelling & Software 46, 283-291. Renn, Ortwin. 2006. Participatory Processes for Designing Environmental Policies. Land Use Policy. 23:34-43. Page 56 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Richards, S.A., Possingham, H.P., and Tizard, J. (1999) Optimal fire management for maintaining habitat diversity. Ecological Applications, 9: 880-892. Ricotta, C., Stanisci, A., Avena, G. C., Blasi, C. (2000). Quantifying the network connectivity of landscape mosaics: a graph-theoretical approach. Community Ecol. 1, 89–94. Rodrigues, A. S. L., H. R. Akcakaya, S. J. Andelman, M. I. Bakarr, L. Boitani, T. M. Brooks, et al. 2004a. Global gap analysis: priority regions for expanding the global protected-area network. Biosci. Rep. 54:1092–1100. Rodrigues, A. S. L., S. J. Andelman, M. I. Bakarr, L. Boitani, T. M. Brooks, R. M. Cowling, et al. 2004b. Effectiveness of the global protected area network in the representing species diversity. Nature, 428:640–643. Rombouts, I., Beaugrand, G., Fizzala, X., Gaill, F., Greenstreet, S.P.R., Lamare, S., Le Loc’h, F., McQuattersGollop, A., Mialet, B., Niquil, N., Percelay, J., Renaud, F., Rossberg, A.G., Féral, J.P. (2013) Food web indicators under the Marine Strategy Framework Directive: From complexity to simplicity?, Ecological Indicators, 29, 246-254. Rout, T., Hauser, C. & Possingham, H.P. (2009). Optimal adaptive management for the translocation of a threatened species. Ecological Applications, 19, 515-526. Rowland, E.R., Cross, M.S., Hartmann, H. (2014) Considering Multiple Futures: Scenario Planning To Address Uncertainty in Natural Resource Conservation. Washington, DC: US Fish and Wildlife Service. Roy, B. (1973) How outranking relation helps multiple criteria decision making. In: Cochrane, J.L., Zeleny, M. (Eds.), Multiple Criteria Decision Making, University of South-Caroline Press, Columbia, SC, pp. 179–201. Runge MC. 2011. Adaptive management for threatened and endangered species. Journal of Fish and Wildlife Management 2:220–233. Runge, M.C., Bean, Ellen, Smith, D.R., and Kokos, Sonja. 2011a. Non-native fish control below Glen Canyon Dam—Report from a structured decision-making project: U.S. Geological Survey Open-File Report 2011– 1012, 74 p., at http://pubs.usgs.gov/of/2011/1012/. Runge, M.C., Converse, S.J. & Lyon, J.E. 2011b. Which uncertainty? Using expert elicitation and expected value of information to design an adaptive program. Biological conservation, 144, 1214-1223. Saaty T.L (1980) The Analytic Hierarchy Process: Planning, Priority Setting, Resource Allocation, ISBN 0-07054371-2, McGraw-Hill. Sachindra, D. A., Huang F., Barton A., Perera B. J. C. (2014). Statistical downscaling of general circulation model outputs to catchment scale hydroclimatic variables: issues, challenges and possible solutions. Journal of Water and Climate Change Vol 5 No 4 pp 496–525 Sanderson, E.W., Jaiteh M., Levy M.A., Redford K.H., Wannebo A.V., Woolmer G. (2002). The human footprint and the last of the wild. Bioscience 52, 891–904. Santelmann, M.V., White, D., Freemark, K., Nassauer, J.I., Eilers, J.M., Vache, K.B., Danielson, B.J., Corry, R.C., Clark, M.E., Polasky, S., Cruse, R.M., Sifneos, J., Rustigian, H., Coiner, C., Wu, J., Debinski, D. ( 2004) Assessing alternative futures for agriculture in Iowa, U.S.A. Landscape Ecol. 19, 357-374. (in Bryan et al. , 2011) SAPM 2009. Madagascar Protected Areas Committee. Madagascar Digital Conservation Atlas Report. Unpublished. (http://41.74.23.114:8080/atlas/index.php?option=com_docman&task=doc_download&gid=31&Itemid =29) Savage, Leonard J. (1954). The Foundations of Statistics. New York, Wiley. Scarlett, L. 2013. Collaborative adaptive management: challenges and opportunities. Ecology and Society 18(3):26. http://dx.doi.org/10.5751/ES-05762-180326 SCBD GBO4 - Secretariat of the Convention on Biological Diversity (2014) "Global Biodiversity Outlook 4." Montréal, 155 pages. Schoemaker, P.J.H. (1993) Multiple scenario development: Its conceptual and behavioral foundation. Strategic Management Journal, 14, 193-213. Schoemaker, P.J.H. (1995) Scenario planning - A tool for strategic thinking. Sloan Management Review, 36, 2540. Page 57 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Schwartz P. (1996). The Art of the Long View: Planning for the Future in an Uncertain World, Doubleday, New York. (in Shearer, A.W. (2009). Schwartz, Michael K.; Luikart, Gordon; and Waples, Robin. (2006), "Genetic monitoring as a promising tool for conservation and management", Trends in Ecology and Evolution 22(1): 25-33. Schwartz, P. (1991) The Art of the Long View, Currency Doubleday. (In Huntley et al., 2010) Shea, K., Possingham, H.P., Murdoch, W.W. & Roush, R. (2002). Active adaptive management in insect pest and weed control: intervention with a plan for learning. Ecological Applications, 12, 927-936. Shearer, A.W. (2005) Approaching scenario-based studies: three perceptions about the future and considerations for landscape planning. Environ. Plan. B-Plan. Des. 32, 67-87. (in Bryan et al. , 2011) Shephard, S., Rindorf, A., Dickey-Collas, M., Hintzen, N. T., Farnsworth, K., and Reid, D. G. (2014) Assessing the state of pelagic fish communities within an ecosystem approach and the European marine strategy framework directive. ICES Journal of Marine Science, 71(7), 1572–1585. Shepherd, G. (2004). The Ecosystem Approach: Five Steps to Implementation. IUCN, Gland, Switzerland and Cambridge, UK. vi + 30 pp.) Shields, D.J., Šolar, S.V., Martin W.E. (2002) The role of values and objectives in communicating indicators of sustainability. Ecological Indicators 2, 149–160. Shin, Y-J., Bundy, A., Shannon, L. J., Simier, M., Coll, M., Fulton, E. A., Link, J. S., Jouffre, D., Ojaveer, H., Mackinson, S., Heymans, J. J., and Raid, T. 2010b. Can simple be useful and reliable? Using ecological indicators to represent and compare the states of marine ecosystems. – ICES Journal of Marine Science, 67: 717–731. Shin, Y-J., Shannon, L. J., Bundy, A., Coll, M., Aydin, K., Bez, N., Blanchard, J. L., Borges, M. F., Diallo, I., Diaz, E., Heymans, J. J., Hill, L., Johannesen, E., Jouffre, D., Kifani, S., Labrosse, P., Link, J. S., Mackinson, S., Masski, H., Mo¨llmann, C., Neira, S., Ojaveer, H., ould Mohammed Abdallahi, K., Perry, I., Thiao, D., Yemane, D., and Cury, P. M. 2010a. Using indicators for evaluating, comparing, and communicating the ecological status of exploited marine ecosystems. 2. Setting the scene. – ICES Journal of Marine Science, 67: 692–716. Sigel K., Bernd Klauer and Claudia Pahl-Wostl, 2010, Conceptualising uncertainty in environmental decisionmaking: the example of the EU Water Framework Directive, Ecological Economics 69(3): 502–510, doi:10.1016/j.ecolecon.2009.11.012. Simon, Herbert A. 1962, The Architecture of Complexity, Proceedings of The American Philosophical Society, Vol. 106, No. 6, 467-482. http://www.cs.brandeis.edu/~cs146a/handouts/papers/simon-complexity.pdf Slovic, P., Fischhoff, B., Lichtenstein, S., 1981. Perceived risk: psychological factors and social implications. Proceedings of the Royal Society of London, Series A, Mathematical and Physical Sciences 376, 17–34 1764. Smith, R. J., J. Easton, B. A. Nhancale, A. J. Armstrong, J. Culverwell, S. D. Dlamini, P. S. Goodman, L. Loffler, W. S. Matthews, A. Monadjem, C.M. Mulqueeny, P. Ngwenya, C.P. Ntumi, B. Soto and N. Leader-Williams. 2008. Designing a transfrontier conservation landscape for the Maputaland centre of endemism using biodiversity, economic and threat data. Biological Conservation 141:2127-2138 Steele, K., Carmel, Y., Cross, J. and Wilcox, C. (2009). Uses and misuses of multi-criteria decision analysis (MCDA) in environmental decision-making. Risk Analysis, 29: 26 – 33. Steinitz, C., Arias, H., Bassett, S. (2003) Alternative Futures for Changing Landscapes the Upper San Pedro River Basin in Arizona and Sonora. Island Press, Washington, DC. (in Bryan et al. , 2011) Stone, A.G. and Redmer T.A.O. (2006) The Case Study Approach to Scenario Planning, Journal of Practical Consulting, Vol. 1 Iss. 1, pp. 7-18. Sutherland, W.J., Clout, M., Côté, I.M., Daszak, P., Depledge, M.H., Fellman, L. et al. (2009). A horizon scan of global conservation issues for 2010. Trends in Ecology and Evolution, 25, 1-7. Taleb, N.N. (2007). The Black Swan: The impact of the highly improbable. Random House, New York, New York, USA. Tallis, H. and Kareiva, P. 2006. Shaping global environmental decisions using socio- ecological models. Trends in Ecology and Evolution 21(10): 562-568. Page 58 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Tang, SM. and E.J. Gustafson. 1997. Perception of scale in forest management planning: Challenges and implications, Landscape and Urban Planning 39: 1-9. Tapinos, E. 2012, Perceived Environmental Uncertainty in scenario planning, Futures 44: 338–345, doi:10.1016/j.futures.2011.11.002. Tauli-Corpuz, V., L. Enkiwe-Abayao and R. de Chavez (2010). Towards an Alternative Development Paradigm: Indigenous Peoples' Self-Determined Development, Tebtebba Foundation. Taylor, P. D., L. Fahrig, K. Henein, and G. Merriam. 1993. Connectivity is a vital element of landscape structure. Oikos 68 (3): 571-573. Thérivel, R., and Paridario, M. R. 2013. The Practice of Strategic Environmental Assessment. Routledge. Traill, L. W., C. J. A. Bradshaw, and B. W. Brook (2007) Minimum viable population size: A meta-analysis of 30 years of published estimates. Biological Conservation 139:159-166. UNDRIP, 2007. United Nations Declaration on the Rights of Indigenous People. http://www.un.org/esa/socdev/unpfii/documents/DRIPS_en.pdf (last accessed: 08 January 2015) UNECE. 1998, Convention on Access to Information, Public Participation In Decision-Making And Access To Justice In Environmental Matters, United Nations Economic Commission for Europe (UNECE), http://www.unece.org/fileadmin/DAM/env/pp/documents/cep43e.pdf UNEP (2000).UNEP Environmental Impact Assessment Training Resource Manual. Geneva: United Nations Environment Programme. Urban, D.L., O'Neil R.V., Shugart, H.H., 1987, Landscape ecology: a hierarchical perspective can help scientists understand spatial patterns. Bioscience, 37, 119-127 van der Heijden, K. (2005) Scenarios: the art of strategic conversation. John Wiley & Sons. van der Keur, P., M. Brugnach, A. Dewulf, J. C. Refsgaard, P. Zorilla, M. Poolman, N. Isendahl, G. T. Raadgever, H. J. Henriksen, J. J. Warmink, M. Lamers, and J. Mysiak, 2010. Identifying Uncertainty Guidelines for Supporting Policy Making inWater Management Illustrated for Upper Guadiana and Rhine Basins, Water Resour Manage 24:3901–3938, DOI 10.1007/s11269-010-9640-x Van der Sluijs, J. (2005) Uncertainty as a monster in the science-policy interface: four coping strategies. Water Sci. Technol. 52, 87–92. (in . Vervoort et al., 2014) Van Ittersum, M.K., Ewert, F., Heckelei, T., Wery, J., Olsson, J.A., Andersen, E., Bezlepkina, I., Brouwer, F., Donatelli, M., Flichman, G., Olsson, L., Rizzoli, A.E., van der Wal, T., Wien, J.E., Wolf, J. (2008) Integrated assessment of agricultural systems e a component-based framework for the European Union (SEAMLESS). Agric. Syst. 96, 150-165. (in Bryan et al. , 2011) Vervoort, J. M., P. K. Thornton, P. Kristjanson, W. Förch, P. J. Ericksen, K. Kok, J. S. I. Ingram, M. Herrero, A. Palazzo, A. E. S. Helfgott, A. Wilkinson, P. Havlík, D. Mason-D'Croz, and C. Jost. (2014). Challenges to scenario-guided adaptive action on food security under climate change. Global Environmental Change. Vervoort, J.M., Thornton, P.K., Kristjanson, P.,Förch, W., Ericksen, P.J., Kok, K.,Ingram, J.S.I., Herrero, M., Palazzo, A., Helfgott, A.E.S., Wilkinson, A. et al. (2014) Challenges to scenario-guided adaptive action on food security under climate change, Global Environmental Change, 28, 383–394. Vigui, V., Hallegatte, S., Rozenberg J. (2014). Downscaling long term socio-economic scenarios at city scale: A case study on Paris, Technological Forecasting & Social Change, 87, 305–324. von Neumann, J. and Morgenstern, O. (1944). Theory of Games and Economic Behavior. Princeton University Press. Walker, W., Harremoes, P., Rotmans, J., Van Der Sluijs, J., Van Asselt, M., Janssen, P., Krayer Von Krauss, M. 2003, Defining Uncertainty: A Conceptual Basis for Uncertainty Management in Model-Based Decision Support. Integrated Assessment, Vol. 4, No. 1, pp. 5–17. Available at: http://journals.sfu.ca/int_assess/index.php/iaj/article/view/122. Date accessed: 07 Jan. 2015. Walters, C. and Maguire, J.J. 1996, Lessons for stock assessment from the northern cod collapse.Reviews in Fish Biology and Fisheries 6, 125-137. Walters, C.J. (1986). Adaptive management of renewable resources. MacMillan, New York, New York, USA. Walters, C.J. (2007). Is adaptive management helping to solve fisheries problems. Ambio, 36, 304-307. Page 59 of 60
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Walters, C.J. and Hilborn R. (1976). Adaptive control of fishing systems, Journal of the Fisheries Research Board of Canada, 33(1): 145-159 Walz, A., Lardelli, C., Behrendt, H., Gret-Regamey, A., Lundstrom, C., Kytzia, S.,Bebi, P. (2007) Participatory scenario analysis for integrated regional modelling, Landsc. Urban Plan., 81, 114-131. (in Bryan et al. , 2011) Wang, Y. M., Shang, J. C., Ding, J. X. (2007). In strategic environment assessment alternative scheme formation and comparative analysis method. Jiangsu Environmental Science and Technology, 20(6), 40-43. Wardekker, J.A. and J.P. van der Sluijs, 2005. Report of the Expert Meeting Uncertainty Communication, Report NWS-E-2005-111, Report commissioned by the Netherlands Environmental Assessment Agency (RIVM/MNP), http://dspace.library.uu.nl/handle/1874/20685. Wardekker, J.A., Jeroen P. van der Sluijs, Peter H.M. Janssen, Penny Kloprogge, Arthur C. Petersen. 2008. Uncertainty communication in environmental assessments: views from the Dutch science-policy interface, Environmental Science and Policy II: 627-641, doi:10.1016/j.envsci.2008.05.005. Warmink, J.J., Janssen, J.A.E.B., Booij, M.J. and Krol, M. 2010, Identification and classifica-tion of uncertainties in the application of environmental models.Envi-onmental Modelling and Software 25(12), 1518–1527. Wei, Y.P., Davidson, B., Chen, D.L., White, R. (2009) Balancing the economic, social and environmental dimensions of agro-ecosystems: an integrated modelling approach. Agric. Ecosyst. Environ. 131, 263273. (in Bryan et al. , 2011) Wennerström, L., L. Laikre, N. Ryman, F.M. Utter, N.I. Ab Ghani, C. Andre´, J. DeFaveri, D. Johansson, L. Kautsky, J. Merila, N. Mikhailova, et al. (2013) Genetic biodiversity in the Baltic Sea: species-specific patterns challenge management, Biodivers Conserv (2013) 22:3045–3065, DOI 10.1007/s10531-013-0570-9 Whitlock, R. (2014) Relationships between adaptive and neutral genetic diversity and ecological structure and functioning: a meta-analysis, Journal of Ecology 102:857–872, doi: 10.1111/1365-2745.12240. Wildlife Conservation Society (WCS) & Bio-Era. (2007) Futures of the wild. A project of the Wildlife Conservation Society’s Futures Group. Wildlife Conservation Society and Bio-Era, New York. Wilkinson, A., Eidinow, E. (2008) Evolving practices in environmental scenarios: a new scenario typology. Environ. Res. Lett. 3, 045017. (in Vervoot et al., 2014) Williams, B.K. 2012, Reducing uncertainty about objective functions in adaptive management, Ecological Modelling 225: 61– 65, doi:10.1016/j.ecolmodel.2011.11.009. Winship, A. J., O’Farrell, M. R., & Mohr, M. S. (2013). Management strategy evaluation applied to the conservation of an endangered population subject to incidental take. Biological Conservation, 158, 155166. Wintle, B.A. & Lindenmayer, D.B. (2008). Adaptive risk management for certifiably sustainable forest management. Forest Ecology and Management, 256, 1311-1319. Wintle, B.A., Bekessy, S.A., Keith, D.A., van Wilgen, B.W., Cabeza, M., Schroder, B. et al. (2011). Ecologicaleconomic optimization of biodiversity conservation under climate change. Nature Climate Change. 1:355-359. Wintle, B.A., Bekessy, S.A., Pearce, J.L., Veneir, L.A. & Chisholm, R.A. (2005). Utility of Dynamic Landscape Metapopulation Models for Sustainable Forest Management. Conservation Biology, 19, 1930-1943. Wintle, B.A., McCarthy, M.A., Volinsky, C.T. & Kavanagh, R.P. (2003). The use of Bayesian Model Averaging to better represent the uncertainty in ecological models. Conservation Biology, 17, 1579-1590. Wintle, B.A., Runge, M. & Bekessy, S.A. (2010). Allocating monitoring effort in the face of unknown unknowns. Ecology Letters, 13, 1325–1337. Woolmer, G., Trombulak S. C., Ray J.C., Doran J. P., Anderson M. G., Baldwin R. F., Morgan A., Sanderson E. W. (2008). Rescaling the Human Footprint: A tool for conservation planning at an ecoregional scale. Landscape and Urban Planning 87 (2008) 42–53 Zacharias, M.A. And J.C. Roff (2000) A Hierarchical Ecological Approach to Conserving Marine Biodiversity, Conservation Biology, 14(5):1327–1334. Zhang, L. J., Xue, B., Sun, D. P. (2004) Alternative plan in the SEA of Hunnan New District in Shenyang. Environmental Protection Science, 30(125), 62–65. Page 60 of 60
First Order Draft
3
IPBES Deliverable 3c
DO NOT CITE
Building scenarios and models of [indirect and direct] drivers of change in biodiversity and ecosystems Coordinating Lead Authors: Michael Obersteiner, Ramón Pichs-Madruga, Britaldo Soares Filho Lead Authors: Mohamed Tawfic Ahmed; Klaus Kellner; Samba Fall; Xuefeng Cui; Peter Verburg; Philippe Cury. Contributing Authors: To be added.
Executive Summary Section 3.1 The analysis of biodiversity and ecosystems changes should start with the distinction between direct or proximate (DD) and indirect or underlying drivers (ID) of those changes. The primary direct causes typically include land use change and conversion of habitat to other land uses, pollution, unsustainable natural resources use, climate change, and invasive alien species. The underlying or indirect causes are usually described by social, economic, political, cultural, and technological processes that ultimately define proximate drivers leading to ecosystem degradation. Drivers associated with human activities have accelerated the rate of species extinction and significantly changed ecosystem properties. The definition of approaches to construct scenarios, the specification of scenario assumptions and the use of modelling approaches to describe interaction between indirect and direct drivers are key elements for assessing the impacts on biodiversity and ecosystem functions. Section 3.2 Participatory as well as modelling methods and tools are key instruments for building scenarios of drivers of change in biodiversity and ecosystems. Participatory methods and tools constitute important channels to collectively define complex problems related to the governance of particular biodiversity and ecosystem services. This social and environmental approach has been widely accepted as an innovative and sound framework for analysing the indirect and direct drivers as well as defining collective, well-grounded solutions and local development planning pathways. This approach reveals that stakeholder participation is critical for identifying drivers of change. Further, it provides a platform for views to be aired, perspectives broadened, and improved understanding of the issues. Indigenous and local knowledge provides a genuine reflection of prevailing conditions and other key inputs. Many typologies of modelling methods and tools of indirect and direct drivers as well as their interactions are possible. Those methods and tools can be categorized depending on their qualitative or quantitative nature, whether the underlying phenomenon can be represented by structural equations or driver processes are captured by data driven approaches, if the model is of deterministic or stochastic nature, and if the entire economy is modelled or a specific sector. Such broad typologies can typically be further broken down in subcategories as appropriate.
Page 1 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Modelling methods and tools to be used depend on the assessment objectives. As a general rule, the short-term prediction skill of data driven approaches is superior to mechanistic structural models. However, for long-run analysis where biophysical boundaries of production systems need to be respected and for the analysis of structural adjustments of drivers due to policy changes, mechanistic models are more suitable. Illustrative examples of models are presented in this chapter by application domain of quantifying direct drivers, such as Marine Habitat Modification, Terrestrial Habitat Modification, Land Degradation, Invasive Species, Climate Change, Pollution, Exploitation and Use of Resources. Section 3.3 Scenarios and models of indirect and direct drivers of change in biodiversity and ecosystems are valuable tools for well-informed decisions by policymakers. A policy cycle serves as a framework to facilitate effective decision making by taking into consideration a comprehensive analysis of the problem, followed by an intervention design, implementation, and finally evaluation of the policy’s success. Due to the inherent complexity of the environment-policy nexus, enactment of environmental policies may result in unforeseen externalities that run counter to the original goals or encourage counterproductive behaviour such as rebound effects. This chapter explores the interactions between several scenarios approaches (e.g., exploratory, goal oriented, ex-ante and ex-post scenarios) and various policy objectives (e.g., non-intervention, policy prescription, proactive or reactive policies). Section 3.4 The development and quantification of scenarios of indirect driver impacts on direct drivers of biodiversity and ecosystem services is multi-faceted; and in many cases, multiple models are required to address the different spatial scales and/or the different indirect drivers. For regional assessments, global scale assessment models are often required to account for the influence of distant drivers on the region, while regional models are used to add more regional specificity and detail to the simulations. No single model can capture all dynamics at a high level of detail and the coupling or integration of models has become a popular tool to integrate the different dimensions. However, the degree of coupling varies amongst studies, and the choice of integrated modelling versus a loose-coupling of models depends on the specific requirements of the assessment but also on the system studied. Basic principles for good modelling practice may include provisions for transparency, completeness, consistency, comparability and accuracy, including the possibility of technical assessment by independent reviewers. The key messages from this chapter are: 1. In order to be relevant, scenarios/models to assess drivers should be aligned with the needs within the policy phase/scale under consideration.
Page 2 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
2. A comprehensive set of measurable drivers/indicators need to be considered for scenarios and models to detect transformational change relevant to biodiversity and ecosystem function. 3. Context-specific solutions can only be found when drivers across different (spatial and temporal) scales are considered and scenarios and modelling tools are adequately integrated.
3.2 Definitions and conceptual framework Ecosystems have been subject to adaptations by exogenous and endogenous drivers ever since the beginning of life on earth. Until human activities started to change the earth surface, leading to considerable impacts on biodiversity, ecosystems and biodiversity evolved due to exogenously given drivers such as changing climatic and lithospheric conditions. It is widely believed that drivers associated with human activities have accelerated the rate of species extinction and significantly changed ecosystem properties. Scientists (Crutzen) have named this new geological epoch the Anthropocene where human activities on the Earth's atmosphere, lithosphere, and biosphere in recent centuries have become the dominant drivers of change. In addition to classification as indirect or direct, drivers can also be grouped as endogenous and/or exogenous. This designation is highly contingent on the spatial and temporal scale under consideration. Where exogenous indirect and direct drivers of ecosystem change affect the decision-maker yet are beyond their spatial and temporal influence (i.e., they influence the decision-maker but the decisionmaker cannot alter them), endogenous drivers imply that the decision-maker has some degree of autonomy over the change. For example, national level government biofuel subsidies may represent an exogenous and indirect driver of land use change at the local level (endogenous and direct). In other words, the exogenous driver (or independent variable) exerts an influence on an endogenous driver (a dependent variable) but is unaffected by changes in the endogenous driver. In the analysis of biodiversity and ecosystems’ change, direct or proximate and indirect or underlying drivers are usually distinguished. Drivers can be classified as indirect and/or direct, and are major factors that affect ecosystems and biodiversity. In many cases drivers can act as both indirect and direct catalysts of change. Drivers should not be looked at as separate, static influences but should be considered as dynamic factors, interacting with and within each other. Indicators are used to reflect the influence and weight of drivers in scenarios and models. Drivers are basic components of scenarios and models, depicting ecosystem conditions and trends as well as biodiversity. Major direct drivers typically include land use change and conversion of habitat to other land uses, pollution, unsustainable natural resources use, climate change, and invasive alien species. For example, agricultural land expansion is estimated to be the proximate or direct driver for around 80% of deforestation worldwide. The underlying or indirect causes are usually accounted for by social, economic, political, cultural, and technological processes that ultimately define proximate drivers leading to ecosystem degradation. Indirect drivers frequently strongly interact, giving rise to complex emerging properties on higher spatial and temporal scales. In many countries “soft” factors such as weak governance and institutions, lack of cross-sectoral coordination, and illegal activity are cited as key
Page 3 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
underlying drivers of ecosystem change (cit. Kissinger). If well targeted, indirect drivers can be realigned to achieve the conservation of ecosystems and biodiversity. Methodologies, such as theory of change (ref), exist to promote processes of social change by outlining causal linkages between drivers along desirable outcomes pathways. The approaches to construct scenarios and the models used for policy and decision making should be relevant, but will depend on the type of ecosystem, scale, and geographical position. Different types of approaches and models are discussed that can be used at different scales and to describe certain changes in biodiversity and ecosystems, as well as their linkages. The impact of drivers for biodiversity and ecosystems as well as their influence on human well-being will be discussed in Chapters 2 and Chapter 4.
3.2.1 Indirect Drivers Within complex ecological systems the role of indirect drivers is an integral aspect of scenario development and subsequent analysis. Socioeconomic factors such as demographic change, agricultural and industrial output, and economic growth among others directly influence consumption patterns with subsequent environmental implications (e.g., Seto & Kaufmann 2003). The impact of demographic change both in terms of overall population growth as well as distribution and composition has long been recognized as a key indirect driver of ecosystem change. Here the cohort-component method and alternative variations of this method (e.g., multistate cohort-component projections taking into consideration education) project population according to age and sex under assumptions regarding fertility, mortality, and migration(e.g., O’Neill et al. 2001). In addition to interacting with socioeconomic and demographic drivers, technological innovation can lead to the adoption of cleaner and more sustainable energy production as well as indirectly contribute to environmental degradation through electronic waste and increased demand for the raw materials used in new technologies. In addition to affecting the aforementioned indirect drivers, governmental policies (or the lack thereof) can impact the environment in myriad ways. Ill-informed and weak governance frequently leads to mismanagement of the commons as well as the adoption of environmentally unfriendly policies. The effects of institutional and governmental policies on the environment is clear in the contrast observed between the Dominican Republic and Haiti where despite geographical similarities, a long history of weak environmental governance in Haiti has led to ecosystem degradation and increased vulnerability to natural disasters (UNEP 2013). Here natural disasters such as prolonged droughts can function as both an indirect and direct driver of ecosystem change by effecting ecological change while simultaneously erecting institutional and governmental barriers to mitigation and adaptation efforts. Other institutional drivers include the role of conservation organizations and other organizations dedicated to environmental management as well as other forms of institutional support. Culture in the form of the values, norms, and beliefs of a group of people can act as an indirect driver of ecosystem change by affecting environmentally relevant attitudes and behaviors. A prominent example is the role of globalization and removal of international trade barriers, resulting in different consumption and
Page 4 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
production patterns as well as the introduction of non-endemic, invasive species.
3.2.2 Direct Drivers Development trajectoria of direct drivers are either produced as outputs of structural models combining indirect driver information as described above or directly by non-structural models using historical data and data driven algorithmic approaches. (to be elaborated on)
Figure 3.1: Key steps for building scenarios and models of drivers of change in biodiversity and ecosystems
3.2.3 Approaches to construct scenarios and scenario assumptions As thought experiments, scenarios allow researchers to learn from a constructed future while avoiding unnecessary experimentation. Scenario construction is a necessary endeavour when attempting to construct possible futures in context of uncertainties, particularly when ecological outcomes are highly contingent on indirect drivers such as economic growth and demography (Carpenter 2002). Thus, scenarios or “variants” as well as probabilistic projections are employed to account for uncertainty within the models. In these cases, rather than attempting to predict from a specific set of ecological variables onto a specific future, it is preferable to employ a variety of scenarios based on scientific knowledge of a range of potential alternative futures (Peterson et al. 2003). Further, goal-seeking scenario construction is a valuable tool for examining the viability of alternative pathways to a desired outcome. Page 5 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Although all modelling implicitly involves some degree of expert opinion, formal expert-based scenario modelling involves identifying and eliciting information (and uncertainty) from multiple experts, either individually or in a group (Krueger et al. 2012). While identification of and disagreement among experts can pose obstacles to this method of scenario construction (as well as the cost and time consumption involved in eliciting information), scientists are increasingly aware of the advantages of the deliberate formal use of expert opinion to inform ecological models. Participatory approaches to scenario development consist of involving a larger group of interested stakeholders through workshops or other formal meetings to share ideas and ultimately develop scenarios based on their collective knowledge. This approach has the benefit of providing local expertise on local scenarios as well as cultivating a culture of participation and better informing local stakeholders (Patel et al. 2007; Palomo et al. 2011). Nonetheless, barriers include insufficient understanding of the issue and considerable differences in opinion among participants as well as difficulty in translating qualitative data into quantitative inputs (Walz et al. 2007).
3.2.5 State of the art Global scale long-run assessments are typically framed in consistency with existing scenario storylines such as the IPCC SRES scenarios (Nakicenovic & Swart 2000). The IPCC, MEA, GBO, GEO and the global desert outlook have used these storylines or close derivatives of these to generate indirect driver scenarios for their sector specific outlooks. Regional assessments of the MEA and GEO as well as the National GEO such as those carried out by the UK, China and Brazil have used globally consistent regional varients of existing storylines. Downscaled gridded scenarios of socio-economic drivers of SRES (Grübler et al. 2007) have been used as indirect drivers of forest cover change (Kindermann et al. 2008). Climate change scenarios are typically provided on the same grid resolution and are used as direct drivers of ecosystem change (e.g., Seidl et al. 2014) Local and more regional specific scenarios of indirect and direct drivers are typically constructed bottom-up and may significantly deviate from the globally established storylines. More recently associations or even directing mapping of such bottom-up scenarios into global storylines were performed allowing for increased comparability across regional case studies (e.g., CCAFS 2015)
3.3 Methods and tools 3.3.1 Participatory methods and tools for developing plausible (indirect) driver scenarios The participatory approach constitutes an interesting and invaluable channel to collectively define complex problems related to the governance of particular biodiversity and ecosystem services. In the terrestrial and marine domain, a number of participatory methods involving stakeholders have been developed which help to prioritize the issues and drivers that need to be considered in scenario building. (Ph. C: We should develop a bit more the overall and overall types of scenarios and their objectives here?) In that sense, through the “agent-based participatory simulation” approach, Briot et al. (2007) have
Page 6 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
studied and illustrated the complex issue of interest conflicts often arising between users of protected areas in the Brazilian Atlantic Forest and specifically, the Ticuja Park. In practice, this approach consists of a social dialogue platform in which multi-agent simulation (MAS) is combined with role-playing games (RPGs). By using properly the MAS/RPG approach, local stakeholders surrounding the Ticuja Park were able to define key indirect and direct drivers of the depletion of their natural resources and were skilled enough to draw collective computer-based scenarios for better management and valuation of the Ticuja Park. In the same vein, Standa-Gunda et al. (2003), have tested and run a participatory approach to investigate a better way of planning the Mafungautsi forest resource management in Zimbabwe, especially the broom grass harvesting which is one of their main economic activities. In a meaningful participatory settings with the action researchers, local communities performed a social learning process in which indirect and direct drivers (overharvesting for example) of the degradation of the broom grass and their interrelationships were assessed and drawn with a network software platform (Simile). By doing so, invaluable thoughts and insights from both local knowledge and results of scenarios have been derived collectively and in turn, will be key inputs for the next round of governance planning for the Mafungautsi forest resources and specifically the broom grass management. In the coastal area of Kenya, researchers and different stakeholders (fishery communities, policy makers, private hotel owners, etc.) have joined efforts to produce a collective diagnostic of drivers that negatively affect their well-being as well as to define common and well-grounded management schemes for fishery activities. First, the mental approach (network system) combined with a constructive and active stakeholder dialogue in which different participatory tools (opinions-voting, plenary, carousel) are used to identify and draw the interrelationship between indirect and direct drivers and in turn, the changes (positive and negative aspects) of their well-being and trade-offs of each fishery governance scheme are scrutinized. Stakeholder participation is critical when identifying drivers of change and their importance for an ecosystem approach to fisheries (EAF) aimed at reconciling exploitation and conservation of marine biodiversity. Based on the FAO code of conduct for responsible fisheries (Attwood et al. 2005) and the Australian ecological sustainable framework (Fletcher et al. 2002), a series of locally adapted Ecological Risk Assessments (ERA) participatory approaches have been developed in the Benguela Current region (i.e., South Africa, Namibia and Angola) to implement this approach (Augustyn et al. 2014). This provides a transparent and structured process among stakeholders which helps to prioritise the issues and drivers that need to be considered. (Nel et al. 2007). Briefly this methodology relies on a three step process (Nel et al. 2007): 1 Identification of concerns or issues
2 Prioritisation of these concerns or issues
3 Development of Performance Reports which describe the appropriate scientific and management responses necessary to address the issue
Page 7 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
The methodology utilizes generic component trees to help participants to identify the main issues that the fishery faces (Figure 2.2). The process starts off by breaking the fishery down into eight major components in three main categories; ‘Ecological Wellbeing’, ‘Human Wellbeing’ and ‘Ability to Achieve’.
Figure 3.2: Diagrammatic representation of the eight major components of the ERA process (Nel et al. 2007)
Identified issues are then prioritized by scoring the consequence of a given risk actually occurring independent from the likelihood of it occurring. The risk value therefore provides a means of prioritizing the issues. Low risk issues require no management action whereas high risk issues require drivers to be taken into consideration. At this step it is necessary, as far as possible, to gain consensus on the consequences and likelihoods. While this can be a contentious stage in these workshops, there was a high level of agreement across the issues that had been identified (Augustyn et al. 2014). Full Performance Reports can then be developed for all issues of sufficient priority according to the template in Table 3.1. Briefly, these required the setting of an operational objective, the identification of indicators, targets, and milestones. These allow for regular progress to measure against agreed targets.
Page 8 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Table 3.1: Diagrammatic representation of the nine steps to developing the Performance Reports (Nel et al. 2007) The work of a wide range of stakeholders including scientific fisheries, conservation institutions, the fishing industry, and NGOs has been essential to identifying indirect and direct drivers of change in biodiversity and ecosystems. Stakeholder participation thus appears critical to the successful implementation of an EAF as complexity creates confusion and frustration, reducing the chances of success when identifying drivers and formulating an implementation process. A structured approach like the ERA provides a platform for views to be aired, broadens perspectives, and improves understanding of the issues. Indigenous populations are a major factor in identifying drivers. Their local knowledge provides a genuine reflection of prevailing conditions, how drivers would evolve, and how to develop a succession of other drivers. Meeting with local tribes of Bedouin communities in Sinai Peninsula (one of the driest ecosystems of Egypt) has revealed that Bedouins developed the ability to predict the pathways of flash floods and the area covered by these floods, allowing them to prepare the land for cropping early enough before the onset of the floods. They resisted any change in the land use that might affect the flow of the flash floods. Changes in land use pattern, caused by gravel and marble extracting contractors, would change the course of flash floods, disperse water flow, and cause a great loss of water resulting in serious problems for Bedouin life and the biodiversity of their environment. The disappearance of many plants they use to treat various maladies or to augment their nutritional needs was attributed to gravel contractors cutting in desert plains and heights to make narrow corridors to allow access to new high altitude resources for mining. This in turn was followed by vegetation
Page 9 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
clearance, habitat fragmentation, and loss of flash flood flow. Meeting with Bedouin (Maglas) during the course of a Subglobal Assessment (SGA), the Millennium Ecosystem Assessment emphasized the role of communicating with local tribe, and refers to their local knowledge as a fundamental source for identifying drivers. The SGA revealed a massive loss of plant diversity in the study area when comparing the current profile with the that of the same area in 1960 (Ahmed 2010). These consultation meetings also revealed how one driver may turn impact a series of other drivers. In this respect, land use change would lead to habitat fragmentation, loss of vegetation, and water scarcity as a sequence of drivers reverberates throughout the ecosystem, culminating in a serious loss of biodiversity. Nevertheless, information collected through local knowledge from indigenous populations should be verified. Further consultation and validation with a number of elderly Bedouin of is one way of authenticating and scrutinizing information.
3.3.2 Modelling methods and tools to quantify the interactions between IDs and DDs 3.3.2.1 Overview of available methods to model the interactions between indirect and direct drivers Many typologies of modelling tools of indirect and direct drivers and their interaction are possible. Modelling tools can for example be categorized depending on their qualitative or quantitative nature, whether the underlying phenomenon can be represented by structural equations or driver processes are captured by data driven approaches be used, if the model is of deterministic or stochastic nature, and if the entire economy is modelled or a specific sector. Such broad typologies can typically be further broken down in sub-categories. For example among structural models we can distinguish again between simulation models such as systems dynamics models and optimization models. Among the latter classical economic models typically maximize a welfare function or minimize production system costs. If such models cover the entire economy they are referred to as general equilibrium models or alternatively partial equilibrium models cover only a specific sector in greater detail. Such economic models can be constructed for comparative static analysis to analyze the introduction of new drivers such as policy shocks or for dynamic assessments to analyse solution pathways. Traditionally structural economic models emulated indirect and direct drivers in deterministic settings and latest developments of these models allow for the assessment of stochastic phenomena such as the impact of volatility of agricultural production on land use changes (Leclère et al. forthcoming). Another class of optimization models are decision support tools such as multiple criteria optimization models which allow for the assessment of management strategies following multiple objective goal functions and their trade-offs. Agent based models allow for a blending of optimization and simulation modelling. They enable the study of the emergent properties on higher spatial and temporal scales of heterogenous drivers impacting optimizing behaviour on the level of individual agents. In some cases
Page 10 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
agent based models operate with heuristic rule sets and then should be categorized as pure simulation models. Short term forecasts of drivers, most frequently economic drivers, are generated by nonstructural models implying that the modelling tool finds patterns in the data itself and projects these into the future. Tools for the extrapolation of current trends include statistical and econometric methods and data mining tools such as artificial neural networks, rough and fuzzy set, and network theory approaches. These tools also allow for projections of an ensemble of variables which interact with each other such as Vector Autoregressive Models. These models data driven typically will not allow for a mechanistic understanding of how and why drivers interact in their particular form. As a general rule the short-term prediction skill of data driven approaches is superior to mechanistic structural models. However, for long-run analysis where biophysical boundaries of production systems need to be respected and for the analysis of structural adjustments of drivers due to policy changes, mechanistic models are more suitable. In the following we will present modelling tools which were applied to specific application domains featuring direct drivers. Table 2.2 provides an overview of the coverage of indirect drivers ranging from demographic to natural disasters for the illustrative model applications.
3.3.2.3 Illustrative examples of models by application domain of quantifying direct drivers 3.3.2.3.1 Marine Habitat Modification A joint participatory mental modelling (network approach) and Toy model(socio-ecological approach)modelling processes were performed to screen out the multiple trade-offs facts between a range of management schemes of the costal ecosystem services in Kenya and to assess its implications regarding the well-being of divergent primary and secondary stakeholders. First, under the guidance and assistance of researcher facilitators, primary and secondary stakeholders (local community) were gathered to define and draw a well-being matrix which consists of identifying direct and indirect factors (drivers) and their interactions that affect their livelihoods and more broadly, the governance of their coastal ecosystem services. Throughout different methods of participatory exercises (opinions-voting, plenary, caroussel), this process called mental modeling (network approach) helps to assess the extent to which the fishery is affected by indirect and direct drivers. The importance of interrelationships are captured by the thickness of their link (Figure 2.4).To have an in-depth understanding of the overall picture of impacts and trade-offs for each fishery governance schemes, an ecological model named Ecopath is used to investigate the fisheries dynamics. The combination of the mental and Toy models permit identification of the factors that should be reinforced in order to improve fishery governance schemes (Figure 3.5).
Page 11 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 3.4: Snapshot of toy model outputs To face the arising complex issues (overfishing, unsuitable governmental schemes, etc.), the Participatory Modelling of Wellbeing Trade-offs in Coastal Kenya (P-mowtick) were set and run with a multiple range of stakeholders to efficiently account for the depletion of the value of the coral reef fishery ecosystem services. In fact, overfishing, unsustainable, and weak governmental management schemes that negatively affect the well-being of local fishery communities and existing trade-offs among stakeholders were not really understood. Therefore, there was a need to investigate such interrelationships through participatory and easily comprehensible networking and ecological modelling approaches. It has been shown clearly that indirect drivers related to the institutional setting (governance schemes, the effectiveness of policy implemented) and population growth have the greatest impact on the tradeoffs in the fishery ecosystem services. On the other hand, the Toy model has revealed that fishing variables (method of fishing, frequency, volume of stocks) influenced directly the well-being and value of the fisheries’ ecosystem services.
Page 12 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 3.5: Identification of indirect and direct drivers In this case, all relevant stakeholders (communities of Mombassa, researchers, policy-makers, etc.) went through a process starting from the design of a mental model up to a socio-ecological model (Toy model). First, the mental modelling approach provides the basic information regarding the indirect and direct factors, the degree of their interconnections, and the importance of their effects on the fishery ecosystem services. The second step will consist of investigating the most important links between factors (indirect to direct ones) and then drawing upon this insight to reveal their impacts on the wellbeing of local communities and fisheries’ ecosystem service
Page 13 of 41
First Order Draft
IPBES Deliverable 3c
Identifying Priority Areas for Conservation of Species at Risk in the Mediterranean Sea Conservation priorities need to take the feasibility of protection measures into account in a dynamical and multi-usage way. The Mediterranean Large Marine Ecosystem (LME) was used as a case study to identify the overlapping areas of low threats and high diversity of vertebrate species at risk (Coll et al. 2014). A large body of knowledge exists on individual anthropogenic threats that impact marine biodiversity in the Mediterranean Sea, although we know little on how these threats accumulate and interact to impact marine species and ecosystems (Coll et al. 2012). In this context, the identification of the main areas where the interaction between marine biodiversity and threats is more pronounced can be performed. Using cumulative threats models, we identified “Priority Areas for Conservation of Species at risk” (Priority Areas), where IUCN diversities are high and threats are low. The Mediterranean LME is globally the first with a complete regional IUCN Red List assessment of the native marine fish. Distributions of marine mammals, marine turtles, and seabirds at risk were considered to calculate the spatial distributions of species at risk (IUCN densities). In times of economic pressures it is essential to identify the “low hanging fruits” for conservation: areas where human impacts are lower and biological diversity is still high, and thus conservation is more feasible. IUCN densities and Priority Areas were not highly correlated spatially among taxa. Continental shelves and deep-sea slopes of the Alboran Sea, Western Mediterranean and Tunisian Plateau/Gulf of Sidra are identified as relevant for fish species at risk (Figure 2.6a). The Eastern side of the Western Mediterranean and the Adriatic Sea are identified as most relevant for endemic fish, and shelf and open sea areas distributed through the LME are most important for marine mammals and turtles at risk (Figure 2.6b), while specific locations of the Western Mediterranean Sea and the Aegean and Levantine Seas are highlighted for seabirds (Figure 2.6c). Large surfaces of Priority Areas fell outside current or proposed frameworks to be prioritized for conservation. Therefore, Priority Areas may be suitable candidates to contribute to the 10% of protection target for the Mediterranean Sea by 2020.
DO NOT CITE
Figure 3.6a-c: Species biodiversity in the Mediterranean Sea
Box 3.1: Identifying priority areas for conservation of species at risk in the Mediterranean Sea
Page 14 of 41
First Order Draft
5
10
15
20
25
30
35
IPBES Deliverable 3c
DO NOT CITE
3.3.2.3.2 Terrestrial Habitat Modification Habitat modification is seen as a prime driver of biodiversity loss and changes in the composition of the bundle of ecosystem services provided at a certain location. Habitat modification is mostly a result of land use change, either induced by human action or a result of changes in the physical determinants of the habitat, e.g. due to changes in hydrology or climate. In most cases the modification of habitat due to human inference is much faster and more pronounced than changes due to climate change (Lehsten et al. forthcoming). However, in specific environments, such as the arctic tundra region, climate change can have major implications for the habitat as well. Land use change is the major human influence on habitats and can include the conversion of land cover (e.g., deforestation), changes in the management of the (agro-)ecosystem (e.g., through intensification of agricultural management or forest harvesting) or changes in the spatial configuration of the landscape (e.g., fragmentation of habitats). At the regional scale, a variety of different models have emerged in the past decades to simulate changes in land use driven by demographic change, policies, and changing demands for land-based commodities or urban use. Model structure and characteristics are often specific to the scale of application, the research questions and the dominant processes involved. For small areas and when (diversity) in land use decision-making is important to be represented explicitly, agent-based models have become popular tools (Matthews et al. 2007; Brown et al. 2014). In such models the changing landscape pattern emerges from the decisions of individual land owners and managers that respond to (often exogenously defined) indirect drivers. At larger spatial and temporal scales, a simpler conceptualization of decision-making is often applied and land change is simulated based on the suitability of locations for a specific land use, with the regional level demands for the different land uses and spatial constraints resulting from regulations and land use planning (van Delden et al. 2011). In such models pixels are the units of simulation and often the state of neighbouring pixels is accounted for to represent neighbourhood effects and processes such as centripetal forces and economies of scale in urban development. Many global scale land use models use macro-economic representations of commodity markets and trade simulation in general or partial equilibrium models to simulate land change between different world regions. In many cases land use decision is represented by simulating the land use choice of a representative farm at the world region level (van Meijl et al. 2006) or at the level of coarse spatial units (Schmitz et al. 2012). Simple land allocation approaches based on land suitability are used to generate spatial patterns or more complex routines that account for competition between alternative land uses (Asselen & Verburg 2013). Independent of the scale, most land use models simulate mainly the major conversions of land cover (urbanization, deforestation etc.) and ignore the more subtle modifications of habitat conditions due to changes in land management and the modifications of the spatial configuration of landscapes (Kuemmerle et al. 2013), either due to lack of fine-resolution data on landscape elements and linear features or due to a simplified representation of landscapes by either dominant or fractional land cover (Verburg et al. 2013a)
40 3.3.2.3.3 Land Degradation Land degradation has a negative impact on the soil, water, and biological parameters of the natural
Page 15 of 41
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
resources. It contributes to soil erosion by wind and water; loss in vegetation cover and density; loss in biodiversity; changes in vegetation structure and composition; and a thickening or encroachment of indigenous, alien, and exotic plant species in the habitat. Land degradation is mainly driven by climatic factors, which include drought and fluctuations of temperature and rainfall, as well as management strategies implemented on the land, depending on the type of environment and biome. Although this is a global phenomenon, it occurs mainly in the drylands of the world where population pressures increase and more people have to depend on the land for their daily livelihoods. Land degradation is therefore driven by bio-physical and socio-economic factors for which long-term datasets under different scenarios are needed. A good understanding of these different drivers based on historical changes over long periods of time (decades and even centuries if possible) is therefore necessary since most of these factors interact with each other. Most models used to describe land degradation are rangeland dynamic and management models based on databases and expert systems culminating in a Decision Support System (DSS) for the implementation of better rangeland management strategies. ‘Decision support system’ is a generic term referring to a variety of computer systems designed to provide decision support (Joubert et al. 2014). Besides the recognition of the value of expert and other decision support systems in the business world (Guimaraes et al. 1992), they are also considered to be potentially useful when applied to natural resource management problems (Davis et al. 1989). Various expert systems and DSS have been developed depending on their aim and how they can be used in the management of rangelands. [BOX – to describe some of these models (or expert systems used as DSS)] 3.3.2.3.4 Invasive Species Invasive species may be indigenous and/or exotic/alien and occur mostly in the terrestrial and aquatic ecosystems. Invasive species compete with the local and indigenous species for natural resources and disturb the biodiversity. The water hyacinth (Eichorniacrassipes) is one of the most notorious invasive species invading the Nile river and is found all along riparian countries. In Egypt water hycaninth covers an area of around 487 km2 in the Nile river irrigation and drainage canals, while covering an area of about 151 km2 in Egypt’s lake. In addition to the loss of a huge volume of water, suppression of the growth of native plants, and negative consequences for microbes, the water hyacinth prevents the growth and abundance of phytoplankton under large mats, ultimately affecting fisheries (Villamagna & Murphy 2010; Gichuki et al. 2012).
Page 16 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Box 3.2: Water hyacinth (Eichornia crassipes) in the Mediterranean Sea
5
10
15
THE MEDITERRANEAN: A BIODIVERSITY HOTSPOT UNDER THREAT The IUCN Red List of Threatened Species (Cuttelod et al. 2009) The type and extent of Weeds and Invasive species (WIPs) will depend on the drivers which mainly include the type of habitat, soil, climatic conditions and the degree of disturbance. Most WIPs do not have natural enemies that occur in the environment they invade and have to be removed by chemical, manual, and mechanical methods. In the arid- and semi-arid savanna and grassland biomes of Southern Africa, invasive species occur in areas that are degraded, mostly in rangelands that have been disturbed by overgrazing or mismanagement, further negatively impacting the grazing capacity of the area. Indigenous woody species, such as Senegaliamellifera(black thorn) and Dichrostachyscinerea(sickle bush) thicken in such degraded rangelands, while the density of the woody alien species, such members of the Prosopissp. (mesquite) increase, competing for moisture with the local species, especially in the lower lying riverine areas and valleys. Although invasive species mostly have a negative effect on the environment and the biodiversity, they can also contribute positively to the ecosystem services on a socio-economic level. WIPs can be used to provide resources for fire, fuel and construction but also create jobs for poverty stricken communities when used in invasive control programs, such as the Working for Water (WfW) program in South Africa.
20
25
A number of related models have been developed and used in depicting invasive species spread, distribution in new areas, and also for quantifying their impacts on environment. A model that could predict the rate of invasive species spread and can be used to optimize the spatial arrangement and frequency of sampling and eradication treatment would greatly assist in the design of effective control measures. Climex, first published in 1980s, is one of the early used models in invasive species. The primary output is a mapped prediction of the favourability of a set of locations for a given species and the model also Page 17 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
produces a suite of additional information to allow for further understanding of how the species responds to climate variations.
5
10
15
20
25
30
35
40
Spatially explicit models (Modular Dispersal in GIS, MDiG) were designed as an open-source modular framework for dispersal simulation integrated within a GIS. The model modules were designed to model an approximation of local diffusion, long distance dispersal, growth, and chance population mortality based on the underlying suitability of a region for establishment of a viable population (Pitt 2008). Dynamic simulation models have been used to map the impact of invasive species. Martelloni et al., (2012) have employed a dynamical population model for the freshwater crayfish P. clarkii to evaluate their impact in yet uninfested sites since P. clarkia may cause a variety of environmental damage. P.clarkii is native to Central and South America and was imported to Europe in 1972. Cook et al, (2007) produced a stochastic bioeconomic model that enables the economic impact of an invasive pest to be estimated before its arrival, based on relatively poorly specified ecological and economic parameters. The model was developed using a hypothetical invasion of the varroa bee mite (Varroa destructor) into Australia and the negative flow-on effects that it would have on pollination by reducing honey bee populations, giving rise to a loss of pollination services. 3.3.2.3.5 Climate The Global Biosphere Management Model (GLOBIOM) developed by the International Institute of Applied Systems Analysis (IIASA) is used to illuminate the complex interplay of agricultural, bioenergy, and forestry production sectors on land use change. GLOBIOM is a partial equilibrium economic model focused on specific economic sectors (accounting for the 18 most important crops) and encompassing 30 world regions in variable degrees of resolution and disaggregation. The model is supplied by a comprehensive geo-spatial database(Skalský et al. 2008) that updates changes in potential land use according to actual or simulated land use in the previous period, thus providing a recursive dynamic depiction of change by 10 year intervals through 2050. A recent study by Leclère et al. (forthcoming)provides a global scenario analysis covering nine different climate scenarios, 17 crops, and four crop management systems, as well as the interactions between crop production, consumption, prices, and trade. It specifically examines adaptations being investmentintensive and not easily reversible, such as building new water management infrastructure for irrigation, or increases and decreases to the production capacity of a region. Such adaptations are referred to as being ‘transformational’ and need to be anticipated, but their implementation is particularly plagued by uncertainty. In accordance with earlier results, the study finds that the impacts on crop yields of changes in climate – such as increased temperature and changing precipitation levels — and CO2 atmospheric concentration could lead to anywhere between an 18% decline in global caloric production from cropland, to as much as a 3% increase by 2050. This biophysical impact varies widely across regions, crops, and management systems, thereby creating opportunities for adaptation at the same time.
Page 18 of 41
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
Indirect GLOBIOM drivers are an exogenously entered and frequently updated GPD and population growth which together with food consumption per capita (FAO-based) allow for the simulation of supply and demand, commodity markets, and international trade. Further, GLOBIOM also assumes technological progress in crop production resulting in a yearly yield increase of 0.5% and places restrictions on the speed of land availability and conversion according to biophysical constraints as well as the pragmatic limitations of policy initiatives. Direct drivers are model outputs includingdynamic maps of land use change, GHG emissions, water use, biomass extraction, erosion, nutrient balances. 3.3.2.3.6 Pollution Pollution is probably the most influential stressor affecting biodiversity. The early reports of the effect of the organochlorine insecticides DDT, along with its analogue DDD, the organochlorine insecticides, on the western grebe (Garrett 1977), is one of the most documented episodes portraying the biodiversity – pollution nexus. The disappearance of the Scandinavian Peregrine, (Newton 1988) is another manifestation of the impact of DDT, the organic pollutants on birds, and biodiversity in general. Incidents of massive killing of marine mammals caused by contamination with the polychlorinated biphenyl PCBs and other persistent organic pollutant (POPs) that belong to the same organochlorine family were frequently reported (Kannan et al. 2000; Shaw et al. 2005). POPs are a group of chemicals that include some pesticides, some industrial chemicals, dioxins, and furans. The use of POPs was banned according to Stockholm Convention on Persistent Organic Pollutant which came into force in 2004 (Ahmed 2006). The tendency of POPs to dissolve and bio accumulate in fat tissues, then bioamplify (biomagnify) through food chains have enabled them to build up in tissues, reaching very high concentrations in organisms in top of the food chain, causing serious impacts, and possible massive death (Figure 3.7).
25
Figure 3.7: Explanatory diagram of the bioamplification of a substance
30
The end of DDT use in early 1970s in many countries has already contributed to the recovery of many of the impacted populations, with special reference to the Scandinavian Peregrines. However, some of these persistent organic pollutants are still used, with their impact extended in the environment. Recently, various reports have emerged to document the deleterious effect of endocrine disturbing chemicals, (EDCs), a group of chemicals that include pesticides, industrial chemicals, metals, personal
Page 19 of 41
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
care products, and others on endocrine systems (Bergman et al. 2013). The effect of EDCs on sex ratio was thoroughly observed in fish and alligators, indicating their ability for sex disruption, with serious impact on biodiversity. (Jobling et al. 1998; Guillette et al. 2000; WWF 2011). One of the most challenging risks posed by EDCs is that the toxic impact they exert at lower concentrations is higher than that at higher concentration levels (non-monotonic effect). Hence indicating the possible drastic impact of these chemicals that could be caused by unnoticeably small concentration levels (Vandenberg et al. 2012). Other potential pollutants that impact biodiversity would include heavy metals, (Mulder and Breure (2006), nutrients, Raúl Ochoa-Hueso et al. (2011) and others. Models have been used to depict changes in ecosystems; however, with the complexity of the biological system, there is little consensus on the basic equations for describing physical systems (James 2002). The European commission has initiated a multidisciplinary project to model the impact of key environmental pollution on freshwater and marine ecosystems and biodiversity, MODELKEY (Brack 2005). Other modelling tools were developed to investigate the potential environmental impact, and inter-region transport of POPs (Prevedouros et al. 2004). One of these models is a regionally segmented multimedia fate model covering the European continent transport. The model examines the environmental fate and behavior of a wide range of chemicals, and investigates a number of emission scenarios and source reduction strategies. As another example, Aquatox is one of the widely used aquatic ecosystem models. It models chemical fate and effects as a prelude to evaluate past and present, direct and indirect impacts of stressors of aquatic ecosystem. Aquatox can simulate flasks and tanks, ponds and pond enclosures, successive stream reaches, lakes and reservoirs and estuaries (Park et al. 2008). The model is frequently used in mapping bioaccumulation of pollutants in plants, and fish, beside shorebirds that feed on aquatic organisms. However, like most water quality models, Aquatox predicts only concentrations of pollutants in water but cannot project the effects of said pollutants.
25 3.3.2.3.7 Exploitation Following a participatory approach, the research project named “SimParc” has been designed and conducted to address the complex issue of interest conflicts among users of protected areas in the Brazilian Atlantic Forest within the Rio de Janeiro’s city. 30
35
Following the “companion modelling” approach, individual behaviour and points of views of multiple stakeholders are assessed by the means of the multi-agent based simulations (MAS) and the role-playing games (RPG) which allow screening natural resource management issues. In fact, the MAS/RPG simulation approach provide an interesting networking platform in which agents (researchers, practitioners, communities, policymakers) expressed their thoughts to solving particular problems and interacted extensively to draft a range of scenarios reflecting a collective and consensual decisionmaking process (Figure 3.8).
Page 20 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 3.8: Process of the “agent-based participatory simulation” approach
5
10
15
Ticuja National Park, one of the protected areas governed under the Brazilian legislation, has been designed in such a way that a buffer zone is delineated to allow surrounding communities to benefit from ecosystem services namely the forest resources and eco-tourism. However, the presence of multiple actors (communities, researchers, officials, etc.) and different use of ecosystem services constitute significant pressure on the whole Ticuja National Park and its buffer zone. Considering the weak effectiveness of law enforcement, Brazilian government and communities are exploring to find out better management schemes of such protected areas. Setting suitable and cost-effective paths to better manage the protected areas is one the main objectives of Brazilian government and local communities which are relying mostly on use and exploitation of their surrounding ecosystem services. In this regard, the application of the “agent-based participatory simulation” approach has fostered a common understanding of institutional and economic drivers (ineffective law enforcement, illegal economic exploitation of resources, etc.) that depleted considerably the Tijuca Park and its buffering zone. In fact, in an effective participatory manner, relevant stakeholders are able to select the main indirect drivers, draw numerous computer-based scenarios of collective governance for a better conservation of biodiversity and ecosystem services.
20
25
By partnering with a local NGO (IBASE), the“agent-based participatory simulation” approach has allowed to establish a comprehensive and inclusive community of practice to define suitable and collective solutions for the management of the Tijuca National Park and its surrounding communities. Following the process (Figure 2.8), this approach starts in a participatory way by identifying the set of indirect drivers (institutional and economical conflict of interests between actors). The next step consists of drawing up with stakeholders within a Role-player games and agent-based simulations to assess
Page 21 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
possible interactions and impacts of their dynamics in order to draw up a range of scenarios of conservation and governance schemes of the Tijuca National Park.
5
The combined MAS/RPG approach appears to be an effective means of setting sustainable and inclusive management schemes of protected areas which are under pressure. In fact, its whole mythology stimulates a participatory consultation process which fosters a sound collective effort to identify relevant indirect and direct drivers of the transformational process and to draw common and agreed scenarios of potential conservation and restoration pathways of biodiversity and ecosystem services.
10
In a broader sense, this participatory simulation method presents advantages namely the effective dialogue and construction of collective decision-making facts among stakeholders, the benefit of mainstreaming sound and agreed scenarios into the planning and governance process of common resources (forest, water, etc.)
15
3.3.2.3.8 Use of resources Within a fruitful partnership with the CIFOR, local communities, researchers and officials in Zimbabwe have tested and run the participatory modelling approach to investigate a better way of planning the Mafungautsi forest resource management, especially the broom grass harvesting which is one of their main economic activities.
20
25
The broom grass model consists of implementing a range of complementary steps. First, the modelling process is explain through a plenary session where stakeholders expressed their thoughts and expectations in using this social learning and were sensitized on the benefit of such exercises. After identifying the indirect and direct drivers of the depletion of the broom grass resource, facillators begin a constructive and inclusive dialogue among stakeholders to draw on all potential and possible interrelationships that will build a preliminary diagram (Figure 3.9).
Figure 3.9: Preliminary model draw by stakeholders
Page 22 of 41
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
To better represent the diagram, a computerized based networking software entitled Simile is used to reflect the nature and means of relationships (Figure 3.10). At this stage, local knowledge and extensive discussions among stakeholders help to swift from the incomplete model (“red model”) towards a ‘black model’ which is more realistic and consensual. From this standpoint, the social learning pillar is truly tested and enhanced in a way that valuable insights were drafted to better manage the broom grass harvesting activity.
Figure 3.10: Model drawn with the Simile software
10
Over-harvesting and a set of unsustainable uses of the broom grass (illegal cutting, digging method, etc.) represent some of the underlying drivers which are depleting considerably the ecosystem services and more broadly, the well-being of the Mafungautsi forest communities. To this regard, the Participatory modelling is used to define suitable conservation governance to reverse the conflict issues and the ineffective governmental legislation schemes (command-and-control method).
15
20
25
Through the participatory modelling approach, local communities were able to generate substantial recommendations which will support the following management planning process of the Mafungautsi forest resources. By drawing the interrelationship among identified issues, local communities have learnt that better management schemes of the broom grass harvesting require a whole range of consultation process and a collection research action and agreed planning procedure. The whole participatory modelling advantage stems from and relies on the success of the social learning process. In fact, the facilitation procedure performed by skillful action researchers from CIFOR has influenced remarkably the identification of indirect drivers (ineffective command-and-control governmental method, increase of population and sales of broom grass, etc.). Building upon the local knowledge and the computed-based modelling, stakeholders were able to scrutinize all possible interrelationships which influenced directly the depletion of the broom grass resource (harvesting unripe grass or by digging). Through the participatory modelling methodology, local beneficiaries were
Page 23 of 41
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
able to foster remarkably a common understanding of the indirect drivers of the depletion the forest resource and to set a collective and well-grounded paths for managing efficiently the broom grass harvesting process. To sum up, by exploring a range of scenarios built on the results of the modelling process, stakeholders have developed a real process of social learning which provides substantial insights. Conservation and sustainable development of the biodiversity and ecosystem services may be hindered by a multiple and complex factors that depleted considerably its values and whole benefits in terms of well-being of local communities. From this angle, the ineffective command-and-control method and the weak community governance in such complex natural resource dynamic system could be better managed and dealt with in such a way that all relevant stakeholders interact together to draft a suitable conservation model and explore collectively expected outcomes drawn from the set of scenarios. Within this approach, social learning throughout the participatory modelling schemes could be of great interest for conflict resolution and multi-sectorial planning process. OSMOSE is a multispecies model (Shin & Cury 2004) which represents the whole life cycle of interacting fish species (growth, predation, reproduction, migration, diverse mortality sources). The model assumes opportunistic predation based on spatial co-occurrence and size adequacy between a predator and its prey. No a priori diets are specified but they emerge from physiological and environmental constraints. Hence the use of OSMOSE is particularly appropriate to explore changes induced by direct drivers such as fishing and climate change which have the potential to trigger changes in habitats and foodweb structure. Furthermore, a recent end-to-end OSMOSE model (OSMOSE E2E) has been developed (Travers et al. 2009; Travers-Trolet et al. 2014) to ensure integration of the main components of the marine ecosystem from the physics, biogeochemistry, exploited fish species up to the fisheries and associated management and socio-economic contexts, while taking into account feedbacks within the environment-human system (Figure 2.11) Based on the coupling of existing disciplinary models (physical, biogeochemical, fish, fisheries and economic models), E2E models benefit from the expertise in each scientific field, and explicitly account for the dynamic forcing effect of climate and human impacts on the marine biodiversity at multiple trophic levels. As such, E2E models present a great potential for exploring a broad range of plausible socioeconomic scenarios (UNEP, MA, IPCC SRES scenarios) and their impacts on biodiversity and associated services. However, for these complex models to be useful beyond the scientific audience, emphasis must be put on producing a synthetic way to communicate simulations results, and on quantifying expected changes into key indicators that could be directly used by decision-makers. In this regard, OSMOSE model allows to produce a variety of indicators in output (size-based, taxon-based, trophicbased) that can be confronted to multiple observational data (surveys and commercial catch data) at different levels of aggregation: at the species level (mean size, mean size-at-age, max size, large fish indicator), and at the community level (Marine Trophic Index, slope and intercept of size spectrum, W statistic of ABC curves, equitability and evenness, etc). A full calibration methodology and software are available to rigorously confront the model to data in hindcast, and best parameterize it to use the model in forecast. All codes of OSMOSE and calibration packages are fully accessible (www.osmosemodel.org). Page 24 of 41
Box 3.3: Osmose model
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 3.11: Example application of OSMOSE end-to-end modelling in the southern Benguela ecosystem (Travers et al. 2014), with climate and fishing drivers explicitly inducing changes in foodweb dynamics (climate induced variability in primary production, change in species and size composition induced by fishing) and in species habitats. Feedbacks between the ecosystem components are represented in the model.
Page 25 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
3.3.3 Summary table of drivers/tools Table 3.2: Summary of drivers, application domains and tools Indirect drivers / Model Demographic application domains Marine habitats Terrestrial habitats
Land degradation Invasive species Climate change Pollution Exploitation of marine resources
AquaMaps, Maxent, DBEM, MARXAN Land change models: global economic models land allocation models multi-agent models
GLOBIOM Aquatox, Modelkey, Fantom EwE, ATLANTIS, OSMOSE, APECOSM
Socio-Economic
Technological
Institutional/ governance/policy
Cultural
Natural Disaster
-
-
X
-
X
-
X
X
X
X
-
X
X
X
X
X
-
X
X
X
X
X
X
X
X
X X
X
X X
X X
X
-
X
X
X
-
-
Page 26 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
3.4 Policy relevance 3.4.1 Recap of policy cycle and different types of scenario approaches
5
10
15
20
Figure 3.12: Policy cycle and scenario approaches (final figure to draw arrows from evaluation to other stages of policy cycle)
A policy cycle serves as a heuristic or framework to facilitate effective decision-making by taking into consideration a comprehensive analysis of the problem, followed by an intervention design, implementation, and finally evaluation of the policy’s success. Although the policy cycle has been criticized for its rationalist approach and perceived emphasis on process over substance,(Bridgman & Davis 2003; Everett 2003) it continues to be the most utilized model for understanding public policy.(Jann & Wegrich 2007) Further, use of a rational framework such as the policy cycle is key to avoiding many of the psychological biases that dominate the decision-making process.(Kunreuther et al. 2002) Thus, while policy making is inherently normative in the establishment of set goals(Robert & Zeckhauser 2011) and there are significant constraints on rational decision-making, particularly at the level of the individual agent, the policy cycle presents a logical depiction of the inputs and outputs at each stage of the policy process. The first stage of the cycle consists of identification of a problem to be addressed through public policy and bringing the problem to the public and formal agenda (agenda setting).(Stone et al. 2001) Here there is an implicit assumption that there are many social problems and limited resources to meet them,
Page 27 of 41
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
necessitating a selection between different problems and issues.(Jann & Wegrich 2007) At this stage various societal stakeholders including the general public, politicians, and researchers vie for influence in the policy arena while the mass media in its framing, tone, and coverage of issues plays a powerful role in increasing public knowledge of and attention to societal problems.(McCombs & Shaw 1972; Scheufele 2000; Scheufele & Tewksbury 2007; Weaver 2007) Within a system of limited political participation, select groups will wield a disproportionate influence over agenda-setting, resulting in biases throughout the process.(Cobb & Elder 1971) To the chagrin of many within the research community, scientific advice has made limited inroads into the policy realm and seldom exerts a direct influence on agendasetting,(Barkenbus 1998) leading to recent studies examining how the scientific community can better translate research into policy.(Vogel et al. 2007; Schenkel 2010). The design of public policy is the stage in which the descriptive is transformed into the prescriptive according to the desired normative approach(Loorbach 2010) (i.e., the will to address a recognized problem is translated into a viable policy formulation with clearly defined objectives.) In order for successful policy decisions to be designed, policy options must be feasible in terms of economic and political resources as well as meet the needs of both the underlying science and interested stakeholders.(Lemos & Morehouse 2005; Jann & Wegrich 2007) Again, while think tanks and scientific research have some impact on policy formulation, their influence relative to other key players should not be overstated. It should also be recognized that although scientific expertise can be used to solve problems, it can also be employed to provide a rational basis for predetermined policies that benefit select stakeholders and interest groups.(Barkenbus 1998). In order to better meet the challenge of translating science into effective practice, boundary workers may facilitate communication at the science-policy interface and help shape the public perception of policy issues(Vogel et al. 2007) although there is some disagreement over their desired involvement.(Hoppe 2009) Finally, it is imperative that the science is useable to meet the pragmatic concerns of local stakeholders as well as those who will be overseeing implementation(Dilling & Lemos 2011) and there is an active dialogue between scientists and stakeholders throughout the duration of the policy.(Carina & Keskitalo 2004; Liu et al. 2008).
30
35
40
Policy implementation has a lower public profile than the identification and design stages yet is an essential and perhaps most challenging stage of the policy cycle model. Here the desired policy is executed via the technical and bureaucratic institutional structures of the state or responsible agency. Public policy can also be implemented through public-private partnerships and subsidization of private research, thus diminishing the potential for market inefficiencies and unwanted spillover effects.(Jaffe et al. 2005) Policy implementation can take the form of top-down (traditional administrative model) or bottom-up models (driven by microlevel actors) and is shaped by the level of ambiguity regarding policy goals and means as well as associated level of policy conflict (e.g., due to disparate values or interests).(Matland 1995) It is also important to take into consideration local level factors such as the political leanings of a population, influence of the manufacturing industry, and degree of fiscal stress as such parameters can impact implementation of climate change mitigation policies.(Sharp et al. 2010)
Page 28 of 41
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
The evaluation stage involves the ex-post reflexive assessment of the extent to which the policy implementation achieved the goals outlined in the initial stage of problem identification. In practice, evaluations are rarely consistent with underlying theory which stipulates that multiple criteria and methods are used, formal policy goals are questioned, and stakeholders are actively involved throughout the process.(Mickwitz 2003; Huitema et al. 2011) Some key obstacles to the realization of policy goals include instrument design oversight, inadequate monitoring, and an absence of effective enforcement mechanisms.(Haug et al. 2010) Further, due to the inherent complexity of the environment-policy nexus, enactment of environmental policies may result in unforeseen externalities that run counter to the original goals or encourage counterproductive behavior such as rebound effects. Climate change policy assessments are also highly contingent on factors such as the choice of discount rates where there is considerable disagreement over the appropriate approach to their formulation.(Goulder & WILLIAMS III 2012)
3.4.2 Description of the role of the different scenario approaches + box with example at different scales for each 3.4.2.1 Exploratory Scenarios
20
25
30
35
According to the IPCC (2001), exploratory (or descriptive) scenarios describe the future according to known processes of change or as extrapolations of past trends. They are sometimes described as "business-as-usual” (BAU) scenarios because often they involve no major interventions or paradigm shifts in the functioning of a system. However, the term "business-as-usual" may be misleading in the policy-making process because exploratory scenarios also can describe futures that bifurcate at some point (an example might be uptake or rejection of a new technology) or that make some assumptions about the functioning of a system. The simplest model is a direct extrapolation of past trends (IPCC 2001). They describe future worlds that might occur in the absence of explicit policies. According to Alcamo and Ribeiro(2001), exploratory scenarios are more common in environmental studies, perhaps because they, in relation to anticipatory, scenarios: require less speculation about the future, tend to be more “value free”. In addition, researchers may be more comfortable with the forward progression of time in exploratory scenarios than with the backward direction of the anticipatory scenarios. Exploratory scenarios can then be constructed both for direct and indirect drivers of change in BD and ecosystems.
40
Main Steps in building exploratory scenarios for direct and indirect drivers of change in BD
Page 29 of 41
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
and ecosystems 1: 1) Identification of research areas (regarding potential changes in BD and ecosystems areas): globally, regionally, nationally, or locally (e.g., Coral reef ecosystems in the Caribbean) 2) Identification of potential changes in BD and ecosystems (e.g.,increasingcoral bleaching and mortality) 3) Identification of main drivers of change (direct and &/or indirect drivers):(e.g., a) climate change (ocean acidification, higher temperatures…); b) Unsustainable socioeconomic activities (turism, fishing,…) 4) Select scenarios axis and scenarios logics: Climate change trends. Socioeconomic stressors in the Caribbean, particularly regarding unsustainable activities in coastal areas and oceans. 5) Building preliminary scenarios Vertical axis:(+)Significant progress in limiting climate change vs. (-) Lack of progress in limiting climate change. Horizontal axis:(+) Significant progress in limiting the effects of socio-economic stressors, particularly in coastal areas and oceans vs. (-) Lack of progress in limiting the effects of socio-economic stressors, particularly in coastal areas and oceans V(+);H (-)
V(+); H(+)
PANEL A: Progress in limiting CC, but adverse impact of unsustainable socioeconomic activities
PANEL B: Progress in limiting CC, combined with growing sustainability of regional socioeconomic development
SHORT NARRATIVE A: Further deterioration of coral reef in the Caribbean Basin (high coral mortality), mainly due to unsustainable socioeconomic practices. HIGH POSSIBILITY of irreversible adverse effects.
1
SHORT NARRATIVE B: Heath of coral reef ecosystems in the Caribbean tend to improve gradually due to the positive change in the main drivers. The possibility of irreversible adverse effects is gradually reduced.
V(-); H(-)
V(-);H(+)
PANEL C: Lack of progressin limiting CC, combined with adverse impact of unsustainable socioeconomic activities
PANEL D: Lack of progressin limiting CC,but growing sustainability of regional socioeconomic development
SHORT NARRATIVE C:Very serious deterioration of coral reef in the Caribbean Basin (extremely
SHORT NARRATIVE D: Further deterioration of coral reef in the Caribbean Basin (very high coral
Based on Willekens, et al. (2010)
Page 30 of 41
First Order Draft
IPBES Deliverable 3c
high coral mortality). EXTREMELY HIGH possibility of irreversible adverse effects.
5
DO NOT CITE
mortality), mainly due to reinforced CC. VERY HIGH possibility of irreversible adverse effects.
Box 3.4: Examples of explanatory scenarios narratives for coral reef ecosystems in the Caribbean Scenario idea has been applied in future climate change project for many years, particularly in the IPCC assessment reports. It started from estimating greenhouse gas emission as the major drivers for climate forcing dating back to the 1990s the Special Report on Emission Scenario (SRES) to the latest Representative Concentration Pathways (RCP). However, these scenarios are normally applied at global scale. Regional scale scenario has been most built under global ones with downscaling techniques for regional information. The specification of model-based scenario assumptions has evolved considerably over time in response to scientific advances in our understanding of climate change as well as the acknowledgement that socio-economic drivers are an integral aspect of formulating potential futures(Abildtrup et al. 2006; Moss et al. 2010). The Special Report on Emissions Scenarios (SRES)(Nakicenovic & Swart 2000) long employed by the IPCC have given way to a new framework formed by the confluence of Representative Concentration Pathways (RCPs) and Shared Socioeconomic Pathways (SSPs). RCPs are constructed from radiative forcing values based on GHG and present a range of potential futures consisting of a low mitigation scenario, two stabilization scenarios, and one high baseline scenario.(Van Vuuren et al. 2011) SSPs as newly formulated by O’Neill et al.(2014) illustrate socioeconomic factors that would make meeting mitigation and adaptation more or less difficult. Building on previous work integrating SRES with socioeconomic scenarios(Abildtrup et al. 2006), this new model takes the form of a dual axis matrix with RCPs representing the possible trajectories of climate change drivers(Moss et al. 2010; Van Vuuren et al. 2011), and SSPs representing possible socioeconomic developments that would impact the ability to mitigate and adapt to climate changes.(van Vuuren et al. 2012). Box 3.5: Evolution of model-based scenario assumptions: From SRES to RCPs to SSPs.
10
3.4.2.2 Goal-oriented scenarios
15
20
This type of scenarios starts with the definition of a clear objective or a set of objectives. These can either be specified in terms of an achievable target (e.g., in terms of the extent of natural habitats remaining, food production self-sufficiency) or as an objective function to be optimized (e.g., minimal environmental damage). Together with these goals and objective functions a set of constraints is defined (e.g., excluding areas for conversion) to ensure somehow realistic outcomes. Optimization techniques are commonly used to find solutions that fulfil the goals and objective functions while accounting for the defined constraints (Castella et al. 2007). Different optimal outcomes are often achievable under the defined set of constraints. Seppelt et al. (2013) indicate that an evaluation of multiple outcomes on a pareto-frontier of tradeoffs between ecosystem services may provide policymakers a range of opportunities of efficient outcomes. The simulation results often show an endpoint of an assumed optimal situation and do not indicate the pathway and the change in indirect drivers required to reach the desired situation. Back-casting techniques and pathway analysis that
Page 31 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
explore the alternative interventions required to reach the optimized situation are scarce. The combination of explorative scenarios with policy options and goal-oriented simulations has the potential to identify policy options that guide the explorative scenario results into the direction of the goal Text box example goal-oriented modelling; zonation tools for protected area allocation under Aichi target
Global protected area expansion is compromised by projected land-use and parochialism According to the Aichi Biodiversity Target 11 adopted by the Convention on Biological Diversity, the protected area network should be expanded to at least 17% of the terrestrial world by 2020. There is considerable risk of ineffective outcomes due to land-use change and uncoordinated actions between countries. The letter shows that with a coordinated global protected area network expansion to 17% of terrestrial land, average protection of species ranges and ecoregions could triple. If projected land-use change by 2040 takes place, it becomes infeasible to reach the currently possible protection levels, and over 1,000 threatened species would lose more than 50% of their present effective ranges worldwide. In addition, a major efficiency gap is found between national and global conservation priorities. Strong evidence is shown that further biodiversity loss is unavoidable unless international action is quickly taken to balance land-use and biodiversity conservation.
5
oriented scenarios (Seppelt et al. 2013) Box 3.6: Example of goal-oriented modeling: zonation tools for protected area allocation under Aichi target
3.4.2.3 Ex-ante scenarios To be elaborated upon 10
3.4.2.4 Ex-post policy assessment To be elaborated upon (Missing sections) 15
3.4.3 Scenarios Approaches and Policy Objectives
Page 32 of 41
First Order Draft
IPBES Deliverable 3c
Table 3.3: Combining scenarios approaches and policy objectives Scenario approaches Brief summary Type of Policy Making / Decision Making Objectives Exploratory Based on plausible Contributing to well Scenarios alternative futures built on informed policies / extrapolations of past trends decisions, but assume and assumptions. absence of explicit policy intervention. Goal-oriented Starts with a prescriptive To identify the pillars scenarios vision of the future & then for policies / decisions (Anticipatory / work backward in time to to build desired future; Normative visualise how this future as well as the various Scenarios) (*) could emerge (*). combinations of drivers (trajectories) to get those objectives. Policy prescriptive Ex-ante scenarios Ex-ante policy assessment Depict the future (Policy Scenarios) (*) effects of environmental policies. Proactive Policy assessment Ex-post scenarios Ex-post policy assessment Looking backward to analyse the gap between environmental policy objectives and real policy results. Reactive Policy Assessment (*) Based on (Alcamo & Ribeiro 2001). 5
DO NOT CITE
Examples
IPCC SRES (2000)
GEO/UNEP, Millenium Ecosystems Assessment (MEA)
Environmental Impact Assessments
2.5 Discussion 2.5.1 Linking scenarios and models
10
The development and quantification of scenarios of indirect driver impacts on direct drivers of biodiversity and ecosystem services is multi-faceted. In many cases multiple models are required to address the different spatial scales and or the different indirect drivers. The case of habitat conversion may require theuse of demographic, economic, and biophysical models to properly represent the development of the different indirect drivers that affect habitat conversion. For regional assessments
Page 33 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
global scale assessment models are often required to account for the influence of distant drivers on the region, while regional models are used to add more regional specificity and detail to the simulations.
5
10
15
20
25
No single model can capture all dynamics at a high level of detail and the coupling or integration of models has become a popular tool to integrate the different dimensions. However, the degree of coupling varies amongst studies. Loose coupling of different, specialized models has as an advantage in that the specific strengths of each model are retained. An example of this approach is the nested modelling approach used by Verburg et al. (2008). Here global economic models explore changes in world consumption and production in terms of the consequences for land use at the level of world regions. Detailed, spatially explicit land change models subsequently downscale calculated areas of land use to individual pixels in order to show the types and location of changes in land use and terrestrial habitats. Based on these results models are used to assess the consequences of land use change for carbon sequestration (Schulp et al. 2008) and ecosystem services. The disadvantage of such loose coupling of models where only limited information is exchanged between the models is the lack of representing feedbacks between the modelled components and the risk of inconsistencies in representation of the same phenomenon in the different models (e.g., a forest in one model can be defined differently in another model). Furthermore, the approach of loose coupling has the risk of the propagation of error and uncertainty between the coupled models which is difficult to track and quantify (Verburg et al. 2013a, 2013b). On the other end of the spectrum, integrated assessment models have been developed that embed the different model representations of the system in a consistent manner. Often, such integrated assessments models are modular and the different modules are built based on simple representations of the system studied. Given the embedding in a single simulation environment, the inclusion of feedbacks and interactions between the different modules is allotted more attention and there is consistent representation of variables across the different modules. Global integrated assessment models have frequently been used for major international assessments such as the IPCC and Global Biodiversity Outlook. For a comprehensive look at integrated assessment modes, see Section 3.3.2.4.3.
30
35
40
For regional scales similar models have been developed that include the most important indirect and direct drivers at regional scales. A disadvantage of this approach is the inherent complexity of the models and the strongly simplified representation of the individual model components. Increased complexity reduces the applicability and transparency of the models (Voinov & Shugart 2013). Although presently these models tend to be used for a wide range of different questions, their model structures often inherit a focus on the specific questions that the models were developed for. Therefore, care should be taken to the range of applications these models are applied to. The choice of integrated modelling versus a loose-coupling of models depends on the specific requirements of the assessment but also on the system studied. An integrated modelling approach is required when feedbacks between the system components or spatial scales studied are important to system outcomes. However, when non-linear dynamics in the individual components are dominating,
Page 34 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
the use of specialised models to capture such dynamics adequately is recommended. Also, in case the study aims to identify leverage points in the dynamics of the indirect drivers, an approach of loosely coupled models may have advantages to study the different components of the system both separately and as part of the full system. 5
3.5.2 Good modelling practice
10
15
20
25
30
35
There is Good Practice Guidance (GPG) to model drivers under the UNFCCC. In particular, there is the GPG for GHG accounting2 and the establishment of the TCCCA principles for good modelling practice to establish forest management reference level (FMRL) scenarios3. The TCCCA principles stand for transparency, completeness, consistency, comparability and accuracy with the final aim to allow a technical assessment of FMRL by an independent reviewer panel organized by the UNFCCC. Modelling of indirect and direct drivers has so far been mainly in the domain of academic research and thus good modelling practice is defined through the peer review process. Even the production of scenarios of key driver scenarios such as GDP and population are not subject to stringent technical quality control measures, many times they are not transparently produced and the credibility of such driver projections typically rest on the credibility of the issuing institutions. In the Macro-economic GDP projections there was an evolutionary transition from the use of originally highly complex structural equation models representing a myriad of interacting economic sectors to very simple even nonstructural time series models, which are mostly used today for projections of up to maximum five years. It is important to notice that these tools were never considered to be useful for long-term GDP projections which are needed for the assessment of BES. Currently there are less than a handful of institutions which issue long-term projections of GDP mostly using tools which were designed for short-term predictions. None of these models consider feedbacks from resource constraints related to BES. More sectorial models of indirect drivers such as integrated assessment model or partial equilibrium models have for good reasons not yet made the transition to simpler models structures as their principle use is less about the forecasting skill of their projections than the study of feasible futures given specific goals under transition constraints. Such models are typically very large and highly complex due to their basic non-linear structures. It is next to impossible to review such model structures with reasonable resources and analyses generated by such models are typically judged on the behaviour of a few output variables which are of interest to the specific paper. Most of these models are used at the stage of policy formulation and very few of these models are actually used for policy planning purposes where review procedures are more biting than academic peer review. Given the fact that there are fundamentally different purposes and subsequent review procedures for different modelling tools it is currently more an art than a science on how to produce consistent scenarios of the long-term behaviour in particular for fundamental drivers such as long-term 2
Information shall be elaborated accordingly with IPCC Guidelines: (IPCC methodologies) 2006 IPCC Guidelines for National Greenhouse Gas Inventories 2013 Supplement to the 2006 Guidelines for National Greenhouse Gas Inventories: Wetlands 2013 Revised Supplementary Methods and Good Practice Guidance Arising from the Kyoto Protocol 3 appendix II to Decision 2/CMP.6 establishes guidelines for submission & review of information on FMRL
Page 35 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
GDP trajectoria. It is unlikely that there will be a major break-through in the science of long-term projections of indirect and direct drivers. Rather there is a tendency to increasingly introduce quality control measures through good practice guidance. 5
10
15
The GPG to model drivers of modulating forest carbon sinks under the UNFCCC serves as a pertinent example. Over more than a decade GPG for GHG accounting for forest ecosystems have been established, which serve as the basic accounting rules for subsequent projections. The modelling process of producing projections is subject TCCCA principles for good modelling practice to establish forest management reference level (FMRL) scenarios. The TCCCA principles stand for transparency, completeness, consistency, comparability and accuracy with the final aim to allow a technical assessment of FMRL by an independent reviewer panel organized by the UNFCCC. There are specific and detailed guidelines for submissions of countries of information on FMRL, requirements for construction and for review of submissions of information on FMRL. The guidelines have differentially been adapted to more and less developed countries conditions after lengthy periods of political negotiation subject to authoritative inputs from the scientific community via the IPCC. Such processes of scientifically informed political negotiations on scenario development finally leading policy changes and potentially also payment streams for ES can serve as useful role models to IPBES.
Page 36 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
References
5
10
15
20
25
30
35
40
45
Abildtrup, J., Audsley, E., Fekete-Farkas, M., Giupponi, C., Gylling, M., Rosato, P., et al. (2006). Socioeconomic scenario development for the assessment of climate change impacts on agricultural land use: a pairwise comparison approach. Environ. Sci. Policy, 9, 101–115. Ahmed, M.T. (2006). Persistent organic pollutants in Egypt-an overview. In: Soil Water Pollut. Monit. Prot. Remediat. Springer, pp. 25–38. Ahmed, M.T. (2010). Ecosystems and human well-being: El-Maghara Northern Sinai Egypt. UNEP. Alcamo, J. & Ribeiro, T. (2001). Scenarios as tools for international environmental assessments. European Environment Agency Copenhagen. Asselen, S. & Verburg, P.H. (2013). Land cover change or land‐use intensification: simulating land system change with a global‐scale land change model. Glob. Change Biol., 19, 3648–3667. Attwood, C., Cochrane, K.L. & Hanks, C. (2005). Putting into practice the ecosystem approach to fisheries. Food & Agriculture Org. Augustyn, J., Petersen, L. & Hamukuaya. (2014). Implementation of the Ecosystem Approach to Fisheries in the Benguela Current LME area (Chapter 19). In: Gov. Mar. Fish. Biodivers. Conserv. Interact. Co-Evol. (eds. Garcia, S., Rice, J. & Charles, A.). John Wiley & Sons. Barkenbus, J. (1998). Expertise and the policy cycle. Energy Environ. Resour. Cent. Univ. Tenn. Sept. Bergman, Å., Heindel, J.J., Jobling, S., Kidd, K., Zoeller, T.R. & Organization, W.H. (2013). State of the science of endocrine disrupting chemicals 2012: summary for decision-makers. Brack, W. (2005). MODELKEY. Models for assessing and forecasting the impact of environmental key pollutants on freshwater and marine ecosystems and biodiversity (5 pp). Environ. Sci. Pollut. Res., 12, 252–256. Bridgman, P. & Davis, G. (2003). What use is a policy cycle? Plenty, if the aim is clear. Aust. J. Public Adm., 62, 98–102. Briot, J.-P., Guyot, P. & Irving, M. (2007). Participatory simulation for collective management of protected areas for biodiversity conservation and social inclusion. AIS-CMS, 7, 183–188. Brown, C., Murray-Rust, D., van Vliet, J., Alam, S.J., Verburg, P.H. & Rounsevell, M.D. (2014). Experiments in Globalisation, Food Security and Land Use Decision Making. PloS One, 9, e114213. Carina, E. & Keskitalo, H. (2004). A framework for multi-level stakeholder studies in response to global change. Local Environ., 9, 425–435. Carpenter, S.R. (2002). Ecological Futures: Building an ecology of the long now 1. Ecology, 83, 2069–2083. Castella, J.-C., Pheng Kam, S., Dinh Quang, D., Verburg, P.H. & Thai Hoanh, C. (2007). Combining top-down and bottom-up modelling approaches of land use/cover change to support public policies: Application to sustainable management of natural resources in northern Vietnam. Land Use Policy, 24, 531–545. CCAFS. (2015). Which way to food security? Four plausible futures for East Africa | CCAFS: CGIAR research program on Climate Change, Agriculture and Food Security [WWW Document]. URL http://ccafs.cgiar.org/blog/which-way-food-security-four-plausible-futures-east-africa#.VK63hHsVRCK. Cobb, R.W. & Elder, C.D. (1971). The politics of agenda-building: An alternative perspective for modern democratic theory. J. Polit., 33, 892–915. Coll, M., Piroddi, C., Albouy, C., Ben Rais Lasram, F., Cheung, W.W., Christensen, V., et al. (2012). The Mediterranean Sea under siege: spatial overlap between marine biodiversity, cumulative threats and marine reserves. Glob. Ecol. Biogeogr., 21, 465–480. Coll, M., Steenbeek, J., Ben Rais Lasram, F., Mouillot, D. & Cury, P. (2014). “Low‐hanging fruit”for conservation of marine vertebrate species at risk in the Mediterranean Sea. Glob. Ecol. Biogeogr. Cook, D.C., Thomas, M.B., Cunningham, S.A., Anderson, D.L. & De Barro, P.J. (2007). Predicting the economic impact of an invasive species on an ecosystem service. Ecol. Appl., 17, 1832–1840.
Page 37 of 41
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Cuttelod, A., García, N., Malak, D.A., Temple, H.J. & Katariya, V. (2009). The Mediterranean: a biodiversity hotspot under threat. Wildl. Chang. World–An Anal. 2008 IUCN Red List Threat. Species, 89. Davis, J.R., Nanninga, P.M., Hoare, J.R.L. & Press, A.J. (1989). Transferring scientific knowledge to natural resource managers using artificial intelligence concepts. Ecol. Model., 46, 73–89. Van Delden, H., Seppelt, R., White, R. & Jakeman, A.J. (2011). A methodology for the design and development of integrated models for policy support. Environ. Model. Softw., 26, 266–279. Dilling, L. & Lemos, M.C. (2011). Creating usable science: Opportunities and constraints for climate knowledge use and their implications for science policy. Glob. Environ. Change, 21, 680–689. Everett, S. (2003). The policy cycle: democratic process or rational paradigm revisited? Aust. J. Public Adm., 62, 65–70. Fletcher, W.J., Chesson, J., Fisher, M., Sainsbury, K.J., Hundloe, T., Smith, A.D.M., et al. (2002). National ESD reporting framework for Australian fisheries: the’how to’guide for wild capture fisheries. Department of Fisheries WA/Fisheries Research & Development Corporation, Canberra, Australia. Garrett, R.L. (1977). Potential western grebe extinction on California lakes. Trans. West. Sect. Wildl. Soc., 13, 80–89. Gichuki, J., Omondi, R., Boera, P., Okorut, T., Matano, A.S., Jembe, T., et al. (2012). Water Hyacinth Eichhornia crassipes (Mart.) Solms-Laubach Dynamics and Succession in the Nyanza Gulf of Lake Victoria (East Africa): Implications for Water Quality and Biodiversity Conservation. Sci. World J., 2012. Goulder, L.H. & WILLIAMS III, R.C. (2012). The choice of discount rate for climate change policy evaluation. Clim. Change Econ., 3. Grübler, A., O’Neill, B., Riahi, K., Chirkov, V., Goujon, A., Kolp, P., et al. (2007). Regional, national, and spatially explicit scenarios of demographic and economic change based on SRES. Technol. Forecast. Soc. Change, 74, 980–1029. Guillette, L.J., Crain, D.A., Gunderson, M.P., Kools, S.A., Milnes, M.R., Orlando, E.F., et al. (2000). Alligators and endocrine disrupting contaminants: a current perspective. Am. Zool., 40, 438–452. Guimaraes, T., Igbaria, M. & Lu, M. (1992). The Determinants of DSS Success: An Integrated Model*. Decis. Sci., 23, 409–430. Haug, C., Rayner, T., Jordan, A., Hildingsson, R., Stripple, J., Monni, S., et al. (2010). Navigating the dilemmas of climate policy in Europe: evidence from policy evaluation studies. Clim. Change, 101, 427–445. Hoppe, R. (2009). Scientific advice and public policy: expert advisers’ and policymakers’ discourses on boundary work. Poiesis Prax., 6, 235–263. Huitema, D., Jordan, A., Massey, E., Rayner, T., Van Asselt, H., Haug, C., et al. (2011). The evaluation of climate policy: theory and emerging practice in Europe. Policy Sci., 44, 179–198. IPCC. (2001). Climate change 2001: impacts, adaptation, and vulnerability: contribution of Working Group II to the third assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press. Jaffe, A.B., Newell, R.G. & Stavins, R.N. (2005). A tale of two market failures: Technology and environmental policy. Ecol. Econ., 54, 164–174. James, I.D. (2002). Modelling pollution dispersion, the ecosystem and water quality in coastal waters: a review. Environ. Model. Softw., 17, 363–385. Jann, W. & Wegrich, K. (2007). 4 Theories of the Policy Cycle. Handb. Public Policy Anal., 43. Jobling, S., Nolan, M., Tyler, C.R., Brighty, G. & Sumpter, J.P. (1998). Widespread sexual disruption in wild fish. Environ. Sci. Technol., 32, 2498–2506. Joubert, D., Zimmermann, I., Fendler, J., Winschiers-Theophilus, H., Graz, F.P., Smit, N., et al. (2014). The development of an expert system for arid rangeland management in central Namibia with emphasis on bush thickening. Afr. J. Range Forage Sci., 1–12. Kannan, K., Blankenship, A.L., Jones, P.D. & Giesy, J.P. (2000). Toxicity reference values for the toxic effects
Page 38 of 41
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
of polychlorinated biphenyls to aquatic mammals. Hum. Ecol. Risk Assess., 6, 181–201. Kindermann, G.E., McCallum, I., Fritz, S. & Obersteiner, M. (2008). A global forest growing stock, biomass and carbon map based on FAO statistics. Silva Fenn., 42, 387. Krueger, T., Page, T., Hubacek, K., Smith, L. & Hiscock, K. (2012). The role of expert opinion in environmental modelling. Environ. Model. Softw., 36, 4–18. Kuemmerle, T., Erb, K., Meyfroidt, P., Müller, D., Verburg, P.H., Estel, S., et al. (2013). Challenges and opportunities in mapping land use intensity globally. Curr. Opin. Environ. Sustain., 5, 484–493. Kunreuther, H., Meyer, R., Zeckhauser, R., Slovic, P., Schwartz, B., Schade, C., et al. (2002). High stakes decision making: Normative, descriptive and prescriptive considerations. Mark. Lett., 13, 259–268. Leclère, D., Havlik, P., Schmid, E., Mosnier, A., Walsh, B., Valin, H., et al. (forthcoming). Climate change induced transformations of agricultural systems: insights from a global model. Environ. Res. Lehsten, V., Sykes, M., Scott, A., Potts, S., Kallimanis, A., Mazaris, A., et al. (forthcoming). Tzanopoulos J, Vogiatzakis I. Disentangling the effects of land use change, climate and CO2 on European habitat types. Glob. Ecol. Biogeogr. Lemos, M.C. & Morehouse, B.J. (2005). The co-production of science and policy in integrated climate assessments. Glob. Environ. Change, 15, 57–68. Liu, Y., Gupta, H., Springer, E. & Wagener, T. (2008). Linking science with environmental decision making: Experiences from an integrated modeling approach to supporting sustainable water resources management. Environ. Model. Softw., 23, 846–858. Loorbach, D. (2010). Transition management for sustainable development: a prescriptive, complexity‐based governance framework. Governance, 23, 161–183. Martelloni, G., Bagnoli, F. & Libelli, S.M. (2012). A dynamical population modeling of invasive species with reference to the crayfish Procambarus Clarkii. ArXiv Prepr. ArXiv12103970. Matland, R.E. (1995). Synthesizing the implementation literature: The ambiguity-conflict model of policy implementation. J. Public Adm. Res. Theory, 5, 145–174. Matthews, R.B., Gilbert, N.G., Roach, A., Polhill, J.G. & Gotts, N.M. (2007). Agent-based land-use models: a review of applications. Landsc. Ecol., 22, 1447–1459. McCombs, M.E. & Shaw, D.L. (1972). The agenda-setting function of mass media. Public Opin. Q., 36, 176– 187. Van Meijl, H., Van Rheenen, T., Tabeau, A. & Eickhout, B. (2006). The impact of different policy environments on agricultural land use in Europe. Agric. Ecosyst. Environ., 114, 21–38. Mickwitz, P. (2003). A framework for evaluating environmental policy instruments context and key concepts. Evaluation, 9, 415–436. Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., Van Vuuren, D.P., et al. (2010). The next generation of scenarios for climate change research and assessment. Nature, 463, 747–756. Mulder, C. & Breure, A.M. (2006). Impact of heavy metal pollution on plants and leaf-miners. Environ. Chem. Lett., 4, 83–86. Nakicenovic, N. & Swart, R. (2000). Special Report on Emissions Scenarios A Special Report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Nel, D.C., Cochrane, K., Petersen, S.L., Shannon, L.J., van Zyl, B. & Honig, M.B. (Eds.). (2007). Ecological risk assessment: a tool for implementing an ecosystem approach for southern African fisheries. WWF Report Series. WWF. Newton, I. (1988). Changes in the status of the Peregrine Falcon in Europe: an overview. Peregrine Falcon Popul. Their Manag. Recovery, 227–234. Ochoa-Hueso, R., Allen, E.B., Branquinho, C., Cruz, C., Dias, T., Fenn, M.E., et al. (2011). Nitrogen deposition effects on Mediterranean-type ecosystems: an ecological assessment. Environ. Pollut., 159, 2265–2279.
Page 39 of 41
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
O’Neill, B.C., Balk, D., Brickman, M. & Ezra, M. (2001). A guide to global population projections. Demogr. Res., 4, 203–288. O’Neill, B.C., Kriegler, E., Riahi, K., Ebi, K.L., Hallegatte, S., Carter, T.R., et al. (2014). A new scenario framework for climate change research: The concept of shared socioeconomic pathways. Clim. Change, 122, 387– 400. Palomo, I., Martín-López, B., López-Santiago, C. & Montes, C. (2011). Participatory scenario planning for protected areas management under the ecosystem services framework: the Doñana social-ecological system in Southwestern Spain. Ecol. Soc., 16, 23. Park, R.A., Clough, J.S. & Wellman, M.C. (2008). AQUATOX: Modeling environmental fate and ecological effects in aquatic ecosystems. Ecol. Model., 213, 1–15. Patel, M., Kok, K. & Rothman, D.S. (2007). Participatory scenario construction in land use analysis: an insight into the experiences created by stakeholder involvement in the Northern Mediterranean. Land Use Policy, 24, 546–561. Peterson, G.D., Cumming, G.S. & Carpenter, S.R. (2003). Scenario planning: a tool for conservation in an uncertain world. Conserv. Biol., 17, 358–366. Pitt, J.P.W. (2008). Modelling the spread of invasive species across heterogeneous landscapes. Lincoln University. Prevedouros, K., MacLeod, M., Jones, K.C. & Sweetman, A.J. (2004). Modelling the fate of persistent organic pollutants in Europe: parameterisation of a gridded distribution model. Environ. Pollut., 128, 251–261. Robert, C. & Zeckhauser, R. (2011). The methodology of normative policy analysis. J. Policy Anal. Manage., 30, 613–643. Schenkel, R. (2010). The challenge of feeding scientific advice into policy-making. Science, 330, 1749–1751. Scheufele, D.A. (2000). Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Commun. Soc., 3, 297–316. Scheufele, D.A. & Tewksbury, D. (2007). Framing, agenda setting, and priming: The evolution of three media effects models. J. Commun., 57, 9–20. Schmitz, C., Biewald, A., Lotze-Campen, H., Popp, A., Dietrich, J.P., Bodirsky, B., et al. (2012). Trading more food: Implications for land use, greenhouse gas emissions, and the food system. Glob. Environ. Change, 22, 189– 209. Schulp, C.J., Nabuurs, G.-J. & Verburg, P.H. (2008). Future carbon sequestration in Europe—effects of land use change. Agric. Ecosyst. Environ., 127, 251–264. Seidl, R., Schelhaas, M.-J., Rammer, W. & Verkerk, P.J. (2014). Increasing forest disturbances in Europe and their impact on carbon storage. Nat. Clim. Change, 4, 806–810. Seppelt, R., Lautenbach, S. & Volk, M. (2013). Identifying trade-offs between ecosystem services, land use, and biodiversity: a plea for combining scenario analysis and optimization on different spatial scales. Curr. Opin. Environ. Sustain., 5, 458–463. Seto, K.C. & Kaufmann, R.K. (2003). Modeling the drivers of urban land use change in the Pearl River Delta, China: integrating remote sensing with socioeconomic data. Land Econ., 79, 106–121. Sharp, E.B., Daley, D.M. & Lynch, M.S. (2010). Understanding local adoption and implementation of climate change mitigation policy. Urban Aff. Rev., 1078087410392348. Shaw, S.D., Brenner, D., Bourakovsky, A., Mahaffey, C.A. & Perkins, C.R. (2005). Polychlorinated biphenyls and chlorinated pesticides in harbor seals (< i> Phoca vitulina concolor) from the northwestern Atlantic coast. Mar. Pollut. Bull., 50, 1069–1084. Shin, Y.-J. & Cury, P. (2004). Using an individual-based model of fish assemblages to study the response of size spectra to changes in fishing. Can. J. Fish. Aquat. Sci., 61, 414–431. Skalský, R., Tarasovičová, Z., Balkovič, J., Schmid, E., Fuchs, M., Moltchanova, E., et al. (2008). GEO-BENE
Page 40 of 41
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
global database for bio-physical modeling v. 1.0. Concepts, methodologies and data. GEO-BENE Database Rep. Standa-Gunda, W., Mutimukuru, T., Nyirenda, R., Prabhu, R., Haggith, M. & Vanclay, J.K. (2003). Participatory modelling to enhance social learning, collective action and mobilization among users of the Mafungautsi Forest, Zimbabwe. Small-Scale For. Econ. Manag. Policy, 2, 313–326. Stone, D., Maxwell, S. & Keating, M. (2001). Bridging research and policy. In: Int. Workshop Funded UK Dep. Int. Dev. Radcl. House Warwick Univ. pp. 16–17. Travers, M., Shin, Y.-J., Jennings, S., Machu, E., Huggett, J.A., Field, J.G., et al. (2009). Two-way coupling versus one-way forcing of plankton and fish models to predict ecosystem changes in the Benguela. Ecol. Model., 220, 3089–3099. Travers-Trolet, M., Shin, Y.J. & Field, J.G. (2014). An end-to-end coupled model ROMS-N2P2Z2D2-OSMOSE of the southern Benguela foodweb: parameterisation, calibration and pattern-oriented validation. Afr. J. Mar. Sci., 36, 11–29. UNEP. (2013). Haiti - Dominican Republic: Environmental challenges in the border zone. United Nations Environment Programme. Vandenberg, L.N., Colborn, T., Hayes, T.B., Heindel, J.J., Jacobs Jr, D.R., Lee, D.-H., et al. (2012). Hormones and endocrine-disrupting chemicals: low-dose effects and nonmonotonic dose responses. Endocr. Rev., 33, 378– 455. Verburg, P.H., van Asselen, S., van der Zanden, E.H. & Stehfest, E. (2013a). The representation of landscapes in global scale assessments of environmental change. Landsc. Ecol., 28, 1067–1080. Verburg, P.H., Eickhout, B. & van Meijl, H. (2008). A multi-scale, multi-model approach for analyzing the future dynamics of European land use. Ann. Reg. Sci., 42, 57–77. Verburg, P.H., Tabeau, A. & Hatna, E. (2013b). Assessing spatial uncertainties of land allocation using a scenario approach and sensitivity analysis: a study for land use in Europe. J. Environ. Manage., 127, S132–S144. Villamagna, A.M. & Murphy, B.R. (2010). Ecological and socio‐economic impacts of invasive water hyacinth (Eichhornia crassipes): a review. Freshw. Biol., 55, 282–298. Vogel, C., Moser, S.C., Kasperson, R.E. & Dabelko, G.D. (2007). Linking vulnerability, adaptation, and resilience science to practice: Pathways, players, and partnerships. Glob. Environ. Change, 17, 349–364. Voinov, A. & Shugart, H.H. (2013). “Integronsters”, integral and integrated modeling. Environ. Model. Softw., 39, 149–158. Van Vuuren, D.P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., et al. (2011). The representative concentration pathways: an overview. Clim. Change, 109, 5–31. Van Vuuren, D.P., Riahi, K., Moss, R., Edmonds, J., Thomson, A., Nakicenovic, N., et al. (2012). A proposal for a new scenario framework to support research and assessment in different climate research communities. Glob. Environ. Change, 22, 21–35. Walz, A., Lardelli, C., Behrendt, H., Grêt-Regamey, A., Lundström, C., Kytzia, S., et al. (2007). Participatory scenario analysis for integrated regional modelling. Landsc. Urban Plan., 81, 114–131. Weaver, D.H. (2007). Thoughts on agenda setting, framing, and priming. J. Commun., 57, 142–147. Willekens, M., Van Poucke, L. & Maes, F. (2010). Developing socio-economic scenarios - Exploratory Scenario Approach. (Belgium Expert Couplet No. 3.3). Establishing future visions. WWF. (2011). Endocrine disruptors and biodiversity. WWF.
Page 41 of 41
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
4 Modelling impacts of drivers on biodiversity and ecosystem functioning 5
10
15
20
25
Coordinating Lead Authors: Lluís Brotons, Villy Christensen & N H Ravindranath Lead Authors: Olivier Maury, Vania Proença, Pablo Peri, Mingchang Cao, Baris Salihoglu, Jung Hwa Chun. Contributing Authors: Rajiv kumar Chaturvedi, Nicolas Titeux… This chapter describes how to use models that represent ecosystems and biodiversity and account for the impacts of human activities. The key messages from the synthesis are: 1. Ecosystems are open systems that are far too complex to be understood and predicted without formal representations (i.e. models). 2. A diverse range of drivers impact global biodiversity and influence ecosystem stability. Modelling biodiversity requires the consideration of relevant drivers and ecological processes to generate projections that are credible. 3. Clarifying the role of modelling to assess the impacts of changes in direct drivers on biodiversity and ecosystem processes might assist policy making over a range of scales. 4. Biodiversity models rest on processes and assumptions. These must be clearly and transparently communicated in order to be of use in informing decision-making. 5. Ecosystems are complex in nature and interact in complex ways with direct drivers. Models should balance complexity and simplicity considering both the decision-making context and an explicit prioritization of processes and feedbacks to be explicitly accounted for. 6. While important gaps remain in the link between ecological theory and models, different approaches are available to conduct assessments and contribute to scenarios development.
4.1 Introduction 30
35
40
Biodiversity and ecosystem dynamics are complex, and environmental drivers, both of natural and anthropogenic origin, induce changes that are commonly translated into changes in ecosystem services and human wellbeing. The present chapter focuses on the approaches and methods currently available to explicitly link these changes in the environment with biodiversity responses from community composition and structure to biogeochemistry fluxes and ecosystem function descriptors. The aim is to identify the range of tools available to unravel mechanisms of change and incorporate this knowledge in models allowing the projection of future biodiversity conditions. The chapter first provides a comprehensive introduction to the context in which biodiversity and ecosystem models are to be developed, and stresses how the decision-making processes should be used to guide model building. Secondly, a description is provided of broad modelling types and drivers to be included in the quest towards better biodiversity projections. Thirdly, we continue with a detailed description of available modelling approaches, and explanations of how environmental drivers of both natural and anthropogenic drivers are introduced into the
Page 1 of 48
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
models, and connected with the respective biological components. Fourthly, we describe the main pros and cons of the different approaches in different contextual applications. Since feedbacks between biodiversity responses and environmental drivers may be important and complexity in biodiversity modelling requires careful integration of the spatial scale and contexts, we illustrate current challenges in these ambits. The issues associated with the sources of uncertainty in model projections are of outmost importance in the context of biodiversity projections for IPBES, and we describe this topic in depth in the context of biodiversity and ecosystem modelling. We finally identify the major challenges for biodiversity projections in the context of the IPBES programme, and highlight major avenues for development to make biodiversity projections more useful for policy makers at a range of administrative scales.
4.1.1 Zooming inside the general IPBES scheme. Links to human drivers and ecosystem services 15
20
25
30
35
40
Currently we are in an era that has been called the Anthropocene where human activities have become global geophysical forces, and major drivers of global environmental change (Steffen et al., 2007). According to Rockström et al. (2009), humanity has transcended the “safe operating space” of the planet with respect to biodiversity loss that play an important role as regulator of ecosystem processes. Traditionally, the discipline of ecology studies aspects related to the ecosphere, and the discipline of economics focus on such flows in social systems. However, within the IPBES framework, there is a need of cross-disciplinary fields such as ecological economics and industrial ecology to understand biodiversity and ecosystem services changes. Thus, environmental problems at the global or regional level may reanalyse the perspective of human-environment systems where social and ecological aspects are interacting at multiple temporal and spatial scales (Clark, 2010). In this context, industrial ecology has in the last few decades emerged as a field aiming at a sustainable development of the anthroposphere, which is the interface between the ecosphere and society. Biodiversity plays multiple roles in the conceptual chain linking direct drivers to nature’s benefits to people. Specifically, it may regulate the ecosystem processes that generate final ecosystem services. Indeed, biodiversity itself may constitute a final ecosystem service, or even a good that is directly enjoyed by people (Mace et al. 2012). In the first case, biodiversity attributes affect the development of ecosystem processes (Cardinale et al. 2012), such as nutrient cycling (Handa et al. 2014), primary productivity (Cardinale et al. 2007) or water infiltration (Eldridge and Freudenberger 2005), which in turn give rise to final ecosystem services, including material (e.g., medicinal plants) and non-material (e.g., water regulation) outputs. In the second case, biodiversity elements are themselves material outputs, with direct use value, such as medicinal plants or fish, but which require human capital inputs (e.g., labour, transport) before being enjoyed by society. Finally, biodiversity elements are a good if directly enjoyed by people without any additional input, which is the case of charismatic species and ecosystems. Although biotic and abiotic ecosystem components interact and are both essential to ecosystem functioning the focus of this chapter will be on the biotic components, as framed by the IPBES conceptual framework (Figure 4.1).
Page 2 of 48
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
Figure 4.1. Role of biodiversity and ecosystems in the IPBES conceptual framework. The present chapter concentrates on the impact of drivers of environmental changes (stressing the role of anthropogenic drivers, Chapter 3) on biodiversity and ecosystem processes as determinants of ecosystem goods and services (Chapter 5)
The role of biodiversity as a regulator of ecosystem processes or as a material output (either a final service or a good), defines the variables of interest when assessing and projecting the impacts of direct drivers. For instance, community data such as species diversity (Cardinale 2007, Mace et al. 2012) or habitat structure (Eldridge and Freudenberger 2005) may be particularly important to assess the impact of drivers when biodiversity has a regulatory role, while population data, such as species distribution (Gaikwad et al. 2011), or population structure (Berkeley et al. 2004) would be more adequate when biodiversity elements have a direct use value. The acknowledgement of the different roles of biodiversity follows an anthropocentric perspective that has ecosystem services, the material and non-material benefits generated by nature, as its main end. In parallel to utilitarian values biodiversity has its own intrinsic value, which is independent of human demand or appreciation.
4.1.2 Decision making context in biodiversity models 20
25
30
Models allowing the assessment of how environmental drivers of change induce changes in biodiversity or ecosystem functions are essential tools to support decision-making. To be effective, models should be able to capture the problem motivating their use and to answer information needs. A formal and accurate definition of the decision-making context and objective is therefore essential in this process (Guisan et al. 2013). A precise definition of the decision-making context will guide the decision of the modelling framework including model complexity, spatial and temporal scales or response variables and data requirements. Response variables should be sensitive to the pressures underlying alternative management scenarios or addressed by policies, and, if possible, be responsive in temporal and spatial scales that are relevant for policy strategies. For example, small farmland birds are responsive to agro-environmental schemes implemented at the field scale whereas large farmland birds are not, being more affected by the conditions found at the larger spatial scales (Concepcíon and Díaz 2011). Moreover, response variables should also be representative of the biodiversity
Page 3 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
attributes underpinning the benefits of nature that are valued in a given decision-making context (see Section 4.1).
5
10
15
Regarding model scope, models should be adjusted to the specific requirements of the decision-making context. Models could rely on observed data to describe the relationship between pressures and response variables, explicitly describe the processes linking those variables, or follow an intermediate approach. The explicit inclusion of mechanisms in modelling approaches will be relevant whenever the understanding of the underlying dynamics is necessary to guide management and where changing environmental conditions call for a mechanistic approach (Gustafson 2013; Collie et al. 2014). The use of correlative approaches, on the other hand, will be adequate if there is limited knowledge about the underlying mechanisms or when model outputs are able to capture the dominant response patterns that are needed to inform policy, such as the evaluation of large scale conservation initiatives (Araújo et al. 2011; Dormann et al. 2012). Regarding model complexity, input data requirements should be balanced against data availability and quality, namely the spatial and temporal resolution of available data, as the lack of or the use of inadequate input data may compromise model feasibility and results quality (Collie et al. 2014).
4.1.3 Why and for what purpose do we need models? 20
25
30
35
40
45
Ecosystems are open systems that are far too complex and multidimensional to be understood without formal abstraction (e.g., models). As humans, we do experience reality but we can only know it through representations. Models are such representations of reality, expressed in scientific terms, most often using mathematical formalisms. Models are tightening together scientific concepts (e.g., energy, population, biomass) using functional hypothesis and the logical grammar of mathematics. They are useful to document the ecological world scientifically, communicate it, and discuss it. Beyond, models are necessary to conceive and think complex scientific objects such as ecosystems and biodiversity, just as words and sentences allow saying but are also the raw material of thinking. Models are therefore at the core of ecological sciences, and ecologists build and use them to represent, explore, share, and discuss their understanding. Ecological models can have various levels of formalisation, from purely qualitative to fully mathematical. The model’s formalisation provides the logical consistency that is needed to manipulate our representations, to make idealized thought experiments that are not possible in real ecosystems, or predict the behaviour of the modelled system beyond the range of observed conditions. Biodiversity is an abstract concept that refers to highly complex living objects (genomes, organisms, populations, communities, ecosystems, Earth system) that are often difficult to observe comprehensively. For that reason, models are useful to formalize the contextdependent scope of biodiversity according to objective definitions, and therefore facilitate discussion and communication. They furthermore allow testing the theoretical consistency of hypothesis on ecosystem functioning and biodiversity response to specific drivers, and their compatibility with observations. They are therefore essential for improving our understanding of ecological patterns and biodiversity. Page 4 of 48
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
Ecological observations often reflect the interaction of multiple processes and convoluted factors acting at various scales. By formalising the ecological complexity at stake, ecological models allow disentangling the factors and processes involved to interpret the data. Revealing theoretical inconsistencies, models stimulate the development of theory. Highlighting observation gaps, models can also be used to guide data collection and stimulate new field and experimental studies. Finally, through their capacity to integrate the effects of various drivers and major interactive processes, and predict their future dynamics, ecological models provide strong basis to build biodiversity scenarios and projections, according to possible evolutions of drivers. They can therefore assist policy makers to understand cause-effect relationship between drivers and the impacts and design efficient policies.
4.1.4 Drivers of environmental change and associated scales 15
20
25
30
35
40
The world has in recent decades experienced global environmental change due to human impact, and this has encouraged research to study the new challenges that biodiversity is exposed to (Pereira et al. 2010). The assessments of the established links between these drivers and biodiversity responses are central to the IPBES and determined by both changes in the environment and to the ecological and physiological processes contributing to the dynamics of these ecological systems even in environmentally static systems (e.g., adaptation can induce range expansion; Lavergne et al. 2010). Thus responses to environmental changes derived directly or indirectly from human activities may be either related to changes in the environment itself, to the biological processes acting within ecosystems or, more frequently, to the combination of both (de Chazal and Rounsevell 2009; Leung et al. 2012). It is therefore important to distinguish between indirect and direct drivers of change in ecosystems and their services, and how changes in ecosystem services have affected human wellbeing. Although biodiversity and ecosystem services experience change due to natural causes, these anthropogenic indirect drivers increasingly dominate current environmental changes. The human sources of these changes lie in two major groups of production and consumption activities: (i) resource extraction with associated processing, use and disposal, and (ii) habitatimpacting measure, e.g., in form of land-use and land-cover change (Ayres, 1989). These activities cause the main direct drivers of environmental change for which we are interested in assessing the responses of biodiversity and ecosystem services. Specifically, temporal and spatial habitat modification in marine and terrestrial habitats, land degradation, the impact of invasive species, climate changes, pollution and exploitation and use of resources (see Chapter 2) induce changes in biodiversity and ecosystem function that are not trivially related to the magnitude of the environmental change recorded. Human impacts on the global environment are operating at a range of rates and spatial scales. Scaling issues are particularly important to assess impacts on biodiversity and ecosystem services because drivers have different impact at different scales. For example, while climate change is a driver that acts at the global scale, habitat modification has an impact on biodiversity and ecosystems services at regional and local scales. In this context, at least half of the ice-free surface of the Earth has already been substantially altered for a variety of human uses (Kates et al., 1990), the methane content of the troposphere has doubled, and the level of carbon dioxide increased by 25%, since preindustrial times (Graedel et al., 1990). Similarly, Page 5 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
humans impact all parts of the world ocean, with a large fraction being strongly impacted by multiple drivers (Halpern et al. 2008).
5
10
15
20
25
A better understanding of habitat modification is of crucial importance to the study of global environmental change. Land uses (agriculture, livestock grazing, settlement and construction, reserves and protected lands, and timber extraction) have cumulatively transformed land cover at a global scale. The main direct drivers affecting this process are external input, (e.g., fertilizer use, pest control, irrigation), technology adaptation and use, resource exploitation, pollutions, species introduction and habitat changes. For marine systems, the main drivers are fisheries (Halpern et al. 2008), notably through bottom-impacting fishing gears (Watling 2005). The consequences of habitat modification have been significant for many aspects of local, regional, and global environments, including climate, atmospheric composition, biodiversity, soil condition, and water and sediment flows. However, global-scale assessments typically mask critical sub-global variations. Local and regional case studies can provide the spatial and temporal resolution required to identify and account for major variations in cause-to-cover relationships and the consequence on biodiversity. But, single-factor explanations, at the macro or the micro scale, have not proven adequate. Many models assessing the impact of environmental drivers on terrestrial ecosystems and biodiversity elements, including those dealing with climate and trace-gas dynamics, require projections of land-cover change as inputs. In this context, Loreau et al. (2003) highlighted that knowledge of spatial processes across ecosystems is critical to predict the effects of landscape changes on both biodiversity and ecosystem functioning and services. Changes in ecosystem connectivity after fragmentation or other anthropogenic and natural perturbations may substantially alter both species diversity and ecosystem processes at local and regional spatial scales. Both increasing and decreasing connectivity can either increase or decrease species diversity and the average magnitude and temporal variability of ecosystem processes, depending on the initial level of connectivity and the dispersal abilities of the organism considered.
30
4.2 Definition of main elements determining ecosystem and biodiversity dynamics 35
40
Scientists and stakeholders in decision-making processes are always faced with the challenge of simplifying reality and selecting key processes and drivers most relevant for their study object (Guisan et al. 2013). Decisions in how and what to include explicitly and what can be simplified or ignored are crucial. The prediction of ecological responses to environmental changes should start with the specification of the major model conceptual components relevant for the case and their critical relationships. In the description of any model of this type, three major conceptual components should be identified: 1. Structural elements describing ecosystem. Target state variable(s) used to describe the biological component of interest, such as biomass, species richness, or habitat heterogeneity (Figure 4.2). State variables should be included based on their sensitivity to pressures and the stability of their response pattern, but also the costs and feasibility of data collection. Page 6 of 48
First Order Draft
2.
5
3.
IPBES Deliverable 3c
DO NOT CITE
Environmental and biotic factors whose spatial dynamics have a direct or indirect effect on the biological component of interest (e.g., climate changes inducing species extinction). In a context of environmental change, changes in the value of environmental and biotic factors will affect the value of state variables. Ecological processes relevant in determining changes in the biological component (e.g., species distribution dynamics such as colonization and extinction).
Figure 4.2. Summary of some of the predicted aspects of climate change and some examples of their likely effects on different levels of biodiversity. From Bellard et al. 2012.
10
15
20
Biodiversity models, as other mathematical models in environmental sciences, are made of a set of components, namely, state variables, external variables, mathematical equations, and parameters (Jørgensen 1994; Smith and Smith 2007, Soetaert and Herman 2009). State variables correspond to the biodiversity variables of interest, the ones that describe the state of the system, such as biomass, species richness, or habitat heterogeneity (Figure 4.2). State variables should be included based on their sensitivity to pressures and the stability of their response pattern, but also the costs and feasibility of data collection (Dale and Beyeler 2001). External variables (or ‘drivers’) correspond to the external factors that influence the state of the system, such as climate or nutrient inputs. In a context of environmental change, changes in the value of external variables will affect the value of state variables. In biodiversity models, Page 7 of 48
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
mathematical equations establish the link between environmental factors (or external variables) and state variables by describing the relationship between them. Models may feature equations that explicitly describe the processes controlling that relationship (mechanistic models), or describe the shape of the response curve by fitting observed data to a predefined function, which is assumed to represent well that process or use a pure statistical approach to derive a relationship from measured values. Parameters are constants in the mathematical equations, which are specific for the system being modelled. The links between external and state variables, and between state variables, as well as the type of processes involved, should be identified at the early stages of the modelling process, namely through the development of a conceptual model (Jørgensen 1994, Smith and Smith 2007). The conceptual model, which can take the form of a diagram, helps to clarify the links between variables but also to define model assumptions – including space and time boundaries (Jørgensen 1994).
15
20
25
30
35
The nature of driver impact on the biological process is key in determining the nature of the model choice and the inclusion of multidisciplinary expertise in the model building process (Guisan et al. 2013). In complex approaches, the impact of drivers on the biodiversity component is compacted through correlative approaches (e.g., SDMs) making the model building process very easy. However, this kind of shortcuts should be performed with caution and full recognition of the assumptions associated to the decision (e.g., ignoring indirect pathways and feedback between element). In a context of environmental change, the effect of environmental pressures on state variables can be direct, (e.g., loss of tree cover after deforestation) or mediated by biological processes, (e.g., ocean acidification and warming affecting coral recruitment and growth, and hence coral abundance and reef structure). In addition, processes also mediate interactions among state variables, (e.g., trophic cascades). The choice of the state variables and processes to figure in a model requires some previous knowledge about the system dynamics. The more realistic a model is, the more complex it will be as realism tends to require description based on solid knowledge of the multiple ecosystem components and processes involved. There is a trade-off with regard to a models predictive capability involved here, where increased complexity leads to a decrease in predictive capability (Walters 1986). Because biodiversity response to environmental change can assume many forms as a consequence of its inherent complexity, there is a large diversity of potential indicators or state variables. A way of addressing this diversity is to reduce it to a few meaningful dimensions.
4.2.1 Describing biodiversity and ecosystem functioning 40
45
Biodiversity indicators or state variables can be arranged along two dimensions representing key aspects of biodiversity complexity, that is, biological organization (species, populations, ecosystems, etc.) and biodiversity attributes (composition, structure, and function). These two dimensions define a conceptual space that can be useful for the identification of state variables and processes ( Table 4.1). More specifically, state variables correspond to composition and structural elements, such as species richness or biomass, and processes to function elements, such as primary productivity, herbivory or competition. Composition and structure emerge from Page 8 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
processes but also affect them (Dale & Beyeler 2001). From an ecological perspective, composition and structure variables describe the structural elements of ecosystems, while processes describe the fluxes of energy and matter and the interactions within and between organization levels. 5 Table 4.1 Examples of compositional, structural, and functional biodiversity variables (Noss 1990, Dale and Beyeler 2001). Based on the representation of the key characteristics of composition, structure and function (from Dale & Beyeler 2001, based on Franklin 1988 at represent the many aspects of biodiversity that warrant attention in environmental monitoring and assessment programs.
Individuals Species/ Populations Community/ Ecosystem
Landscape/ Region
Composition
Structure
Function
Genes Presence, abundance, cover, biomass, density Species richness, evenness and diversity, similarity
Genetic structure Population structure, range, morphological variability Canopy structure, habitat structure,
Habitat richness
Spatial heterogeneity, fragmentation, connectivity
Genetic process Demography, dispersion, growth, metabolism, phenology Species interactions (herbivory, predation, competition, parasitism) Succession, decomposition Landscapes processes (hydrologic processes, geomorphic processes), disturbances
10
15
20
25
30
Ecosystems are open systems. They harness solar energy and transfer it through their various structural elements and organization levels, via different biological and ecological processes. At the biosphere level, water and nutrients, such as carbon, nitrogen and phosphorus, are key structural elements of all living components, and key abiotic components of ecosystems. Their flux across the Earth system is described by the biogeochemical cycles. This flux of energy permits life on Earth and fuels the ecological functions that are useful for societies, (e.g., ecological services). To model the dynamics of biodiversity, the major ecological processes involved in the transfer of energy through ecosystems must be taken into account. These include production, consumption, respiration, and recycling. Other processes such as regulation and evolution are critical to the maintenance of ecosystems over time and also involve biodiversity. Broadly speaking, mechanisms determining ecosystem dynamics can be divided into three broad groups (Lavergne et al. 2010). First, we find mechanisms related to the ecology of the species populations, such as habitat selection, dispersal, and population dynamics. Those processes are primarily determined by species traits expressing the capability of the target species to deal with environmental variability in space and time, (e.g., Hanski et al. 2013; Thuiller et al. 2013). At this level of biological organization, the main structural elements, besides individuals, are the cohorts or life stages, such as seedlings and mature trees. Primary production and respiration are major ecological processes, occurring at the organism level and underlying population dynamics. Organic matter from primary production is at the basis of all Page 9 of 48
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
life on Earth. Numerous factors such as light, the availability of inorganic nutrients, water and temperature influence primary production and have to be considered explicitly in models. Respiration, which comprehends all the living processes using oxygen, is at the core of metabolism. Respiration can be considered at every organization levels, from mitochondria to ecosystems. At the organism level, respiration processes are influenced by many factors including the species considered (body-size scaling rules imply that many metabolic processes vary with the maximum size that a species can reach, Kearney et al. 2010a), the size of individuals, their condition, the availability of food, oxygen and temperature. At the population level, respiration integrates the metabolism of all individuals. It is therefore highly dependent of the size- and state-structure of the population. At the community level, respiration integrates the metabolism of all populations and is therefore controlled by their relative abundance and the structure of the community. Modelling the variability and environmental influences on metabolic biomass production and dissipation therefore requires integrating over several organisation levels, from individuals to communities (Maury and Poggiale, 2013). In the second place, species interactions can restrict or expand the set of places that the species is able to inhabit (Davis et al. 1998). Competition, facilitation or trophic relationships are site and species-specific and account for a great deal of variability in the species capability to survive in a given environment. At the community level, the main structural elements of interest are related with the trophic dynamics (e.g., producers, consumers, detritivores) and other species interactions (e.g., plant-pollinator interactions). Consumption and recycling are main processes associated to trophic interactions. Consumption constitutes a major process of ecosystem dynamics that transfers solar energy along food chains, from primary producers up to top predators. Trophic interactions are influenced by various factors, including by spatialtemporal co-occurrence of grazers/predators and their food/prey, which is often constrained by environmental features. In the oceans, both horizontal and vertical distributions and movements of organisms control trophic interactions, but the vertical distributions are especially dynamic and can change over short time scales. Alternative modelling approaches considering explicit modelling of vertical distributions through three-dimensional models with the complexity this entails have to be explored along with implicit modelling of vertical distributions. Recycling is a major component of the dynamics of ecosystems that must considered in models for closing the system-level mass-balance. Recycling has a critical importance in the biogeochemistry of the Earth System. It is in particular responsible for the storage of carbon in soils and its export to the deep marine sediments via the biological carbon pump (Sarmiento and Gruber, 2006). Finally, evolutionary processes including contemporary microevolution and macro evolutionary trends, allow for critical changes in species traits (Lavergne et al. 2010). Any biological process of interest should have an explicit link with the components formulated in the model. However, this link does not need to be one-to-one.
Page 10 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
4.3 Modelling approaches for assessing the impact of drivers on biodiversity and ecosystem functioning 5
10
15
20
25
30
35
Various types of ecological models are used for assessing the impacts of large-scale environmental drivers on the ecosystem. Ecological models can be classified in two broad categories: phenomenological and mechanistic. Amongst the phenomenological approaches, purely statistical models, that only rest on correlations between variables, must be distinguished from phenomenological process-based models that explicit some relevant processes in a phenomenological way. Mechanistic approaches attempt to derive ecosystem dynamics from first principles, essentially like theoretical physics attempts to describe the nonliving universe. Practically, since the development of a purely mechanistic theory of ecosystems is still to come, existing approaches can be classified along a gradient from purely phenomenological to purely mechanistic. Correlative models use statistical methods to establish relationships between environmental variables and biodiversity data, such as species richness, abundance or distribution (Morin and Lechowicz 2008). These models deliver information on biodiversity patterns and responses to drivers based on empirical observations, but do not explain the mechanisms underlying those patterns and responses (Rahbeck et al. 2007). Model performance and realism is therefore dependent on a careful choice of environmental predictors and of the spatial resolution and extent of the data, but also on the judicious use of the model given model assumptions (Elith and Leathwick 2009, Araújo and Peterson 2012).
Figure 4.3. A simple schematic showing the relationship between two observations of a species distribution in the ‘real world’, ‘statistical models’ and ‘dynamic, process based models’. From McInerny & Etienne 2012.
Process-based phenomenological models are empirical constructions. They use the inductive scientific method, going from empirical observations to likely generalizations. Unlike purely statistical models, process-based phenomenological models explicitly consider some relevant processes that they represent using mathematics as a tank of flexible (potentially non parametric) shapes that are fitted to data to mimic the observations. That is providing them with some functional capacity and the possibility to interpret the data within a formal theoretical framework. However, process-based phenomenological models may have limited cognitive capacity (they describe rather than they explain) and they are limited by (1) the sensitivity of the system dynamics to the arbitrariness of the mathematical form used to represent the process, (2) the sensitivity to the data used to estimate the parameters and the Page 11 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
impossibility to predict beyond the range of observed conditions, (3) the disconnection of processes that leads to a lack of internal consistency and over-parameterization.
5
10
15
20
25
At the other end of the formalization gradient, pure mechanistic models also termed as theoretical models are axiomatic constructions (Gallien et al. 2010). Like in theoretical physics, they use the hypothetico-deductive scientific method, starting from hypothesis (the axioms) to deduce predictions that can be tested empirically, to falsify or conversely to corroborate the hypothesis made (but never to prove it or “validate” it). Mechanistic models are based on few basic rules and general principles in which processes are explicitly accounted for and from which their dynamics can be deduced logically and mathematically. They consist in mathematically derived parameterizations where parameters are meaningful constants. They are based on clear processes and causality chains formulated from “first principles”. Mechanistic models of biodiversity have to be internally consistent with both biological theory and mathematical rules. They must represent the biological and ecological processes, respecting dimension rules and verifying the principle of energy and mass conservation. In the literature, the term mechanistic is often used in a wider perspective to include models explicitly accounting for processes in their internal structure (i.e. correlative mechanistic models, Kearney et al. 2010). In this chapter we use the term process based in the context of explicit implementation of ecological processes in the model and mechanistic to describe models with stronger foundations on ecological theory. Mechanistic models of biodiversity may help improve our understanding, interpret observations, guide data collection, and stimulate new field and experimental studies. Mechanistic models have a high cognitive capacity (they explain more than they describe). However, they are often very complex, empirically undetermined, and difficult to use practically.
4.3.1 Phenomenological correlative modelling approaches 30
35
40
45
Phenomenological correlative models are frequently used to assess the impacts of human activities on biodiversity, forecast future impacts of environmental changes, support human productive activities (e.g., enhance agricultural production) and conservation actions (e.g., identify sites for translocations and reintroductions, predict the location of rare and endangered species), understand species’ ecological requirements, among other uses (Peterson, 2006; Muñoz et al. 2011). Correlative models have the advantage of offering greater interpretability as to causation of phenomena, and permit predictability of phenomena that depend on the differences between components—e.g., the invasive potential of a species depends on the difference between potential and actual distributional areas (Peterson, 2006). Correlative models can be applied at all spatial scales. The choice of the appropriate spatial scale should consider the scale of variation of relevant environmental predictors (Elith and Leathwick 2009). For instance, the effect of climate variables is better assessed at large spatial extents, such as regions, and coarse resolution data may be acceptable, whereas the effect of soil nutrients requires fine resolution data, to cover fine-scale variations, and is usually modelled at smaller extents, such as landscapes. When the selected environmental predictors act at different scales, hierarchical models with nested sub-models can be used (Elith and Page 12 of 48
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
Leathwick 2009). Similarly, advances in modelling marine ecosystems will require coupling of different trophic levels that may have different resolution. A way in representing biodiversity in complex marine systems would be to concentrate the detail of representation at the target species. Regarding temporal scales, correlative models are often static (i.e., assume a constant time frame) failing to capture species or community dynamics, such as species dispersal. Nevertheless, temporal predictors, such as variability of food resources, may be added to models to capture variation in the state of biodiversity variables. More complex dynamics, such as species dispersal, can be integrated in modelling approaches, such as Dynamic Bioclimate Envelope models (DBEM), where an envelope niche model is combined with dynamics of species dispersal (Cheung et al. 2008). An important caveat of correlative models, concerns extrapolation to new spatial and time frames, namely hind-casting and forecasting applications, since the conditions associated to training data (i.e., the data used to fit the model), may not be constant over time (Elith and Leathwick 2009, Araújo and Peterson 2012) or be inadequate to represent the conditions found outside their area of distribution. For instance, the data informing niche models are not drawn from the entire abiotic niche or even from the potential distribution, but from the actual area of distribution, which calls for caution when interpreting results (Araújo et al. 2005; Pearson and Dawson 2003). Alternative approaches based explicitly on mechanisms are likely to be robust under novel environmental combinations but are limited by the availability of data to build the models. In contrast, data required to fit correlative models is readily available across a range of scales and the models can implicitly capture many complex ecological responses. Because of this, Elith et al. (2010) anticipate ongoing use of correlative models for biodiversity projections.
25
30
35
Species occurrence and abundance are common state variables in correlative modelling approaches at the species level whereas variables of community diversity and structure are common in community-level approaches (Iverson et al. 2001). Environmental datasets often comprise climatic variables (e.g., mean annual temperature, annual precipitation, and annual solar radiation), topographic variables (e.g., average elevation, elevation coefficient of variation, elevation range), and habitat variables (e.g., habitat type, habitat extent). For instance, a study to assess the impacts of climate change on the distribution of Pinus densiflora in South Korea (Chun and Lee 2013) utilised 27 environmental variables from 4 categories of climate characteristics, geographic and topographic characteristics, soil and geological properties, and MODIS EVI, and about 4,000 species occurrence and abundance records from National Forest Inventory sites across the whole country.
4.3.1.1 Species-level modelling 40
Species distribution models (SDMs), also known as niche models or bioclimatic envelope models, have been widely used to model the effects of environmental changes on species distribution across all realms (Pearson & Dawson 2003, Elith and Leathwick 2009). A bioclimatic envelope can be defined as a set of physical and biological conditions that are suitable to a given species, it is generally identified through the relationship between species observed distribution and environmental attributes using statistical methods (e.g., generalized Page 13 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
additive model, Luoto et al. 2005) or artificial intelligence models (e.g., artificial neural network, Pearson et al. 2002).
5
10
15
20
25
30
35
40
Main critiques to bioclimatic envelope models relate to their omission of biotic interactions, evolutionary change and species dispersal processes (Pearson & Dawson, 2003). Still, examples exist where bioclimatic envelope models were developed to include some of these processes. The Dynamic Bioclimate Envelope Model developed by Cheung et al. (2008) simulates changes in the relative abundance of marine species through changes in population growth, mortality, larval dispersal and adult movement following the shifting of the bioclimate envelopes induced by changes of climatic variables. The model does not account for species interactions and potential food web changes. A new dynamic bioclimate envelope model is being developed to account for effects of ocean biogeochemistry such as oxygen level and pH on the ecophysiology and distribution of marine fish (Stock et al. 2011). SDMs often use species presence-absence data, however, in some situations there is only access to presence data (e.g., museum data). In those situations other models, such as GARP and MaxEnt, can be used (Elith and Leathwick 2009). GARP (Genetic Algorithm for Rule set Production) is a machine-learning system that has shown predictive ability in delineating species’ ecological niches and predicting geographic distributions (Peterson et al., 2002). At an initial stage, GARP produces four types of rule sets (atomic, range, negated range, and logistic regression) explaining the relationship between the distribution of species presence data and explanatory environmental variables. Then it uses an iterative process of rule modification and selection until it reaches the most appropriate set of rules, which can then be used to predict species distributions. MaxEnt is a presence/background method, that is, model fitting uses presence-only data and data on environmental variation across the study area (i.e., background data) (Peterson et al. 2011). The algorithm estimates the species range according to a probability distribution of maximum entropy (i.e., closest to uniform) constrained by information on environmental data at presence sites (Araújo et al. 2007, Phillips and Dudík 2008, Merow et al. 2013). Maxent has been used for several purposes, such as, to assist conservation planning, to predict potential distributions for invasive species, and to predict climate and land use change impacts (Elith et al. 2010). Case study: Conservation of Iberian lynx populations (Fordham et al. 2013) Decision-making context: Iberian lynx populations have suffered severe declines during the last century due to habitat change, changes in abundance of preys (due to habitat change and overexploitation), direct persecution and road killing. Today, the Iberian lynx is the world’s most endangered cat. On-going conservation actions include habitat protection and restoration for the species and its preys (mainly rabbits), ex-situ breeding and reintroductions. Policymakers are considering the establishment of new Iberian lynx populations in areas within its recent historical range, namely in different autonomous regions in the southern half of Spain. However, this solution may be compromised by climate change due to impacts on habitat suitability and barriers to dispersal. An alternative would be to perform reintroductions in habitat-favorable areas across all Iberian Peninsula, even if outside the recent historical range. Page 14 of 48
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
Modelling approach: The future distribution of Iberian lynx populations was assessed using coupled ecological niche models (ENM) of Iberian lynx and European rabbit range dynamics under different climate change and conservation scenarios. Climate change scenarios include a high-CO2 stabilizing reference scenario and a greenhouse gas mitigation Policy scenario. Conservation scenarios, included a business as usual scenario, an extramanagement scenario (no reintroductions), and two reintroductions scenarios, within the recent historical range and Peninsula-wide. ENM were coupled to metapopulation simulations of source-sink dynamics to account for stochastic changes in Iberian lynx populations driven by changes in prey abundance, due to disease, climate and land-use changes, and by changes in habitat suitability, due to climate and landuse changes. Main results: Probability of extinction is high for the business as usual and extra management conservation scenarios (>85%) for the two climate-change scenarios. Planned reintroduction programs, on the other hand, can reduce extinction risk (<5%). Planning reintroductions in habitat-favorable areas across all Iberian Peninsula may be more successful than the restricting reintroductions to the recent historical range.
4.3.1.2 Species traits approaches
30
Trait based ecological risk assessments (TERA) is a trait based approach to assess ecological responses to natural and anthropogenic stressors based on species characteristics related to their functional roles in ecosystems (Baird et al. 2008). In case there is enough information on functional characteristics and roles of taxonomic species or communities in the ecosystems, TERA has potentials in ecological risk assessment adding trait attributes to map onto existing taxonomic entities but more cooperative effort is needed to address the challenges such as task of trait definition, linkage of traits to specific environmental drivers, and the extraction of trait information by observation, experiment, and literature data mining (Baird et al. 2007).
35
4.3.1.3 Community-level modelling
40
Community-level models are used to produce information at the collective level (Ferrier, 2002). The community metric of interest can take a number of forms, such as predictive mapping of community types (groups of locations with similar species composition), species groups (groups of species with similar distributions), axes or gradients of compositional variation, levels of compositional dissimilarity between pairs of locations, and various macroecological properties (e.g., species richness) and even phylogenetic diversity (Ferrier & Guisan, 2006). Community-level modelling offers an opportunity to move beyond species-level predictions to predicting broader impacts of environmental changes (e.g., Hilbert & Ostendorf Page 15 of 48
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
2001; Peppler-Lisbach & Schröder 2004), which may be relevant in certain decision-making contexts. Examples are when time and financial resources are limited, when existing data is spatially sparse or the knowledge on individual species distribution is limited (Ferrier, 2002), and when species diversity is beyond what can feasibly be modelled at the individual species level. Three community-level modelling approaches have been proposed (Ferrier & Guisan, 2006): (1) ‘Assemble first, predict later’, whereby species distributions are first combined with classification or ordination methods and the resulting assemblages are then modelled using machine learning or regression-based approaches, (2) ‘Predict first, assemble later’, whereby individual species distributions are modelled first and the resulting potential species distributions are then combined (i.e. the result is in fact the summation of individualistic models), and (3) ‘Assemble and predict together’, whereby species distribution models are fitted using both environmental predictors and information on species co-occurrence. Reviews on usefulness of community-level modelling can be found in Ferrier et al. (2002) and Ferrier & Guisan (2006) Table 4.2. The six main type of spatial output that can be generated using community level modeling (Ferrier and Guisan, 2006) Community metrics
Description
Structure of derived grid layers
Individual species
Predicted distribution of multiple species, as for species-level modelling
Community types
Each’ community type’ defined as a group of locations (grid cells) that closely resemble one another in terms of predicted species composition. Grouping normally achieved through some form of numerical classification
Species groups
Each ‘species group’ defined as a subset of species with similar predicted distributions. Grouping again achieved through numerical classification, but in this case the objects classified are species rather than locations A set of continuous axes (or gradients) representing dimensions of a reduced space that summarizes the compositional pattern exhibited by multiple species. These axes most commonly derived through some form of ordination The predicted level of dissimilarity in community composition between all possible pairs of grid cells in a region
A separate layer for each species, indicating the predicted probability of occurrence or abundance of that species in each cell Either (i) a single layer with each cell assigned exclusively to one community type (depicted as a map with different colours indicating different types) or (ii) a separate layer for each community type, indicating the probability of that type occurring in each cell (depicted as multiple grey-scale or colour-ramp maps) A separate layer for each species group, indicating the predicted prevalence or abundance of that group in each cell (depicted as multiple grey-scale of colour-ramp maps)
Axes of compositional variation
Levels of compositional dissimilarity between pairs of cells Macroecological properties
Most commonly modelled property is species richness, either of a whole group (e.g. all vascular plants) or of a functional subgroup (e.g. annuals and trees). Many other macro-ecological properties (e.g. mean range size and endemism) can potentially be modelled
A separate layer for each axis, indicating the predicted score for that axis in each cell (depicted either as multiple grey-scale or colour-ramp maps, or as a single map by assigning each of the first three axes to a different colour dimension, e.g. red, blue, green) In theory a complete matrix of pair-wise dissimilarities, but in practice these values are usually predicted dynamically as required by the application of interest (difficult to depict spatially without prior conversion to community types or axes of compositional variation) A separate layer for each macro-ecological property (depicted as a grey-scale or colour-ramp map)
Page 16 of 48
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
Also important at the community-level are species-area relationship (SAR) models, which use habitat area as a predictor of community species richness. SARs have been applied to a wide range of taxa and across all scales, from local to global. SAR functions are often used to predict the impacts of changes in habitat availability, driven by land use change (e.g., van Vuuren et al. 2006, Desrochers et al. 2011) or climate change (e.g., Malcom et al. 2006, van Vuuren et al. 2006), on community richness, but also to assess the impacts of direct exploitation on community parameters, such as species-turnover rates (e.g., Tittensor et al. 2007). Reviews on the use of SARs can be found in Rosenzweig 1995, Drakare et al. 2006, and Triantis et al. 2012).
4.3.2 Process-based and mechanistic models Considering the relevant processes provides process-based phenomenological models with some functional capacity and the possibility to interpret the data within a formal theoretical framework. However, process-based phenomenological models have poor cognitive capacity (they describe rather than they explain) and they are limited by (1) the sensitivity of the system dynamics to the arbitrariness of the mathematical form used to represent the process, (2) the sensitivity to the data used to estimate the parameters and the impossibility to predict beyond the range of observed conditions, (3) the disconnection of processes that leads to a lack of internal consistency and over-parameterization. Let’s look at a few illustrative examples to clarify these limitations: 1. Even though they can theoretically be derived from ecological mechanisms, functional response curves are usually used as flexible and practical regression functions linking the amount of food eaten by a predator to the density of prey, disregarding the underlying mechanisms that control their shape. Yet, when embedded in ecosystem models, the choice of the functional response’s form (e.g. type I, II or III) leads to profound differences in the modelled ecosystem dynamics (Fulton, 2003; Anderson et al., 2010). 2. Other classical examples of phenomenological representation include the allometric equations that are typically used for representing the various processes of metabolism (e.g., respiration is proportional to some power – named the allometric coefficient – of individual weight). Because allometric functions are stationary and do not explicit the influence of environmental factors on metabolism, any change of environmental conditions is absorbed by a corresponding change in parameter value. Changing the dataset used to estimate the allometric coefficients therefore leads to different estimated values and associated responses of metabolism to weight (e.g. Bonnet et al., 2003). As a consequence, parameters are meaningless system- and state-specific “constants that vary” in space and time to capture the observations. 3. Because metabolism includes very different processes with specific dynamics, fitting allometric regressions to different processes of metabolism (e.g., respiration, reproductive output, age at maturity, aging mortality) leads to as many allometric coefficients as processes considered. Using independent allometric curves to represent the various metabolic processes at stake in an ecosystem model therefore leads to overparameterization and do not consider the interactions between the processes represented.
Page 17 of 48
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
Process-based models allow a more explicit representation than correlative approaches of ecological processes mediating biodiversity responses to environmental drivers. However, and despite their wide use in biology and ecology, process-based phenomenological ecosystem models, as all model developments suffer from fundamental and practical limitations that should be explicitly acknowledged. Examples Various strategies and levels of process formalisation can be distinguished amongst the available process-based phenomenological models of ecosystem:
10
15
20
1. Box models. This is the simplest and most developed category. It describes the ecosystem structure as a set of bulk functional groups that are connected together by fixed uptake/predation terms based on given functional responses, and that dissipate energy through simple respiration terms that do not explicit the various components of metabolism. The main advantage of these models is that they balance the mass and energy fluxes at the scale of the system represented. The counterpart is that they neglect important phenomenon such as the importance of size in controlling metabolism, predator–prey interactions and life-history omnivory (i.e. diet changes when organisms grow). Examples include nutrient-phytoplankton-zooplanktondetritus (NPZD) type models (e.g. PISCES ref, MEDUSA ref, ERSEM ref, TOPAZ ref, etc…) that represent 3D marine biogeochemistry in climate models, mass-balanced trophic models (e.g. EwE Polovina, 1984; Christensen and Pauly, 1992; Christensen and Walters 2004) that represent marine communities in 0-D (Ecopath, Ecosim) or in 2-D (Ecospace), often with a focus on upper trophic level organisms.
25
30
35
40
45
2. Age/stage-structured models. These models are box models that are structured along a dimension (usually age or stage) that is supposed to be functionally important. They explicit some processes of metabolism such as growth, reproduction and the agedependence of respiration. This group, among others includes dynamic global vegetation models (DGVM) that represent the dynamics and biogeochemistry of the global vegetation cover, the Atlantis model (Fulton et al. 2005) that represents the trophic interactions and biogeochemistry of marine trophic networks in 3-D boxes, and the dynamic EwE models, which incorporates age-structured population models (Walters et al. 2000; Walters et al. 2010). Spatial population dynamics models (e.g. Lehodey, 2008) use habitat-preference functions to describe population movements with advection-diffusion equations. 3. Size-structured models. Some approaches emphasize the interest of size to structure ecosystem models. In marine ecosystems, size rather than taxonomic identity controls predation (e.g., Shin and Cury, 2004) and metabolism (e.g., Gillooly et al., 2001). Also, size-based models are easier and cheaper to parameterize than food-web models. A wide variety of size-structured models exist. This include 0-D marine ecosystem sizespectrum models, where ecosystem dynamics is projected along the size dimension “from bacteria to whales”, disregarding the metabolic and physiological differences between species (e.g., Platt and Denman, 1978; Benoit and Rochet, 2004; Maury et al., Page 18 of 48
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
2007). In the same perspective, Madingley (Harfoot et al., 2014) represents the 2-D energy fluxes through size-structured functional groups with a lagrangian approach, in both terrestrial and marine environments. Recently, the need to consider explicitly the functional role of species diversity in the dynamics of functional groups has emerged. This has led to the development of trait-based approaches (e.g., Andersen and Beyer, 2006; Hartvig et al., 2011), where the community-level dynamics emerges from the interactions of a large number of species that are distinguished solely on the basis of their maximum size, which captures most of the inter-specific differences of metabolism and life history (Kooijman, 1986). Other approaches, like the 2-D lagrangian models, Osmose (Shin and Cury, 2004) and size-based Ecospace model (Walters et al. 2010), represent explicitly a set of key size-structured species and emphasize the importance of size and spatial co-occurrence in controlling trophic interaction and the emergent structure of marine communities. Dynamic Global Vegetation Models (DGVM) The most advanced tool to estimate the impact of climate change on vegetation dynamics at global scale are Dynamic Global Vegetation Models (DGVMs). These are process-based models that simulate various biogeochemical, biogeophysical and hydrological processes such as photosynthesis, heterotrophic respiration, autotrophic respiration, evaporation, transpiration and decomposition. DGVM models simulate shifts in potential vegetation and the associated biogeochemical and hydrological cycles as a response to shifts in climate. DGVMs use time series of climate data and, given the constraints of latitude, topography, and soil characteristics, simulate monthly or daily dynamics of ecosystem processes. DGVMs are most often used to simulate the effects of future climate change on natural vegetation and its carbon and water cycles. DGVMs integrate biogeochemistry, biogeography, and disturbance sub-models, and disturbance is generally limited to wildfires (for example in IBIS wildfires are represented by static values). DGVMs are usually run in a spatially distributed mode, with simulations carried out for thousands of “grids” or geographic points, which are assumed to have homogeneous conditions within each grid. Under DGVM framework, simulations are carried out across a range of spatial scales, from global to landscape. With a growing interest in the biodiversity changes in an ecosystem, these ESM’s generally comprise of the coupling of DGVMs with AOGCMs. Basic structure of a DGVM is given in Figure 4.4.
Page 19 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 4.4. Structure of DGVMs. Source: http://seib-dgvm.com/oview.html
5
10
15
20
25
Dynamic models capture the transient response of vegetation to a changing environment using explicit representation of key ecological processes such as establishment, tree growth, competition, death, nutrient cycling (Shugart 1984, Botkin 1993). Plant functional types (PFTs) are central to DGVMs as, on the one hand, they are assigned different parameterizations with respect to ecosystem processes (e.g., phenology, leaf thickness, minimum stomatal conductance, photosynthetic pathway, allocation, rooting depth), while on the other, the proportion of different PFTs at any point in time and space defines the structural characteristics of the vegetation (Woodward and Cramer 1996). Input and outs: DVMs use spatially explicit time series of climate data (e.g., temperature, precipitation, humidity, sunshine days, wind) and take into account the features of topography and soil characteristics in order to simulate monthly or daily dynamics of ecosystem processes. Plant species are represented as groups with similar physiological and structural properties, termed Plant Functional Types (PFTs), which are designed to represent all major types of plants (Sitch et al. 2008). DGVMs are most often used to simulate the effects of future climate change on natural vegetation and its carbon and water cycles. DGVM provides policy relevant outputs with respect to ecosystems such as shifts in vegetation types, NPP and carbon stock changes, biomes and habitat change. Advantage: Key advantages of using DGVMs include its capacity to simultaneously models the transient responses related to dynamics of plant growth, competition and, in a few cases, migration. As such it allows the identification of future trends in ecosystem function and structure and these models can be used to explore feedbacks between biosphere and atmospheric processes (Bellard et al 2012).
4.3.2.1 Mechanistic models
30
Mechanistic models of biodiversity are tools to improve our understanding, interpret observations, guide data collection, and stimulate new field and experimental studies. Their consistency and structural stationarity (e.g., mechanistic models keep valid in other environments, including beyond the range of observed conditions) provide a basis to build and analyse scenarios and projections. Mechanistic models have a high cognitive capacity (they Page 20 of 48
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
explain more than they describe). However, they are often very complex, empirically undetermined and difficult to use practically. Purely mechanistic models and theories in the fields of biodiversity are rare. In practice, they are often incomplete and articulate mechanistic descriptions of the components under focus with phenomenologic parameterizations of the less known or less important phenomenon. However, in recent years there has been growing effort to better place existing approaches into more available ecological theories allowing wider use and greater potential of the models derived from this approach. Some of these examples are the theory based biodiversity model introduced by Thuiller et al. (2013), which used metapopulation theory as the cornerstone for a new generation of biodiversity models. This kind of approaches has the advantage of being explicit about the processes involved and provides interpretable parameters while at the same time has solid foundations in population dynamics. Furthermore, the new developments build on the recent addition of environmental heterogeneity, dispersal limitation, and biotic interactions into the incidence function and include the addition of an evolutionary perspective as a new challenge.
15
20
25
30
35
40
Another theory used in mechanistic modeling is the Dynamic Energy Budget (DEB) theory (e.g., Kooijman 2010) which is a mechanistic theory aiming at capturing the quantitative aspects of metabolism at the organism level from a small set of key assumptions (Sousa et al., 2008a). The DEB theory allows accounting for environmental variability effects on organisms through food and temperature changes. It captures the diversity of all possible living forms on earth in a single mechanistic framework. This allows representing the energetics and major life history traits of all possible species in a community with the same set of unspecific taxa-dependent DEB parameters. The existence of inter-specific scaling rules is a fundamental property of the DEB theory. It emphasizes the fundamental importance of the maximum size of the species and implies that functional biodiversity (in terms of energetics and life history) has actually far less degrees of freedom that one would expect given the high number of species on Earth. This major finding opens the way to a mechanistic understanding of functional biodiversity and ecosystems dynamics. DEB-based individual bioenergetics can indeed be scaled up to build physiologically-structured mechanistic population dynamics models that can in turn be scaled up using interspecific scaling rules to derive size-based mechanistic models of the dynamics of communities (Maury and Poggiale, 2013). The APECOSM model adopts this strategy to represent the dynamics of marine ecosystems in a mechanistic way (Maury, 2010). Embedded in a 3-D mechanistic framework representing organisms movements (Faugeras and Maury, 2007) and environmentally-driven behaviours, the trait-based dynamics of canonical DEB-structured marine communities is modelled from regional to global scales. APECOSM can be coupled to an Ocean General Circulation Models (OGCM) and an Ocean General Biogeochemical Model (OGBM) that provide the timedependent and tri-dimensional environmental forcing variables, both for hind- and forecasting (Lefort et al., 2014; Dueri et al., 2014).
4.3.2.2 Other approaches
45
Deterministic Models: In these models one cause is modeled to have one or many effects. The term “forest dynamics” spans a huge range both in time and space. For example the enzymatic reactions of photosynthesis operate within fractions of a second; foliage development takes a Page 21 of 48
First Order Draft
5
10
15
20
25
30
35
IPBES Deliverable 3c
DO NOT CITE
few weeks, while tree growth lasts decades to centuries, and the dynamics of soil organic matter span millennia (Bugmann 1994). On the other hand, the germination of a seed takes place on a few square cm3, a sunfleck moving over the forest floor covers a few m2; a dominant tree in the canopy occupies 0.01-0.1 ha, and the quasi-equilibrium of a forest landscape may be reached on the scale of several hectares (Shugart and Urban 1989). There have been two basic approaches to modeling vegetation response to changing climates, static (timeindependent) and dynamic (time-dependent) models. Static or equilibrium models: Static models assume equilibrium conditions in both climate and terrestrial vegetation in order to predict the distribution of potential vegetation by relating the geographic distribution of climatic parameters to the vegetation. The equilibrium approach is implicitly large scale in nature as it ignores dynamic processes. It generally requires far less information and provides estimates of potential magnitude of the vegetation response at regional to global scales. The equilibrium models are restricted to the estimation of the steadystate conditions. Prominent examples of static models include BIOME4 (Kaplan 2003) and Mapped Atmosphere Plant Soil System (MAPPS; Neilson 1995). Biogeochemistry Models: Biogeochemistry models project changes in basic ecosystem processes such as the cycling of carbon, nutrients, and water. These models are designed to predict changes in nutrient cycling and primary productivity. The inputs to these models are temperature, precipitation, solar radiation, soil texture, and atmospheric CO2 concentration. The plant and soil processes simulated are photosynthesis, decomposition, soil nitrogen transformations mediated by microorganisms, evaporation and transpiration. Common outputs from biogeochemistry models are estimates of net primary productivity, evapotranspiration fluxes and the storage of carbon and nitrogen in vegetation and soil. Some of the popular models include BIOME-BGC (Running and Hunt, 1993), CENTURY (Parton et al. 1993), and the Terrestrial Ecosystem Model (TEM; McGuire et al. 1992). Biogeography Models: Biogeography models simulate shifts in the geographic distribution of major plant species and communities. They analyze the essential environmental conditions over entire continents, to estimate the type of vegetation that is most likely to cover a given area. These types of models are best suited for assessing broad-scale changes in vegetation. They are based on ecophysiological constraints, which determine the broad distribution of major categories of woody plants, and response limitations, which determine specific aspects of community composition, such as the competitive balance of trees and grasses. Examples of biogeography models include the Mapped Atmosphere Plant Soil System (MAPSS; Neilson, 1995), BIOME3 (Haxeltine and Prentice 1996), and IBIS (Foley et al. 1996). Input datasets for biogeography models mainly include latitude, mean monthly temperature, wind speed, solar radiation, and soil properties such as texture and depth.
40
4.3.3 Expert-based systems
45
Knowledge-based systems generally refer to expert system approaches to prediction or decision support. An expert is someone who has achieved high knowledge on a subject through her/his life experience (Kuhnert, Martin, & Griffiths 2010), therefore being a reliable source of information in a specific domain. Expert knowledge-based species-habitat Page 22 of 48
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
relationships are used extensively to guide conservation planning, particularly when data are scarce (Iglecia et al. 2012). Expert knowledge is quite commonly utilized in conservation science (Janssen et al. 2010). Ecological expert knowledge has been frequently incorporated in aquatic habitat suitability model to link environmental conditions to the quantitative habitat suitability of aquatic species (Mouton et al. 2009). Research on species distribution modelling, which incorporated expert knowledge into modelling process is relatively limited. In a study on “finessing atlas data for species distribution models”, Niamir et al. (2011) incorporated existing knowledge into a Bayesian expert system to estimate the probability of a species being recorded at a finer than the original atlas data to predict a bird species distribution. They noted that knowledge-based species distribution maps of finer scale using a hybrid MaxEnt/expert system had a higher discriminative capacity than conventional approach even though such an approach might be limited to well-known species. In a study on “tradeoffs of different types of species occurrence data for use in systematic conservation planning”, Rondinini et al. (2006) noted that the geographic range data of species generated by expert knowledge had advantage of avoiding the potential propagation of errors simplifying data processing steps. However, eliciting expert information involves dealing with multiple expert judgements, with different sources of biases in the elicited information and with uncertainty around expert estimates (Martin et al. 2012). For instance, expertise may vary geographically, with relevant information restricted to the region of interest of the experts (Murray et al. 2009).
4.3.4 Hybrid modelling 25
30
35
40
Hybrid models combine multiple modelling approaches to represent complex, integrated systems of human and biophysical components (Parrot et al. 2011). These models are highly generally data driven and serve to aid in pattern extraction and knowledge synthesis, providing an important link between data sources and decision support systems. Developments of hybrid models take a pathway where some of the ecological processes defining the ecological system under study are (i.e. realized niche) are modelled explicitly (process-based or mechanistically), while others are kept in the correlative niche model. Hybrid approaches derive from the interest to balance realism and flexibility in model building with limited knowledge. However, this approach is not without problems. The way different models are integrated into hybrid approaches often is a difficult issue. Gallien et al. (2010) indicate that one of the current limitations of these hybrid approaches is the form and strength of the relationship between habitat suitability and demographic parameters. Changes in habitat suitability are normally integrated with population processes by limiting carrying capacity. Furthermore, how ecological processes (e.g., growth, dispersal, and thermal tolerance) respond to environmental changes is unclear, and is often assumed to be unimodal or linear. Nonlinear functional response could make the model more complex. More importantly, most functional responses used in the model are assumed, rather than confirmed from experiments. Fortunately, when simple occupancy models are used, model integration becomes easier: cells that become unsuitable force or induce extinction, whereas colonization is allowed in those cells that become suitable only.
Page 23 of 48
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
Broadly speaking, mechanisms determining ecosystem dynamics can be related to the ecology of species, species interactions and evolutionary processes (Lavergne et al. 2010). Any biological process of interest should have an explicit link with the components formulated in the model. However, this link does not need to be one-to-one (Lurgui et al 2014). The implementation of these processes in the model may be carried out in a wide variety of ways spanning a broad range of complexities from cellular automata (e.g. Iverson et al. 2004; Smolik et al. 2010), metapopulation models (e.g. Wilson et al. 2009), structured metapopulation models (e.g., Akçakaya 2000) to spatially-explicit population models (e.g., Cabral & Schurr 2010), individual-based models (Grimm & Railsback 2005) and reaction-diffusion models (e.g., Wikle 2003; Hui et al. 2011). Models with emergent dynamics may also include species interactions (e.g., Albert et al. 2008), or abiotic processes included via feedbacks (e.g., wildfires vs. vegetation growth; Grigulis et al. 2005). Emergent dynamics can be spatially auto correlated or not. For example, a spread model for an invasive species will generate autocorrelated spatial patterns (e.g., Merow et al. 2011), whereas a module simulating local birth and death may generate temporal dependency but may not generate auto correlated spatial structures by itself.
4.4 Modelling biodiversity feedbacks and issues across temporal and spatial scales. 20
25
30
Both human and non-living environmental drivers influence biodiversity through a number of processes. In turn, biodiversity exerts feedbacks on both systems (Fig. 4.6). Considering the feedbacks is important because they are susceptible to cause non-linearity in the interaction dynamics. This is especially important as non-linearity can potentially lead to whole-system bifurcations (e.g. regime shift) when forcing variables such as atmospheric CO2 evolve beyond observed levels. Changes in biodiversity interact with different drivers of biodiversity change (i.e. climate change, disturbance regimes such as forest fires, invasive and pests, and ecosystem processes) over different temporal and spatial scales. Changes in biodiversity and range shifts (plant traits) can influence climate at global and regional scales. For instance, General Circulation Models (GCMs) based simulations indicate that widespread replacement of deep-rooted tropical trees by shallow-rooted pasture grasses would reduce evapotranspiration and lead to a warmer, drier climate (Shukla et al. 1990). Similarly the replacement of snowcovered tundra by a dark conifer canopy at high latitudes may increase energy absorption sufficiently to act as a powerful positive feedback to regional warming (Foley et al 1994).
Page 24 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 4.5 Schematic diagram of the interactions between biodiversity, the human system and the nonliving environment. The figure represents the feedbacks between biodiversity and the drivers to biodiversity change and their interactions (Source: Chapin et al 2000).
5
10
15
20
25
30
Besides, changes in biodiversity are known to induce changes in the microclimate. Such changes in microclimate may further impact the biodiversity. For example, in a boreal forest ecosystems, where soil temperatures have a strong influence on nutrient supply and productivity, the presence of moss, which reduces heat flux into the soil, provides stability to the permafrost leading to the low rates of nutrient cycling (Von Cleve et al 1991). As fire frequency increases in response to high-latitude warming, moss biomass decline, the stability of permafrost declines, nutrient supply increases, and the species compositions of forests are further altered. Chapin et al. (2010) outlines the possible interactions and feedbacks between biodiversity, biodiversity change drivers and the interactions between these drivers and suggests that anthropogenic activities (arrow 1 on Figure 4.5) cause environmental and ecological changes of global significance (2). By a variety of mechanisms, these global changes impact biodiversity, and changing biodiversity feeds to the susceptibility to species invasions (3, purple arrows). Changes in biodiversity can have direct consequences for ecosystem services and as well as human economic and social activities (4) In addition, changes in biodiversity can influence ecosystem processes (5) and ecosystem services (6). These further feedbacks to alter biodiversity (7, red arrow). Global changes may also directly affect ecosystem processes (8, blue arrows). Chapin et al. (2010) further conclude that “depending on the circumstances, the direct effects of global change may be either stronger or weaker than effects mediated by changes in diversity”. Changes in biodiversity influence annual rates of primary productivity (NPP) and decomposition of organic matters in soils. Increased NPP production relates to increased carbon sequestration from the atmosphere, and thereby provides negative feedback to Page 25 of 48
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
climate warming. Similarly, increased decomposition of soil organic matters and litter relates to a loss of carbon from the biosphere to the atmosphere, and thereby provides a positive feedback to climate warming. Several studies using experimental species assemblages conclude “annual rates of primary productivity and nutrient retention increase with increasing plant species richness, but saturate at a rather low number of species” (Tillman et al 1996, Hector et al 1999). Salonius (1981) concludes that microbial richness can lead to increased decomposition of organic matter. Further, particular species can have strong effects on ecosystem processes by directly mediating energy and material fluxes or by altering abiotic conditions that regulate the rates of these processes. Introduction of the deep-rooted salt cedar (Tamarix sp.) to the Mojave and Sonoran Deserts of North America increased the water and soil solutes accessed by vegetation, enhanced productivity, and increased surface litter and salts. This inhibited the regeneration of many native species, leading to a general reduction in biodiversity.
15
4.4.1 Tools for assessing feedbacks
20
25
30
35
40
45
Feedbacks between drivers and biodiversity or ecosystem levels usually involve high level of complexity in the models because changes in state variables at different levels (either biological or biological and other) should be able to interact and cause emergent dynamics. Changes in biodiversity for instance can impact disturbance regimes such as fire, which in turn is strongly determined by climate (Pausas and Keely 2009) and fire-suppression efforts (Brotons et al. 2013). As an illustration, several species of nutritious but flammable grasses were introduced to the Hawaiian Islands to support cattle grazing. Some of these grasses spread into protected woodlands, where they caused a 300-fold increase in the extent of fire. Most of the woody plants, including some endangered species, are eliminated by fire, whereas grasses rebound quickly (D’Antonio et al 1992). Similar increases in the ecological role of fire resulting from grass invasions have been widely observed in the Americas, Australia, and elsewhere in Oceania. Species diversity is also known to reduce the probability of outbreaks by 'pest' species or invasives by diluting the availability of their hosts. This phenomenon particularly decreases host-specific diseases (Burdon, 1993), plant-feeding nematodes (Wasilewska, 1995) and consumption of preferred plant species (Bertness and Leonard, 1997). Further, biodiversity characteristics can impact the ability of exotic species to invade communities through either the influence of traits of resident species or some cumulative effect of species richness. Early theoretical models and observations of invasions on islands indicate that species-poor communities would be more vulnerable to invasions because they offered more empty niches (Elton, 1958). Fire disturbance models at the landscape scale have been used to integrate different processes and model feedbacks between vegetation and disturbances regimes. The LANDIS for instance model simulates forest succession, disturbance (including fire, wind, harvesting, insects), climate change, and seed dispersal across large landscapes and allows tracking the spatial distribution of discrete tree and shrub species (Scheller et al. 2007). If feedbacks are to be incorporated across scales or very different domains of applications, approaches explicitly integrating different models should be developed. One example of this complexity is exemplified by the use of Integrated Assessment Models (IAMs, see Figure 4.7). Page 26 of 48
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
In the IPCC Third Assessment Report (IPCC, 2001) integrated assessment was defined as “an interdisciplinary process that combines, interprets, and communicates knowledge from diverse scientific disciplines from the natural and social sciences to investigate and understand causal relationships within and between complicated systems”. It is generally agreed that there are two main principles to integrated assessment, i.e. integration over a range of relevant disciplines; and the provision of information suitable for decision-making (Harremoes and Turner, 2001). IAMs thus aim to describe the complex relations between environmental, social and economic drivers that determine current and future state of the ecosystem and the effects of climate change, in order to derive policy-relevant insights (Vuuren et al., 2009). One of the essential characteristics of integrated assessment is the simultaneous consideration of the multiple dimensions of environmental problems. IAMs typically describe the cause-effect chain of climate change from economic activities and emissions to changes in climate and related impacts on e.g. ecosystems, human health and agriculture, including some of the feedbacks between these elements. In order to make their construction and use tractable, many IAMs use relatively simple equations to capture relevant phenomena. This simplification is most obvious for the climate system and carbon cycle, which, in many IAMs, consists of only a few equations (Goodess et al. 2003). However, the behaviour of these components can have a significant impact on IAM results and the quality of policy advice, with the possibility of simplifications in the earth system projections leading to imprecision (or even error) in projecting impacts and costs of mitigation.
Figure 4.6 General structure of an IAM.
25
30
Over the last decade, IAMs have expanded their coverage in terms of land use and terrestrial carbon cycle representation, non-CO2 gases and air pollutants and by looking into specific impacts of climate change. Some IAMs have a stronger focus on economics, such as multisectoral computable general equilibrium models that are combined with climate modules and models focussed on cost–benefit analysis. Whereas, other IAMs are more focussed on the physical processes in both the natural system and the economy (integrated structural models/biophysical impact models). Few examples of IAMs are: IMAGE, DICE, FUND and MERGE. The basic structure of IMAGE 2.4 version is described below (see Figure 4.7). The key drivers of change, population and the macro-economy, can be derived from various external Page 27 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
and internal sources. For macro-economic drivers the exogenous source depends on the study in which the model is applied.
5
10
However, as IAMs aim to integrate different aspects of the environment, they run the risk of becoming extremely complex. Thus, the developers of such models have to make decisions about the focus of their study and how they wish to express the impacts they are attempting to estimate, whether it is through the reporting of physical changes in emissions, shifts in landuse activity or mortality rates, or through cost-benefit analysis of damages resulting from climate change (Goodess et al., 2003). The data requirements for these IAMs are also large and not always feasible.
Figure 4.7 Framework of IMAGE 2.4. Source: Bouwman et al., 2006
4.5 Accounting for uncertainty 15
20
Biodiversity and ecosystems are under stress due to drivers, and the stress is likely to increase in the future due to drivers such as climate change, habitat modification, exploitation, and pollution. Policy making on biodiversity and ecosystem functioning may have to take place based on current knowledge. This must be done recognizing that uncertainty is associated with all modelling, due to limitations of data, representation of processes, and resolution of the ecosystem scale. However, policy makers have to take decisions even in the face of uncertainty, to act on the drivers in order to conserve ecosystems and biodiversity. Page 28 of 48
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
No discussion of modelling frameworks would be complete without consideration of uncertainty (Leung et al 2012). Uncertainty in modelling of biodiversity and ecosystem processes under changing climate and land-use dynamics arises from multiple factors. For example, existing impact assessment studies including the biophysical and integrated assessment model (IAM) based studies, generally tend to examine the ‘means’ of the impacts probability distribution, neglecting the low-probability, high impact tails of the distribution (Marten et al 2013; Ackerman 2010; Weitzman 2009). Further, impact projections generally model linear climate change and inadequately account for factors such as non-linearity and tipping elements in the climate system, and irreversibility of the impacts – which are difficult to model (Whiteman et al. 2013). Impact studies generally focus on single-sector or single region-based assessments. The potential interactions, among sectors and regions, which can adversely impact the biodiversity and ecosystems, are not adequately included in the quantitative estimates (Warren 2011). Additionally, impact assessment models generally leave out the natural processes and feedbacks that are difficult to model at current state of knowledge, even though these processes may cause large impacts. For example, pest attack and fire dynamics in terrestrial and forest ecosystems are often not included in the biophysical impact assessment processes. Similarly the ambient policy and management practices and socio-economic stresses leading to degradation of natural resources are also not included in majority of the sectorial impact assessment models. Key human related issues such as armed conflicts, migrations and loss of cultural heritage has a lot of potential to impact the natural ecosystems, however impact assessment models do not include these stresses related to the human systems (Hope 2013; Theisen 2013). IAMs based economic analyses of impacts are generally conservative as these studies make optimistic assumptions about the scale and effectiveness of adaptation (Marten et al 2013; Hope 2013). Fischling et al. (2007) conclude that the most advanced tool to assess the impact of climate change on vegetation dynamics at global scale include DGVMs.
4.5.1 Sources of uncertainty Link et al. (2012) and Leung et al. (2013) highlighted six major sources of uncertainty confronting ecosystem modellers (Figure 4.8).
Page 29 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 4.8 A conceptual diagram of the flow of information and actions in a typical LMR management system. Rectangles represent components of the system, solid arrows indicate flows of information and actions between components, and ellipses represent major sources of uncertainty (Link et al. 2012)
5
10
15
20
25
30
1) Natural variability Natural variability or stochasticity includes biological differences among individuals within a population, differences among populations within a community, changes in spatial distributions through time, density-dependent or independent variation in a vital rate, seasonal or interannual variability in realized environmental conditions, or shifts in productivity regimes. Natural variability increases ecosystem model uncertainty by reducing the precision of parameter estimates. 2) Observation error Observation error is inevitable when studying organisms in either a single species or an ecosystem context (e.g., Morris and Doak 2002; Ives et al. 2003 as cited in Link et al. 2012). The environmental characteristics of a particular area (even those that we can measure fairly accurately) are difficult to relate directly to the full experience of mobile organisms that move into and out of that area. Thus, natural variability can actually exacerbate observation error. Observation error adds uncertainty to ecosystem models through reduced precision, misspecified parameter distributions, and biased parameter estimates. 3) Structural complexity The structural complexity of a model arises from many factors, such as the number of parameters it includes; the number of ecosystem components and processes it simulates; the temporal scale; the nonlinearities, log effects, thresholds, and cumulative effects incorporated in those processes; and whether or not it includes features such as spatial dynamics or stochasticity (Fulton et al., 2003 as cited in Link et al. 2012). Structurally complex ecosystem models are gaining in use, in part due to improved computing capabilities and also to the intricate, multi-sector, cross-disciplinary questions commonly being addressed in ecosystemPage 30 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
based management. Thus, models have optimal levels of complexity (Walters 1986, Fulton et al., 2003).
5
10
15
20
25
30
35
40
45
Epistemic uncertainty contains parameter uncertainty, model (structural) uncertainty reflected in the model structural complexity and data/observation error. Epistemic uncertainty reflects our level of knowledge about a system, and can be reduced with additional information. We build verbal (qualitative) or mathematical (quantitative) models to represent our understanding of processes underlying a system. We use empirical data to parameterise the model. However, given stochastic elements present, we never have perfect estimates of underlying parameters and processes. Furthermore, given that models are our conceptual abstractions of the real world, even the basic structure of the model is uncertain. In addition, there will always be some uncertainty in our data sources, which may propagate through the risk model. 4) Inadequate communication Inadequate communication (or linguistic uncertainty) relates to the difficulty of effectively conveying information between scientists, managers and stakeholders. Communication problems range from the subtle to the obvious (Peterman, 2004 as cited in Link et al. 2012). When communication is ineffective, information is lost, which can manifest itself as uncertainty in many ways (Martin et al. 2012). 5) Unclear management objectives There are numerous obstacles to the development of clear objectives. Many issues related to resource use or conservation are poorly understood and are not amenable to clear objectives, or involve vague constructs that are difficult to link to performance metrics (sensu Smith et al. 1999 as cited in Link et al. 2012) 6) Outcome uncertainty Outcome uncertainty arises when realized in situ values deviate from expected values derived from a model. It is also referred to as “implementation error” or “implementation uncertainty” because it is commonly associated with differences between a management goal and the implementation of the management plan (Rice and Richards 1996; Dichmont et al. 2006; Holt and Peterman 2006; Dorner et al. 2009 as cited in Link et al. 2012). Outcome uncertainty may also be connected to other forms of uncertainty described above. For example, unclear management objectives or other forms of inadequate communication could lead stakeholders to reject management plans because the stakeholders feel their best interests are being overlooked (Watson-Wright et al. 2009 as cited in Link et al. 2012). All of the ecosystem models are diverse in terms of scope and approach, but share the general feature of a large number of parameters with complex interactions. These models are necessarily built with imperfect information. Given these inevitable uncertainties, large and complex ecosystem models must be evaluated through sensitivity analyses before their output can be effectively applied to conservation problems (Hilborn and Mangel 1997; Saltelli et al 2000b; Regan et al. 2002; Clark 2003; Harwood and Stokes 2003; Pielke and Conant 2003; Tang et al. 2006 as cited in McElhany et al. 2010). Uncertainty in climate scenarios arises from Page 31 of 48
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
different greenhouse gas emission storylines and from differences between climate models even if driven with the same greenhouse gas emission scenario (e.g., Buisson et al. 2010 as cited in McElhany et al. 2010). This can be partly addressed by using climate change scenario data from several emission scenarios but also by using results from multi‐model studies (i.e. an ensemble of climate models). Process‐based models (PBMs) are important tools in forest and ecological science and widely used to assess the impacts of climate change on forest ecosystems (Landsberg 2003; Fontes et al. 2010 as cited in McElhany et al. 2010). Parameter uncertainty has received less attention in climate change impact studies. Climate change impact studies that do not integrate parameter uncertainty may over‐ or underestimate climate change impacts on forest ecosystems. Box. Uncertainty in DGVMs Process based, dynamic models capture the transient response of vegetation to a changing environment using explicit representation of key ecological processes such as establishment, tree growth, competition, death, nutrient cycling (Shugart 1984, Botkin 1993). With the adoption of DGVMs reliability of results has improved in relation to previous generations, however validation of some of these models is an ongoing exercise (e.g., Prentice et al. 2007). The large structural and parametric uncertainty in the DGVMs concerning the processes of recruitment, competition and tree mortality means that existing DGVMs produce a wide variety of predictions regarding the future strength and direction of the climate carbon cycle feedback (Thornton et al., 2007; Sitch et al., 2008). For example, some models predict catastrophic declines in the Amazon and boreal forests, while others predict relatively stable ecosystem composition and carbon storage, even with the same future climate drivers (as illustrated by Sitch et al., 2008). Model biases introduced by these uncertainties are not readily estimated because limited observations exist to constrain demographic processes under rapidly altering climates (Allen et al., 2009). For issues such as climate change effects on forest growth, there are large uncertainties in both the magnitude of future climate change, and the magnitude of various responses to a changed climate (Dixon and Wisniewski 1995). Many of the DGVMs do not yet simulate forest fires dynamically as well as pest attacks and land use change dynamics. Further, migration and seed dispersal constraints are not yet incorporated in DGVMs. Many of the DGVMs are known to over-simulate grasslands (Bonan et al. 2003) and at the same time there are suggestions that compared with other dynamic vegetation models, some DGVMs tends to simulate a fairly strong CO2 fertilization effect (Cramer et al. 2001; McGuire et al. 2001). Additionally uncertainty in the projections of biodiversity and ecosystem processes arises from uncertainty in climate projections as well, particularly in precipitation at downscaled regional levels (Chaturvedi et al 2012). Poor land use projections (including afforestation, reforestation and forest regeneration) due to anthropogenic influences further add to the uncertainty. However it is important that the uncertainty is adequately represented and acknowledged in the impact assessment studies. We suggest the following specific improvements for the future research to capture the full range of uncertainty in the impact assessments: Use of multiple climate models and use of multiple impact assessment models; Integrated modelling of the natural and production system Developing Plant functional Types for more vegetation types and at species level and for consideration of fauna. Page 32 of 48
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
Similarly the results of complex models, such as lake ecosystem models, often suffer, however, from limitations due to various sources of error and uncertainty such as the initial conditions, input data, model structure, model parameters, validation data, etc. (Beck 1987 as cited in Gal 2012). Parameter uncertainty is a key issue when dealing with a complex model due to the large number of parameters and the uncertainty as to their true values (Helton et al. 2006). In order to increase the reliability of the model as a management tool it is important to estimate the degree of uncertainty surrounding those predicted relationships. There are diverse approaches for estimating and quantifying uncertainty of the various model components affecting model outcome (Walker et al. 2003; Helton et al. 2006).
10
15
20
25
Box. Uncertainty in species distributions models (SDMs). Modelling exercises should aim at minimising and when possible explicitly accounting for known sources of uncertainty and critical sources of uncertainty. Sources of uncertainty in biodiversity modelling used in a decision making context are not equivalent and therefore, a critical analysis should be performed in assessing which potential sources of uncertainty may impact the outcomes of the modelling approach (see Chapter 5 on decision making). Recent reviews on the role of uncertainty in biodiversity modelling approaches such as SDMs have indentified a number of uncertainty sources of different types and suggested solutions and good practices to deal with them (Rocchini et al. 2011). They identified significant sources of uncertainty in the SDMs modeling building approach and explicitly mention possible solutions that could be indentified as good practice. As an example uncertainty in input data cold be related to uncertainty in coordinate information or positional error associated to presence records that should be dealt with by improving screening of data previous to model building. This work strongly suggests that the collection of these uncertainty sources and the collection of possible solutions and good practices should be favoured in the future and greatly facilitate the development of new applications.
4.6 Options for reducing uncertainty Simeon et al. (XX) highlighted some considerations on reducing uncertainty in the structural form of models of ecosystem dynamics. 30 1. 2. 35 3. 4. 40 5.
Identify the purpose of the modelling exercise in terms of management objectives, and the performance measures by which the attainment of these objectives will be assessed. Identify the key uncertainties about the system. This should occur during the process of assembling information and formulating models; it is necessary to identify and highlight uncertainties rather than to make assumptions that constrain the models to a single view of any important process. Develop models or parameterisations that represent plausible limits to each key uncertainty. Consider more than one basic model structure. Always include models, assumptions, and parameterisations that are not on the bounds, (and hence may be more plausible), and ensure that the choice of models, parameter values, and assumptions is balanced given the purpose of the modelling. Establish the full range of model behaviours by considering different combinations of models and parameters. Sensitivity analysis is useful at this stage to determine the importance of each source of uncertainty Page 33 of 48
First Order Draft
6.
5 7.
8. 10 9. 10. 15 11.
20
12.
IPBES Deliverable 3c
DO NOT CITE
Consider the interaction between models and data. Specifically, do the models capture the full range of potential conditions, or just the conditions represented by the data? If the latter is the case, it may be advisable to consider adopting plausible future scenarios that extend outside of the range of historic data, as may occur, for example, under current predictions for climate change. Ensure that each model is logically consistent. For example, assumptions or fixed values for key parameters in an ecosystem model will also need to be made in models used to derive prior density functions for the ecosystem model’s input parameters. Weigh models by plausibility if information exists to do this. Ideally this weighting would be on the basis of posterior probability, but a more subjective weighting by prior probability might be necessary. Run each model multiple times to incorporate the effects of parameter uncertainty and natural variability. Avoid averaging model results unless the distribution of results suggested by all models is unimodal. If it is possible to weight models, present the results in terms of the risk that each management objective will not be met. If it is not possible to weight models, present the results in terms of the trade-offs associated with each management action for each alternative model. Make sure the assumptions and limitations of the approach are presented along with the results.
4.7 Ways forward in biodiversity and ecosystem modeling 25
30
35
Modelling biodiversity and ecosystem response to direct drivers of change adequately for predicting future outcomes is a challenge that can hardly be surmounted because of the inherent complexity of real-world processes (Araújo et al. 2005). In the quest towards models with enhanced predictability, biodiversity and ecosystem models will certainly become more complex and include different components. One of the costs inherent to increasing model complexity is the need for appropriate model parameters. Building complex models with unrealistic parameters derived from lack of knowledge or data is often less preferable to simple models with realistic parameters. Both model calibration and evaluation call for data, and whereas the need of obtaining realistic models for a world with many species is on the rise, little information on their spatial dynamics is currently available. An increase in the number of processes explicitly formulated in process-based models is leading to a larger number of parameters and unsurprisingly to less predictive capability and increase the uncertainty associated with the results (Walters 1986). Therefore, increased model complexity should be justified and match the scientific question at hand. In addition, increased model complexity will challenge the ability of modellers to communicate what is being formulated into the model and how this is being conceptually done.
40 Increased complexity in biodiversity model applications can be due to: (1) increased number of drivers acting in a model through input layers; or (2) increased number of processes explicitly considered in the model. In correlative niche models many biological processes are collapsed and implicitly included (Guisan & Thuiller 2005; Soberón 2007). The pathway of increasing the Page 34 of 48
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
number of external inputs may lead to poor predictive performance of temporal extrapolations from correlative niche models (Vallecillo et al. 2009), because there is no way to specify or control the interaction between the processes implicitly included. In process-based models biological processes are included explicitly (Kearney & Porter 2009). This means that an increase of the number of external inputs can be handled appropriately if knowledge is available on the functional responses that relate all inputs with species performance. However, interactions between even a few processes can make such functions very complex (e.g., Rayfield et al. 2009). Therefore, the pathway of increasing the number of processes modelled externally may be either limited by the knowledge available or condemned to poor performance.
4.7.1 Handling increased complexity 15
Matching model complexity with biological objectives at hand is a major challenge in future development of biodiversity and ecosystem models. We describe here three general strategies that should help limiting model complexity.
4.7.1.1 Formulating critical biological processes
20
Avoiding unnecessary increases in model complexity requires a careful assessment of the biological processes that most directly affect species distributions at the spatial and temporal scales of interest for each particular study (Guisan & Thuiller 2005). Although there is no general recipe to select the relevant biological processes, processes related to species autoecology will always have a central role. Habitat selection and population dynamics in species level models may be formulated with more or less detail, but are fundamentally important to predict species distribution dynamics (Willis et al. 2009; Kunstler et al. 2011).
25
30
35
40
Moreover, movement has been recently identified as one of the most important biological processes because it allows the modelling of emergent spatial dynamics and constrains species distributions (Soberón & Nakamura 2009). However, movement is a function not only of individual species’ characteristics, but also the structure of the landscape (or seascape) over which dispersal is occurring, including the presence of natural barriers or fragmentation of habitats (Nathan & Muller-Landau 2000; Schurr et al. 2007). At the community level, biotic interactions have also been identified as important but their importance varies across species (generalist species are less dependent upon interactions with other species than specialists) and scales (interactions are more likely to be determinant at finer scales). Despite its importance, biotic interactions are likely to constitute one of the most difficult challenges faced in biodiversity models. There are some examples that have explicitly simulated the effects of competition in the distributions of individual species (e.g. Lischke et al. 2006; Albert et al. 2008), but complex sets of interactions among species are seldom considered in modelling exercises (Araújo et al. 2011; Guisan & Rahbek 2011). Finally, plasticity and evolutionary processes are rarely considered in the context of species distribution dynamics but they may result critical for species whose short generation time allows them to rapidly adapt to environmental changes (e.g., Kearney et al. 2009).
Page 35 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
4.7.1.2 Hierarchical modular modelling approaches
5
10
Biological processes should be formulated explicitly and internally (i.e. process based models) only when they are critical for the questions at hand. The remaining processes can be modelled externally and formulated into the model by means of input spatial layers or parameters modified by additional modelling frameworks (Smith et al. 2001). Such an approach may facilitate the structure of species distribution models by allowing sub-models to be plugged into one another (e.g., McRae et al. 2008). In this modular structure, the upper levels provide external contextual information (and hence external dynamics) to the lower ones. Hierarchical modular structures have the advantage of: (1) more easily integrate across different spatial and temporal scales (e.g. to downscale the results of processes formulated at higher levels) (e.g., Del Barrio et al. 2006); and (2) assess the levels of uncertainty added at each stage (Larson et al. 2004, Chisholm & Wintle 2007). However, modularity may be limited for those target species that modify their environment or interact with other biotic entities (Midgley et al. 2010).
15
4.7.1.3 Comparing models with different levels of complexity
20
Research is needed to compare the outputs of models with different degrees of complexity in light of validation data appropriate for the process or driver under study (Roura-Pascual et al. 2010). Only in this case will it be possible to build a body of reference regarding the minimum acceptable levels of complexity to analyze a given problem.
4.7.1.4 Model communication: making use of standard protocols
25
30
35
40
In the quest for more reliable predictions, SDM applications are moving beyond simple correlative models towards a wider range of process-based modeling strategies, thus converging with developments in other modeling disciplines such as population and metapopulation dynamics. This trend will likely complicate the interpretation and communication of model structure and functionality in a similar way to the problem identified more generally in other fields such as agent based modeling (Grimm et al. 2006, Grimm et al. 2010). One way to better communicate the origin of dynamics in SDMs would be the general application of standard protocols for model communication such as the ODD (Overview, Design concepts, and Detail, Grimm et al. 2006). The ODD protocol forces the modeler to be explicit in the critical issues we have identified in our conceptual review. First, the Overview section of the ODD protocol is a conceptual description of the model that includes (1) the specific purpose of the model; (2) what are basic biological entities and state variables used in model formulation; and (3) which are the processes explicitly included in the model. In line with the ODD protocol, the conceptual description of a SDM should include the biological question in hand (section 1 in Overview), a detailed description of target state variable and drivers included (section 2 in Overview) and the biological processes of interest (section 3 in Overview). The second main component in the ODD protocol tackles the design concepts defining the model formulation. In this section emergence is especially important allowing the identification of which system-level phenomena truly emerge from the explicitly formulated processes (emergent dynamics) and which are merely imposed from the inputs (imposed dynamics). When considering specific models from scientific applications, it will be important to see whether it is possible to tease apart (i) what was explicitly included in the model vs. Page 36 of 48
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
what was imported from other sources and how; and (ii) what was actually modelled in the model vs. what was considered by the authors to be modelled, and what is the link. Finally, there is a need for approaches that combine the rigor of the small-scale studies precision with the breadth of broad-scale assessments (Naidoo and Ricketts 2006; Egoh et al. 2008; and Nelson et al. 2008 for some initial attempts). Spatially explicit values of biodiversity across landscapes that might inform land-use and management decisions are still lacking. Without quantitative assessments, and some incentives for landowners to preserve biodiversity, this tends to be ignored by land-management decisions. Similarly, rights-based fisheries have shown to have positive impacts on management of marine resources (Costello et al. 2008). Without information on the impacts of land-use management practices on biodiversity, it is impossible to design policies or payment programs that will provide the desired ecosystem services.
4.7.2 Conclusions 15
20
25
30
35
40
Modelling is the only option to policy makers to assess the implications of different drivers on biodiversity and ecosystem functioning, especially in the future context. The best examples are from climate change, where Global Circulation Models and Earth System Models help understand the role of greenhouse gas emission driver on future climate. Similarly DGVMs assist policy makers on the impact of projected climate change driver on terrestrial ecosystems and biodiversity. Modelling complex systems is characterized by inherent limitations since all processes and drivers and their linkage to biodiversity or ecosystem functions cannot be quantified and included. Further no single model that exists today covers all potential biophysical and socio-economic drivers, their inter-relationships and feedbacks. But policy-making requires robust model projections of impacts of drivers on biodiversity and ecosystem functioning. All the limitations and multiplicity of models was presented in this chapter. In the coming decades drivers such as climate change, land use and pollution will become more important. Thus scientific community has to develop strategies to address the limitations of current models and reduce uncertainty involved. Some potential examples of strategies include the following: • Formation of Model inter-comparison group, similar to CMIP - Coupled Model Intercomparison Project Phase 5 (http://cmip-pcmdi.llnl.gov/cmip5/) or ISIMIP Inter-Sectoral Impact Model Intercomparison Project (https://www.pikpotsdam.de/research/climate-impacts-and-vulnerabilities/research/rd2-crosscutting-activities/isi-mip) projects, involving a large number of modelling groups working on biodiversity and ecosystem modelling • Development of Integrated models that can be applied at sea/landscape or ecosystem level to assess the impact of drivers on biodiversity and ecosystem functioning. These integrated models should consider both bio-physical and socio-economic drivers at scales relevant to decision making. • Apply multiple models, and get a range of outputs to policy makers, rather than giving single model outputs, since there are multiple future pathways. • Need to generate data for different ecosystems at multiple locations and multiple scales by forming research networks.
Page 37 of 48
First Order Draft
•
• 5 •
IPBES Deliverable 3c
DO NOT CITE
Initiate long term monitoring studies to assess the response of ecosystems to changing drivers such as climate change, pollution and hydrological changes (http://unesdoc.unesco.org/images/0009/000938/093876eo.pdf). Develop protocols for modelling drivers impacting biodiversity and ecosystem functions. Encourage more collaboration between modelling groups and field ecologists.
Page 38 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
References
5
10
15
20
25
30
35
40
45
Ackerman, F., Stanton, E.A., Bueno, R (2010) Fat tails, exponents, extreme uncertainty: Simulating catastrophe in DICE. Ecological Economics 69 (2010) 1657–1665 Akçakaya, H., Radeloff, V. & Mladenoff, D. (2004). Integrating landscape and metapopulation modeling approaches: viability of the sharp-tailed grouse in a dynamic landscape. Conservation Biology, 18:526-537. Alkemade R, et al. (2009) GLOBIO3: A framework to investigate options for reducing global terrestrial biodiversity loss. Ecosystems (NY) 12:374-390. Araújo, M. B., & New, M. (2007). Ensemble forecasting of species distributions. Trends in ecology & evolution, 22(1), 42-47. Araújo, M. B., & Peterson, A. T. (2012). Uses and misuses of bioclimatic envelope modeling. Ecology, 93(7), 1527-1539. Araújo, M. B., Alagador, D., Cabeza, M., Nogués‐Bravo, D., & Thuiller, W. (2011). Climate change threatens European conservation areas. Ecology letters, 14(5), 484-492. Araújo, M. B., and R.G. Pearson. (2005). Equilibrium of species' distributions with climate. Ecography 28, 693-695. Arrhenius, O. (1921). Species and area. Journal of Ecology, 9, 95–99. Ayres, R.U. 1989. Industrial metabolism and global change. Int. Soc. Sci. J. 121, 363-73. Baird D.J., Rubach M.N. & Van den Brink P.J. (2008), Trait-based ecological risk assessment (TERA): the new frontier? Integrated Environmental Assessment and Management. 4, 2–3. Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W. & Courchamp, F. (2012). Impacts of climate change on the future of biodiversity. Ecol Lett, 15, 365–377. Berkeley, S. A., Hixon, M. A., Larson, R. J., & Love, M. S. (2004). Fisheries sustainability via protection of age structure and spatial distribution of fish populations. Fisheries, 29(8), 23-32. Bonan GB, Levis S, Sitch S, Vertenstein M, Oleson KW (2003) A dynamic global vegetation model for use with climate models: concepts and description of simulated vegetation dynamics, Global Change Biol. 9:1543–1566. Botkin DB (1993) Forest dynamics: An Ecological Model. Oxford University Press. Oxford. Bouwman AF (eds. Kram T, Klein Goldewijk K), 2006, Integrated Modelling of Global Environmental Change. An Overview of IMAGE 2.4, Report no.500110002, ISBN-9069601516. Brotons, L., Aquilué, N., De Cáceres, M., Fortin, M.J. & Fall, A. (2013). How fire history, fire suppression practices and climate change affect wildfire regimes in Mediterranean landscapes. Plos One. Bugmann H (1994) On the Ecology of Mountainous Forests in a Changing Climate: A Simulation Study Ph.D. Thesis No. 10'638. Swiss Federal Institute of Technology, Zurich (ETHZ), Switzerland 258 pp. Cardinale, B. J., Duffy, J. E., Gonzalez, A., Hooper, D. U., Perrings, C., Venail, P., et al. (2012). Biodiversity loss and its impact on humanity. Nature, 486(7401), 59-67. Cardinale, B. J., Wright, J. P., Cadotte, M. W., Carroll, I. T., Hector, A., Srivastava, D. S., et al. (2007). Impacts of plant diversity on biomass production increase through time because of species complementarity. Proceedings of the National Academy of Sciences, 104(46), 18123-18128. Cariboni J., Galtelli D., Liska R., Saltelli A., (2007). The role of sensitivity analysis in ecological modelling. Ecological Modelling. 203, 167-182. Chapin, F. S., Zavaleta, E. S., Eviner, V. T., Naylor, R. L., Vitousek, P. M., Reynolds, H. L., et al. (2000). Consequences of changing biodiversity. Nature, 405 (6783), 234–242. Chaturvedi RK, Jaideep Joshi, Mathangi Jayaraman, G. Bala, N.H. Ravindranath (2012). Multi-model climate change projections for India under Representative Concentration Pathways. Current Science, 103(7):791-802 Cheung, W. W., Sarmiento, J. L., Dunne, J., Frölicher, T. L., Lam, V. W., Palomares, M. D., ... & Pauly, D. (2013). Shrinking of fishes exacerbates impacts of global ocean changes on marine ecosystems. Nature Climate Change, 3(3), 254-258. Page 39 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Cheung, W.W.L., Lam, V.W., Pauly, D., 2008. Dynamic bioclimate envelope model to predict climateinduced changes in distribution of marine fishes and invertebrates. In: Cheung, W.W.L., Lam, V.W.Y., Pauly, D. (Eds.), Modeling Present and Climate-shifted Distribution of Marine Fishes and Invertebrates, vol. 163. Fisheries Centre, University of British Columbia: Fisheries Centre Research Reports, pp. 5–45. Cheung, W.W.L., Lam, V.W.Y., Sarmiento, J.L., Kearney, K., Watson, R. & Pauly, D. (2009). Projecting global marine biodiversity impacts under climate change scenarios. Fish and Fisheries, 10, 235– 251. Christensen, V., & Pauly, D. (1992). ECOPATH II --- a software for balancing steady-state ecosystem models and calculating network characteristics. Ecological Modelling, 61(3-4), 169–185. doi:10.1016/0304-3800(92)90016-8 Christensen, V., & Walters, C. J. (2004). Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling, 172(2-4), 109–139. Chun J. H. and C. B. Lee. (2013). Assessing the Effects of Climate Change on the Geographic Distribution of Pinus densiflora in Korea using Ecological Niche Model. Korean Journal of Agricultural and Forest Meteorology, 15(4), 219-233. Clark WC (2010) Sustainable development and sustainability science. In: Levin, S. A., Clark, W. C. (Eds), Toward a science of sustainability: Report from Toward a Science of Sustainability Conference. Princeton: Princeton University Press. Collie, J. S., Botsford, L. W., Hastings, A., Kaplan, I. C., Largier, J. L., Livingston, P. A., et al. (2014). Ecosystem models for fisheries management: finding the sweet spot. Fish and Fisheries. Concepción, E. D., & Díaz, M. (2011). Field, landscape and regional effects of farmland management on specialist open-land birds: Does body size matter?. Agriculture, Ecosystems & Environment, 142(3), 303-310. Costello, C., Gaines, S.D., Lynham, J., 2008. Can Catch Shares Prevent Fisheries Collapse? Science 321, 1678–1681. doi:10.1126/science.1159478 Cramer et al (2001) Response of terrestrial ecosystem structure and function to CO2 and climate change: results from six dynamic global vegetation models, Global Change Biology 7:357-373. Dale, V. H., & Beyeler, S. C. (2001). Challenges in the development and use of ecological indicators. Ecological indicators, 1(1), 3-10. Davis, A. J., Jenkinson, L. S., Lawton, J. H., Shorrocks, B., & Wood, S. (1998). Making mistakes when predicting shifts in species range in response to global warming. Nature, 391(6669), 783-786. de Chazal, J. & Rounsevell, M.D.A. (2009). Land-use and climate change within assessments of biodiversity change: A review. Global Environmental Change, 19, 306–315. Dengler, J. (2009). Which function describes the species–area relationship best? A review and empirical evaluation. Journal of Biogeography, 36(4), 728-744. Desrochers, R. E., Kerr, J. T., & Currie, D. J. (2011). How, and how much, natural cover loss increases species richness. Global Ecology and Biogeography, 20(6), 857-867. Detlef P. van Vuuren, Jason Lowe, Elke Stehfest, Laila Gohar, Andries F. Hof, Chris Hope, Rachel Warren, Malte Meinshausen, Gian-Kasper Plattner, 2009, How well do integrated assessment models simulate climate change?, Springer, Climate change, DOI 10.1007/s10584-009-9764-2 Dixon RK and Wisniewski J (1995) Global forest systems: An uncertain response to atmospheric pollutants and global climate change?, Water, Air, Soil Pollution 85:1:101–110. Dormann, C.F., Schymanski, S.J., Cabral, J., Chuine, I., Graham, C., Hartig, F., et al. (2012). Correlation and process in species distribution models: bridging a dichotomy. J. Biogeogr., 39, 2119–2131. Drakare, S., Lennon, J. J., & Hillebrand, H. (2006). The imprint of the geographical, evolutionary and ecological context on species–area relationships. Ecology letters, 9(2), 215-227. Egoh B, Reyers B, Rouget M, et al. 2008. Mapping ecosystem service for planning and management. Agri Ecosyst Environ 127: 135–40. Page 40 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Eldridge, D. J., & Freudenberger, D. (2005). Ecosystem wicks: woodland trees enhance water infiltration in a fragmented agricultural landscape in eastern Australia. Austral Ecology, 30(3), 336-347. Elith, J., & Leathwick, J. R. (2009). Species distribution models: ecological explanation and prediction across space and time. Annual Review of Ecology, Evolution, and Systematics, 40(1), 677. Elith, J., Kearney, M., & Phillips, S. (2010). The art of modelling range-shifting species. Methods in Ecology and Evolution, 1, 330-342. Elith, J., Phillips, S. J., Hastie, T., Dudík, M., Chee, Y. E., & Yates, C. J. (2011). A statistical explanation of MaxEnt for ecologists. Diversity and Distributions, 17(1), 43-57. Ewers, R.M., Kapos, V., Coomes, D.A., Lafortezza, R., Didham, R.K., 2009. Mapping community change in modified landscapes. Biological Conservation 142, 2800-2872. Faith DP, Ferrier S, Williams KJ. 2008. Getting biodiversity intactness indices right: ensuring that ‘biodiversity’ reflects‘diversity’. Glob Chang Biol 14:207-21. Ferrier, S., Manion, G., Elith, J., Richardson, K., 2007. Using generalized dissimilarity modelling to analyse and predict patterns of beta diversity in regional biodiversity assessment. Diversity and Distributions 13, 252-264. Ferrier, S., Watson, G., Pearce, J., Drielsma, M., 2002. Extended statistical approaches to modelling spatial pattern in biodiversity in northeast New South Wales. I.Species-level modelling. Biodiversity and Conservation 11, 2275-2307. Ferrier,S., Guisan,A. 2006. Spatial modelling of biodiversity at the community level. Journal of Applied Ecology 43, 393-404. Ferrier,S.(2002) Mapping spatial pattern in biodiversity for regional conservation planning: where to from here? Systematic Biology, 51, 331-363. Fischlin A, Midgley GF, Price JT, Leemans R, Gopal B, Turley C, Rounsevell MDA, Dube OP, Tarazona J and Velichko AA (2007) Ecosystems, their properties, goods, and services. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change.(eds.Parry ML, Canziani OF, Palutikof JP, van der Linden PJ and Hanson CE, Cambridge University Press, Cambridge, 211272. Foley AJ, Levis S, Costa MH, Cramer W and Pollard D (2000) Incorporating dynamic vegetation cover within global climate models ecological applications, Ecological Society of America 10(6):16201632. Fordham, D. A., Akçakaya, H. R., Brook, B. W., Rodríguez, A., Alves, P. C., Civantos, E. et al. (2013). Adapted conservation measures are required to save the Iberian lynx in a changing climate. Nature Climate Change, 3(10), 899-903 Fulton, E.A., Fuller, M., Smith, A.D.M., Punt, A.E. 2005. Ecological indicators of the ecosystem effects of fishing: final report. Australian Fisheries Management Authority Report, R99/1546, 239 pp. Gaikwad, J., Wilson, P. D., & Ranganathan, S. (2011). Ecological niche modeling of customary medicinal plant species used by Australian Aborigines to identify species-rich and culturally valuable areas for conservation. Ecological Modelling, 222(18), 3437-3443. Gal, G., V. Makler-Pick, and N. Shachara (2012). Dealing with uncertainty in ecosystem model scenarios: application of single model ensemble approach. International Environmental Modelling and Software Society (iEMSs) 2012 International Congress on Environmental Modelling and Software. http://www.iemss.org/society/index.php/iemss 2012 proceedings Gallien, L., Münkemüller, T., Albert, C.H., Boulangeat, I. & Thuiller, W. (2010). Predicting potential distributions of invasive species: where to go from here? Diversity and Distributions, 16, 331– 342. Garcia, S. M., Kolding, J., Rice, J., Rochet, M. J., Zhou, S., Arimoto, T., ... & Smith, A. D. M. (2012). Reconsidering the consequences of selective fisheries. Science, 335(6072), 1045-1047
Page 41 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Gaston, K.J., Davies, R.G., Orme, C.D.L., Olson, V.A., Thomas, G.H., Tzung-Su, D.,Rasmussen, P.C., Lennon, J.J., Bennett, P.M., Owens, I.P.F., Blackburn, T.M., 2007.Spatial turnover in the global avifauna. Proceedings of the Royal Society of London Series B 274, 1567-1574. Goodess CM, Hanson C, Hulme M, Osborn TJ (2003) Representing climate and extreme weather events in integrated assessment models: a review of existing methods and options for development. Integr Assess 4:145–171 Graedel, T.E. and Crutzen, P.J. 1990. Atmospheric trace constituents. In: The Earth as Transformedbv HumanAction. Turner, B.L., II, Clark, W.C., Kates, R.W., Richards, J.F., Mathews, J.T. and Meyer, W.B. (eds). Cambridge University Press, Cambridge, p. 295-311. Guilhaumon, F., Gimenez, O., Gaston, K. J., & Mouillot, D. (2008). Taxonomic and regional uncertainty in species–area relationships and the identification of richness hotspots. Proceedings of the National Academy of Sciences, 105, 15458–15463. Guisan, A., Tingley, R., Baumgartner, J.B., Naujokaitis-Lewis, I., Sutcliffe, P.R., Tulloch, A.I.T., et al. (2013). Predicting species distributions for conservation decisions. Ecol Lett, 16, 1424–1435. Guisan, A.,Zimmermann,N.E. (2000). Predictive habitat distribution models in ecology. EcologicalModelling. 135, 147-186. Gustafson, E. J. (2013). When relationships estimated in the past cannot be used to predict the future: using mechanistic models to predict landscape ecological dynamics in a changing world. Landscape ecology, 28(8), 1429-1437. Halpern, B. S., Walbridge, S., Selkoe, K. A., Kappel, C. V., Micheli, F., D'Agrosa, C., et al. (2008). A global map of human impact on marine ecosystems. Science, 319(5865), 948–952. Handa, I. T., Aerts, R., Berendse, F., Berg, M. P., Bruder, A., Butenschoen, O., et al. (2014). Consequences of biodiversity loss for litter decomposition across biomes. Nature, 509(7499), 218-221. Hanski, I., Zurita, G. A., Bellocq, M. I., & Rybicki, J. (2013). Species–fragmented area relationship. Proceedings of the National Academy of Sciences, 110(31), 12715-12720. Harfoot, M. B., Newbold, T., Tittensor, D. P., Emmott, S., Hutton, J., Lyutsarev, V., ... & Purves, D. W. (2014). Emergent Global Patterns of Ecosystem Structure and Function from a Mechanistic General Ecosystem Model. PLoS biology, 12(4), e1001841. Harremoes P, Turner RK (2001) Methods for integrated assessment. Reg Environ Change 2:57–65 Haxeltine A and Prentice IC (1996) BIOME3, an equilibrium terrestrial biosphere model based on ecophysiological constraints, resource availability and competition among plant functional types, Global Biogeochemical Cycles 10.4:693-710. Hector et al. 1999. Helton, J.C., Johnson, J.D., Sallaberry, C., Storlie, C.B. (2006). Survey of sampling-based methods for uncertainty and sensitivity analysis. Reliability Engineering & System Safety, 91, 1175-1209. Hickler, T., Vohland, K., Feehan, J., Miller, P. A., Smith, B., Costa, L., ... & Sykes, M. T. (2012). Projecting the future distribution of European potential natural vegetation zones with a generalized, tree species‐based dynamic vegetation model. Global Ecology and Biogeography, 21(1), 50-63. Hilbert,D.W.& Ostendorf,B(2001) The utility of artificial neural networks for modelling the distribution of vegetation in past, present and future climates. Ecological Modelling,146, 311-327. Hope C (2013) Critical issues for the calculation of the social cost of CO2: why the estimates from PAGE09 are higher than those from PAGE2002. Climatic Change. 117:531-543 Iglecia, M. N., J. A. Collazo, and A. J. McKerrow. 2012. Use of occupancy models to valuate expert and knowledge-based species-habitat relationships. Avian Conservation and Ecology. 7(2), 5. http://dx.doi.org/10.5751/ACE-00551-070205. Iverson, L. and Anantha M. Prasad (2001). Potential Changes in Tree Species Richness and Forest Community Types following Climate Change. Ecosystems, 4, 186-199. Jacobs, S., Burkhard, B., Van Daele, T., Staes, J. & Schneiders, A. (2015). Ecological Modelling. Ecol Model, 295, 21–30. Page 42 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Jager, T., & Klok, C. (2010). Extrapolating toxic effects on individuals to the population level: the role of dynamic energy budgets. Philosophical Transactions of the Royal Society B: Biological Sciences, 365(1557), 3531-3540. Janssen, J.a.E.B., Krol, M.S., Schielen, R.M.J., Hoekstra, A.Y. & De Kok, J.L. (2010) Assessment of uncertainties in expert knowledge, illustrated in fuzzy rule-based models. Ecological Modelling, 221, 1245-1251. Jennings, S., Mélin, F., Blanchard, J.L., Forster, R.M., Dulvy, N.K., Wilson, R.W., 2008. Global-scale predictions of community and ecosystem properties from simple ecological theory. Proceedings of the Royal Society of London B 275, 1375–1383. Jennings, S., Pinnegar, J.K., Polunin, N.V.C., Warr, K.J., 2002. Linking size-based and trophic analyses of benthic community structure. Mar Ecol Prog Ser 226, 77–85. Jørgensen, S. E., 1994. Fundamentals of Ecological Modelling. Developments in Environmental Modelling, 2nd edition, vol. 19. Elsevier, Amsterdam Kaplan JO et al. (2003) Climate change and Arctic Ecosystems: Modeling, Paleodata-Model comparisons, and future projections, Journal of Geophysical Research-Atmosphere 108:8171. Neilson RP (1995) A model for predicting continental-scale vegetation distribution and water balance, Ecological Applications 5:362-385. Kates, R.W., Tumer, B.L., II and Clark, W.C. 1990. The great transformation. In: The Earth as Transformed by Human Action. Turner, B.L., II, Clark, W.C., Kates, R.W., Richards, J.F., Mathews, J.T. and Meyer, W.B. (eds). Cambridge University Press, Cambridge, p. 1-17. Kearney, M., Simpson, S.J., Raubenheimer, D. & Helmuth, B. (2010a). Modelling the ecological niche from functional traits. Philosophical Transactions of the Royal Society B: Biological Sciences, 365, 3469–3483. Kearney, M.R., Wintle, B.A. & Porter, W.P. (2010b). Correlative and mechanistic models of species distribution provide congruent forecasts under climate change. Conservation Letters, 3, 203–213. Koleff, P., Gaston, K.J. & Lennon, J.K. (2003) Measuring beta diversity for presence-absence data. Journal of Animal Ecology, 72, 367-382. Kuhnert, P.M., Martin, T.G. & Griffiths, S.P. (2010) A guide to eliciting and using expert knowledge in Bayesian ecological models. Ecology Letters, 13, 900–914. Lavergne, S., Mouquet, N., Thuiller, W. & Ronce, O. (2010). Biodiversity and Climate Change: Integrating Evolutionary and Ecological Responses of Species and Communities. Annual Review of Ecology, Evolution, and Systematics, 41, 321–350. Legendre, P., Lapointe, F.-J. & Casgrain, P. (1994) Modeling brain evolution from behavior: a permutational regression approach. Evolution, 48, 1487-1499. Lenhardt, P. P., Schäfer, R. B., Theissinger, K., & Brühl, C. A. (2013). An expert-based landscape permeability model for assessing the impact of agricultural management on amphibian migration. Basic and Applied Ecology, 14(5), 442-451 Lennon, J.J., Koleff, P., Greenwood, J.J.D., Gaston, K.J., 2001. The geographical structure of British bird distributions: diversity, spatial turnover and scale. Journal of Animal Ecology 70, 966-979. Leung, B., Roura-Pascual, N., Bacher, S., Heikkilä, J., Brotons, L., Burgman, M.A., et al. (2012). TEASIng apart alien species risk assessments: a framework for best practices. Ecol Lett, 15, 1475–1493. Link, J. S., Ihde, T. F., Harvey, C. J., Gaichas, S. K., Field, J. C., Brodziak, J. K. T., et al. (2012). Dealing with uncertainty in ecosystem models: The paradox of use for living marine resource management. Progress in Oceanography, 102, 102–114. doi:10.1016/j.pocean.2012.03.008 Loh J, Green RE, Ricketts T, Lamoreux J, Jenkins M, Kapos V,Randers J. 2005. The Living Plant Index: using species population time series to track trends in biodiversity. Phil Trans R Soc B 360:289-295. Loreau M., Mouquet N., Gonzalez A. (2003) Biodiversity as spatial insurance in heterogeneous landscapes. PNAS 100(22): 12765–12770.
Page 43 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Luoto, M.,Poyry,J.,Heikkinen,R.K.,Saarinen,K., (2005). Uncertainty of bioclimate envelope models based on the geographical distribution of species. GlobalEcology and Biogeography 14, 575–584. Lurgi, M., Brook, B.W., Saltré, F. & Fordham, D.A. (2014). Modelling range dynamics under global change: which framework and why? Methods in Ecology and Evolution, in press. Mace, G. M., Norris, K., & Fitter, A. H. (2012). Biodiversity and ecosystem services: a multilayered relationship. Trends in ecology & evolution, 27(1), 19-26 Majer JD, Beeston G. 1996. The Biodiversity Integrity Index: an illustration using ants in Western Australia. Conserv Biol 10:65-73. Malcolm, J. R., Liu, C., Neilson, R. P., Hansen, L., & Hannah, L. E. E. (2006). Global warming and extinctions of endemic species from biodiversity hotspots. Conservation biology, 20(2), 538-548. Marten, A.L., Robert E. Kopp., Kate C Shouse., Charles W. Griffiths., Elke L. Hodson., Elizabeth Kopits., Bryan K. Mignone& Chris Moore & Steve C. Newbold & Stephanie Waldhoff & Ann Wolverton (2013) Improving the assessment and valuation of climate change impacts for policy and regulatory analysis. Climatic Change. 117:433–438 Martin, T.G., Burgman, M. a., Fidler, F., Kuhnert, P.M., Low-Choy, S., McBride, M.F. & Mengersen, K. (2012) Eliciting expert knowledge in conservation science. Conservation biology, 26, 29–38. Maury, O. & J.C. Poggiale (2013) From individuals to populations to communities: A dynamic energy budget model of marine ecosystem size-spectrum including life history diversity, Journal of Theoretical Biology, 324, 52-71 McElhany, P., E.A. Steel, D. Jensen, K. Avery, N. Yoder, C. Busack and B. Thompson. (2010). Dealing with uncertainty in ecosystem models: lessons from a complex salmon model. Ecological Applications. 20, 465-482. McGuire AD, Melillo JM, Joyce LA, Kicklighter DW, Grace AL, Moore III B and Vorosmarty CJ (1992) Interactions between carbon and nitrogen dynamics in estimating net primary productivity for potential vegetation in North America, Global Biogeochemical Cycles 6:101-124. McGuire AD, Sitch S, Clein JS, Dargaville R, Esser G et al (2001) carbon balance of the terrestrial biosphere in the twentieth century: analyses of CO2, climate and land-use effects with four process-based ecosystem models, Global Biogeochemical Cycles 15:183-206. McInerny, G.J. & Etienne, R.S. (2012). Ditch the niche - is the niche a useful concept in ecology or species distribution modelling? J. Biogeogr., 39, 2096–2102. McMahon, S.M., Harrison, S.P., Armbruster, W.S., Bartlein, P.J., Beale, C.M., Edwards, M.E., et al. (2011). Improving assessment and modelling of climate change impacts on global terrestrial biodiversity. TREE, 26, 249–259. Merow, C., Smith, M. J., & Silander, J. A. (2013). A practical guide to MaxEnt for modeling species’ distributions: what it does, and why inputs and settings matter. Ecography, 36(10), 1058-1069. Midgley, G., Davies, I., Albert, C., Altwegg, R. & et al. (2009). BioMove an integrated platform simulating the dynamic response of species to environmental change. Ecography, 33:1-5. Molnár, P. K., Derocher, A. E., Klanjscek, T., & Lewis, M. A. (2011). Predicting climate change impacts on polar bear litter size. Nature Communications, 2, 186. Morin, X., & Lechowicz, M. J. (2008). Contemporary perspectives on the niche that can improve models of species range shifts under climate change. Biology Letters, 4(5), 573-576. Morin, X., & Thuiller, W. (2009). Comparing niche-and process-based models to reduce prediction uncertainty in species range shifts under climate change. Ecology, 90(5), 1301-1313. Mouton, A.M., De Baets, B., Goethals, P.L.M. (2009). Knowledge-based versus data-driven fuzzy habitat suitability models for river management. Environmental Modelling & Software, 24, 982-993. Muñoz, M.E.S. Giovanni, R., Siqueira, M.F., Sutton, T., Brewer, P., Pereira, R.S., Canhos, D.A.L. & Canhos, V.P. (2011). "openModeller: a generic approach to species' potential distribution modelling". GeoInformatica. 15, 111–135.
Page 44 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Murphy, J.M, D M. H. Sexton, D N. Barnett, G S. Jones, MJ. Webb, M Collins & D A. Stainforth, 2004, Quantification of modeling uncertainties in a large ensemble of climate change simulations, Nature 430, 768-772 (12 August 2004) Murray, J. V., Goldizen, A.W., O’Leary, R.A., McAlpine, C. a., Possingham, H.P., Choy, S.L. & O’Leary, R. a. (2009) How useful is expert opinion for predicting the distribution of a species within and beyond the region of expertise? A case study using brush-tailed rock-wallabies Petrogale penicillata. Journal of Applied Ecology, 46, 842–851. Naidoo R and Ricketts TH. 2006. Mapping the economic costs and benefits of conservation. PLoS Biol 4: 2153–64 Niamir, A., Skidmore, A.K., Toxopeus, A.G., Muñoz, A.R.M. & Real, R. (2011) Finessing atlas data for species distribution models. Diversity and Distributions. 17, 1173–1185. Noss, R. F. (1990). Indicators for monitoring biodiversity: a hierarchical approach. Conservation biology, 4(4), 355-364. Parrott, L. (2011). Hybrid modelling of complex ecological systems for decision support: Recent successes and future perspectives. Ecological Informatics, 6, 44–49. Parton WJ, Scurlock JMO, Ojima DS, Gilmanov TG, Scholes RJ, Schimel DS, Kirchner T, Menaut J-C, Seastedt T, Moya EG, Kamnalrut A, Kinyamario JL (1993) Observations and modeling of biomass and soil organic matter dynamics for the grassland biome worldwide, Global Biogeochemical Cycles 7:785-809. Pausas, J.G. & Keeley, J.E. (2009). A Burning Story: The Role of Fire in the History of Life. BioScience, 59, 593–601. Payne, K. and Stockwell, D. (2006). GARP Modelling System User's Guide and Technical Reference, http://landshape.org/enm/garp-modelling-system-users-guide-and-technical-reference/ Pearson, R.G. and T.P. Dawson. (2003). Predicting the impacts of climate change on the distribution of species: Are bioclimate envelope models useful? Global Ecology and Biogeography, 12, 361-371. Pearson, R.G. C.J. Raxworthy, M. Nakamura, and A.T. Peterson. (2007). Predicting species' distributions from small numbers of occurrence records: A test case using cryptic geckos in Madagascar. Journal of Biogeography 34, 102-117. Pearson, R.G. T.P. Dawson, P.M. Berry, and P.A. Harrison. (2002). Species: A spatial evaluation of climate impact on the envelope of species. Ecological Modelling 154, 289-300. Peppler-Lisbach, C. & Schröder, B. (2004) Predicting the species composition of Nardus stricta communities by logistic regression modelling. Journal of Vegetation Science, 15, 623-634. Pereira, H. M., & Daily, G. C. (2006). Modeling biodiversity dynamics in countryside landscapes. Ecology, 87, 1877–1885 Pereira, H.M., Leadley, P.W., Proenca, V., Alkemade, R., Scharlemann, J.P.W., Fernandez-Manjarres, J.F., et al. (2010). Scenarios for Global Biodiversity in the 21st Century. Science, 330, 1496–1501. Pereira, H.M., Navarro, L.M. & Martins, I.S. (2012). Global Biodiversity Change: The Bad, the Good, and the Unknown. Annu. Rev. Environ. Resourc., 37, 25–50. Peterson G., Allen C.R., Holling C.S. (1998) Ecological resilience, biodiversity, and scale. Ecosystems 1: 6– 18 Peterson, A. T., & Vieglais, D. A. (2001). Predicting Species Invasions Using Ecological Niche Modeling: New Approaches from Bioinformatics Attack a Pressing Problem. BioScience, 51(5), 363-371. Peterson, A. T., Soberón, J., Pearson, R. G., Anderson, R. P., Martínez-Meyer, E., Nakamura, M., Araújo, M.B. (2011). Ecological niches and geographic distributions (MPB-49). Princeton University Press. Peterson, A.T. (2006). Uses and requirements of ecological niche models and related distributional models. Biodiversity Informatics, 3. 59-72. Peterson, A.T. M.A. Ortega-Huerta, J. Bartley, V. Sanchez-Cordero, J. Soberon, R.W. Buddemeier, and D.R.B. Stockwell. (2002). Future projections for mexican faunas under global climate change scenarios. Nature 416, 626-629. Page 45 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Phillips, S. J., & Dudík, M. (2008). Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography, 31(2), 161-175. Polovina, J. J. (1984). Model of a coral reef ecosystem. Coral Reefs, 3(1), 1–11. Prentice IC, Bondeau A, Cramer W, Harrison SP, Hickler T, Lucht W, Sitch S, Smith B and Sykes MT (2007) Dynamic Global Vegetation Modelling: Quantifying Terrestrial Ecosystem Responses to Largescale Environmental Change (eds. J.G. Canadell, D.E. Pataki and L.F. Pitelka) Springer-Verlag, Berlin, Terrestrial Ecosystems in a ChangingWorld, 175-192. Proença, V., & Pereira, H. M. (2013). Species–area models to assess biodiversity change in multi-habitat landscapes: The importance of species habitat affinity. Basic and Applied Ecology, 14(2), 102-114. Rahbek, C., Gotelli, N. J., Colwell, R. K., Entsminger, G. L., Rangel, T. F. L., & Graves, G. R. (2007). Predicting continental-scale patterns of bird species richness with spatially explicit models. Proceedings of the Royal Society B: Biological Sciences, 274(1607), 165-174 Rajiv K Chaturvedi, 2012, Climate change mitigation and adaptation in Indian forests, Lambert Academic Publishers, Germany ISBN: 978-3-659-13500-2. Rajiv K Chaturvedi, Ranjith Gopalakrishnan, Mathangi Jayaraman, Govindasamy Bala, NV Joshi, Raman Sukumar, NH Ravindranath, 2011, Impact of climate change on Indian forests: a dynamic vegetation modeling approach, Mitigation and adaptation strategies for global change, Vol 6, issue 2 Raxworthy, C. J., Martinez-Meyer, E., Horning, N., Nussbaum, R. A., Schneider, G. E., Ortega-Huerta, M. A., & Peterson, A. T. (2003). Predicting distributions of known and unknown reptile species in Madagascar. Nature, 426(6968), 837-841. Rocchini, D., Hortal, J., Lengyel, S., Lobo, J.M., Jimenez-Valverde, A., Ricotta, C., et al. (2011). Accounting for uncertainty when mapping species distributions: The need for maps of ignorance. Prog Phys Geogr, 35, 211–226. Rockström J, Steffen W, Noone K, Person A, Chapin FS, Lambin EF, et al. (2009) A safe operating space for humanity. Nature 461: 472-475. Rodrigues M, Oliveira A, Costa M, Fortunato AB, Y. Zhang (2009) Sensitivity analysis of an ecological model applied to the Ria de Aveiro. J Coastal Research. SI 56, 448-452. Rondinini C. Kerrie, A. Wilson, Luigi Boitani, Hedley Grantham and Hugh P. Possingham (2006). Tradeoffs of different types of species occurrence data for use in systematic conservation planning. Ecology Letters. 9, 1136-1145. Rosenzweig, M. L. (1995). Species diversity in space and time. Cambridge: Cambridge University Press. Running SW, Hunt Jr. ER (1993) Generalization of a Forest Ecosystem Process Model for Other Biomes, BIOME-BGC, and an Application for Global-scale Models, (eds. Ehleringer JR, Field C). Academic Press, San Diego, CaliforniaScaling processes between leaf and landscape levels.141-158. Salonius 1981 Sarmiento, J. L. S., and N. Gruber (2006), Ocean Biogeochemical Dynamics, 503 pp., Princton Univ. Press, Princeton, N. J. Scheller, R. M., J. B. Domingo, B. R. Sturtevant, J. S. Williams, A. Rudy, E. J. Gustafson, and D. J. Mladenoff. 2007. Design, development, and application of LANDIS-II, a spatial landscape simulation model with flexible spatial and temporal resolution. Ecological Modelling 201 (3-4): 409-419. Scholes RJ, Biggs R. 2005. A biodiversity intactness index. Nature 434:45-9. Shin, Y.J., Cury, P.M., 2004. Using an individual-based model of fish assemblages to study the response of size spectra to changes in fishing. Canadian Journal of Fisheries and Aquatic Sciences 61, 414– 431. Shugart HH (1984). A theory of forest dynamics: The Ecological Implication of Forest Succession Models. Springer-Verlag, New York.
Page 46 of 48
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Shugart HH, Urban DL (1989). Factors Affecting the Relative Abundances of Tree Species. Eds.Grubb PJ, Whittaker JB, Towards a more exact ecology. 30th Symposium of the British Ecological Society, Blackwell, Oxford -, 249-273. Sitch S (2000) The Role of Vegetation Dynamics in the Control of Atmospheric CO2 Content, PhD Thesis, Lund University. Allen CD (2009) Climate-induced forest dieback: an escalating global phenomenon?, Unasylva 60:231-232. Smith, J.U., & Smith, P. (2007). Environmental modelling. An introduction. Oxford University Press, Oxford Snowling, S., and Kramer, J. (2001). Evaluating modelling uncertainty for model selection. Ecological Modelling. 138, 17-30. Soetaert, K., & Herman, P. M. (Eds.). (2009). A practical guide to ecological modelling: using R as a simulation platform. Springer. Steffen W, Crutzen PJ, McNeill JR (2007) The anthropocene: Are humans now overwhelming the great forces of nature? Ambio 36(8): 614-621. Steffen W, Sanderson A, Tyson PD, Jäger J, Matson P, Moore III B, et al. (2004) Global change and the earth system: A planet under pressure. The IGBP global change series. Berlin, Germany: SpringerVerlag. Steffen WL, Walker BH, Ingram JSI and Koch GW (1992) Global change and terrestrial ecosystems: The operational plan. International Geosphere-Biosphere Programme. Stock, C.A. et al. (2011). On the use of IPCC-class models to assess the impact of climate on living marine resources. Progress in Oceanography, 88, 1–27. Stockwell, D.R.B. and I.R. Noble. (1992). Induction of sets of rules from animal distribution data: a robust and informative method of data analysis. Mathematics and Computers in Simulation, 33, 385390. Theisen, OM., Nils Petter Gleditsch, Halvard Buhaug (2013) Is climate change a driver of armed conflict? Climatic Change. 117:613-625 Thornton, P. E., & Zimmermann N. E. (2007). An Improved Canopy Integration Scheme for a Land Surface Model with Prognostic Canopy Structure. Journal of Climate. 20(15), 3902 - 3923. Thuiller, W., Münkemüller, T., Lavergne, S., Mouillot, D., Mouquet, N., Schiffers, K., et al. (2013). A road map for integrating eco-evolutionary processes into biodiversity models. Ecol Lett, 16, 94–105. Tikkanen, O. P., Punttila, P., & Heikkilä, R. (2009). Species–area relationships of red‐listed species in old boreal forests: a large‐scale data analysis. Diversity and Distributions, 15(5), 852-862. Tillman et al. 1996 Tittensor, D. P., Micheli, F., Nyström, M., & Worm, B. (2007). Human impacts on the species–area relationship in reef fish assemblages. Ecology Letters, 10(9), 760-772. Triantis, K. A., Guilhaumon, F., & Whittaker, R. J. (2012). The island species–area relationship: biology and statistics. Journal of Biogeography, 39(2), 215-231. UNEP. (2004) Decisions adopted by the conference of the parties to the convention on biological diversity at its seventh meeting (UNEP/CBD/COP/7/21/Part 2), decision VII/30 (CBD2004). http://www.biodiv.org/decisions/. Vallecillo S, Brotons L, Thuiller W. (2009) Dangers of predicting bird species distributions in response to land-cover changes: the role of dynamic processes. Ecological Applications 19(2): 538-549. van Vuuren, D. P., Sala, O. E., & Pereira, H. M. (2006). The future of vascular plant diversity under four global scenarios. Ecology and Society, 11, 25. Walker, W.E., Harremoes, P., Rotmans, J., Van der Sluijs, J., Van Asselt, M.,Janssen, P., Von Krauss, M.P.K. (2003). Defining uncertainty: a conceptual basis for uncertainty management in modelbased decision support. Integrated Assessment, 4, 5-17
Page 47 of 48
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
Walters, C., 1986. Adaptive management of renewable resources. McMillan, New York [Reprint edition available from Blackburn Press, 2001]. Walters, C., Pauly, D., Christensen, V., & Kitchell, J. F. (2000). Representing density dependent consequences of life history strategies in aquatic ecosystems: EcoSim II. Ecosystems, 3(1), 70–83. Walters, C., Christensen, V., Walters, W., & Rose, K. (2010). Representation of multi-stanza life histories in Ecospace models for spatial organization of ecosystem trophic interaction patterns. Bulletin of Marine Science, 86, 439–459. Warren R (2011) The role of interactions in a world implementing adaptation and mitigation solutions to climate change. Phil Trans Roy Soc A 369:217–241 Watling, L. (2005). The global destruction of bottom habitats by mobile fishing gear. In E. A. Norse & L. B. Crowder, Marine conservation biology the science of maintaining the seas biodiversity (pp. 198– 210). Washington, DC: 496 pp. Weitzman M. 2009. On Modeling and Interpreting the Economics of Catastrophic Climate Change. Review of Economics and Statistics, 91(1):1-19 Whiteman G, Chris Hope & Peter Wadhams. 2013. Climate science: Vast costs of Arctic change. Nature 499, 401–403 Wiersma, Y.F., Urban, D.L., 2005. Beta diversity and nature reserve system design in the Yukon, Canada. Conservation Biology 19, 1262-1272. Wohlgemuth, T. (1998) Modelling floristic species richness on a regional scale: a case study in Switzerland. Biodiversity and Conservation, 7, 159-177.
Page 48 of 48
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
5 Modelling consequences of change in biodiversity and ecosystems for nature’s benefits to people Coordinating Lead Authors: Garry Peterson (Sweden) & Jane Kabubo-Mariara (Kenya) Lead Authors: Jonathan Anticamara (Philippines), Neville Crossman (Australia), Ainars Aunins
(Latvia), Pablo Munoz (Chile); Brenda Rashleigh (USA); Makarius Victor Mdemu (Tanzania).
5
Contributing Authors: William Cheung (Canada), Adrian Vogl (USA)
Key messages: ●
Value of ecosystem service approach comes from considering bundles and tradeoffs among multiple ecosystem services. Assessments of an ecosystem service in isolation can be useful for specific contexts, but risk hiding important social conflicts & ecological interactions.
●
No single type of modeling approaches, tool, or processes is appropriate for all decision contexts, because the requirements of decision contexts vary and the strengths and weaknesses of tools vary. This chapter provides guidance on how to match tools and decision contexts.
●
Modelling the impact of ecological changes on human well-being is not well developed. Developing such tools will require investment and trans-disciplinary collaboration of policy makers, with natural and social scientists to develop new frameworks, methods, and tools. These types of tools are particularly needed to bridge different knowledge systems.
●
Applying multiple models to the same case produce more robust decisions because applying models with different strengths and weaknesses can provide a more complete picture and comparing model results can indicate where models provide variable or consistent results. This chapter explains how different types of models can effectively complement one another.
●
Developing a community of practice for ES modeling – standards, open data, and capacity building will improve the ability of people to use, develop, and advance ecosystem service modeling as well as encourage the development of new frameworks to assess the relationship between people and nature. Building this capacity is important for enhancing tool development from diverse knowledge systems.
10
15
20
25
30
5.1 Introduction 35
Modeling the impact of ecological changes on human well-being has rapidly progressed over the past decade. This chapter presents the types of dynamics that can be captured in these models with reference to the IPBES conceptual framework. It identifies and reviews the key components of this framework for these models, and then reviews the different assessment and decision contexts in which different types of Ecosystem Benefits and Values are assessed with regards to IPBES. This chapter then presents the reviews of major modeling frameworks for assessing ecosystem services, such as InVEST, Page 1 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Aries, and Ecosim, as well as alternative approaches. Moreover, this chapter shows the relationships between modeling approaches to the decision contexts in which ecosystem service models can be used. This chapter ends with an assessment of gaps, and recommendations for action and future research that would develop the capacity to better use and develop models in IPBES. 5
5.2 Conceptualizing how changes in biodiversity and ecosystem services influence the supply of benefits to people 10
15
Understanding how biophysical change impacts the supply of benefits to people requires models that include biophysical, people and the connections among them. IPBES has conceptualized these relationships (Figure 1), and this chapter focuses on models that relate ecosystem services to human well-being, each of the components of this chapter is briefly explained below.
Figure 5.1 This chapter focuses on models that relate ecosystem services to human well-being. This chapter spans part of IPBES conceptual framework.
Page 2 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
5.2.1 Conceptualizing relationships between Biodiversity and Ecosystem Services (BES) and Human Well-being (HWB)
5
10
15
20
25
30
35
40
Biodiversity and ecosystem services contributes to human wellbeing, but modeling the interconnectedness (and trade-offs) among the several dimensions of BES in relation of multivariate aspects of HWB remains a great scientific challenge (Bennett et al. 2009; McShane et al 2011). Some BES dimensions are directly or strongly linked with HWB (e.g., multiple goods and services provided by high diversity coral reefs and rainforest to local communities' food, livelihood, traditional medicine, income, and culture). In fact, large part of the consumed material goods and services by humans rely ultimately on BES provisioning, and income generated from BES such as food production, raw materials, and tourism - and in these cases degrading biodiversity could easily result to reduction in HWB (Butler and Olouch-Kosura 2006; McMichael et al. 2005; Raudsepp-Hearne et al. 2010). However, other HWB dimensions are weakly linked with the production of BES, such as in the case where biodiversity play a supporting role on the consumption of other activities such as recreation (i.e., biodiversity providing opportunities for social connectedness or relaxation), wherein reduced biodiversity may not necessarily reduce social connectedness (Reyers et al. 2013). Moreover, it is a notable fact that while the global biodiversity is declining and huge portions of various ecosystems are in degraded conditions, the overall measure of HWB such as material wealth is actually increasing. Thus, a clear understanding of the relationships and flows of various metrics and aspects of BES benefits to various aspects of HWB of different sectors and beneficiaries is greatly needed in order to prevent further catastrophic collapse of global biodiversity and therefore ensure the continued supply of BES for HWB (McAfee 2012). To date, there are several challenges in the pursuit of modeling BES and HWB relationships for management and policy purposes to sustain BES and improve HWB for most countries around the world. The first major challenge is the lack of a universally acceptable definitions and metrics HWB that could help ensure sustainability and meet global targets such as Aichi Targets for 2020 (McGillivray and Clarke 2006). The Millennium Ecosystem Assessment (MA 2005) assumes that there are multiple constituents of HWB, which are commonly articulated by the following five dimensions: Security, including personal safety, secure resource access, security from disasters; Basic material for life, comprising adequate livelihoods, sufficient nutritious food, shelter access to goods; Health, considering strength, feeling well, access to clean air and water; Good social relations, including social cohesion, mutual respect and ability to help others; and Freedom of choice and action, considering the opportunity to be able achieve what an individual values doing and being. Other studies (Kopmann and Rehdanz 2013; Ringold et al. 2013, De Gasper 2007, Stiglitz et al. 2009, McGillivray and Clarke 2006) also included other components of HWB not explicitly indicated in the MA 2005, such as education, work-life balance, and political voice, and governance. Moreover, there are several other terms which are interchangeably used as a proxy of HWB, as for example, happiness, life satisfaction, sense of place and security, spiritual experience, quality of life, welfare, living standards, utility, prosperity, needs fulfillment, among several others (Diener et al. 2009; McGillivray and Clarke 2006). These various articulations of many dimensions of HWB still needs to be explored and its relationships with various aspects of BES needs to established, modeled, and clarified to help reconcile the often varied and conflicting world views on BES and HWB relationships (Tengö et al. 2013 ). Page 3 of 32
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
Given the multi-dimensional and multivariate nature of HWB, to date, there are a number of accounting frameworks developed to track changes and flow of aggregate indices of HWB that can be linked to BES. Such measures can be grouped into the following three categories: (1) composite indices such as Better Life Index (BLI) and Human Development Index (HDI) (Morse 2003; McGillivray and White 1993; Kasparian and Rolland 2012); (2) green accounting indices (e.g., the utilitarian framework of inclusive wealth, project scale cost benefit analysis, and TEEB) (Serafy 1997; Ring et al. 2010; Sukhdev et al. 2009), and (3) subjective accounting indices such as life satisfaction surveys, studies on happiness (Costanza 2000; Farley 2012; Abdallah et al. 2008)), among others. In addition, the interaction between BES and HWB have been characterized in terms of people’ access to nature through (1) available assets and technology; (2) magnitude or amount of wealth that can be generated per unit area; (3) scale or extent to which the BES can be produced; and (4) timing of BES production, and the frequency of ES delivery such food, water, or wealth in highly seasonal or climate-dependent production systems (Mace et al. 2012; Naidoo et al. 2008). Each of these metrics has its own strengths and weaknesses. To date, the lack of data or appropriate measures for some variables in these indices led to the use of surrogate data (i.e. , environmental quality is represented by amount of particulate matter in the BLI). Therefore, current representations and measures of BES and HWB modeled relationships still needs to be carefully understood and interpreted within the defined context. The remaining sections of this chapter will provide detailed discussions on the various advances in modeling spatio-temporal BES and HWB flows and relationships.
5.2.2 Understanding relationships among anthropogenic assets, institutions, BES, and HWB 25
30
35
40
Anthropogenic assets in combination with BES provide basis of HWB not only to present, but also future generations, and are therefore a key aspect of inter-generational sustainability (Munoz et al. 2014). Assets in general can be defined as “anything which can, either directly or indirectly, yield flows of benefits to people over time (Dickie et al. 2014). In other words, anthropogenic assets provide ‘intertemporal’ flows of goods and services that are valuable for people’s lives. When those benefits to people are derived from nature or biodiversity, we usually refer to them as ecosystem services (ES). Some kind of ES can be accessed ‘directly’ by the final beneficiaries, as, for example, clean air, the enjoyment of some landscapes or recreational services. However, there are several other situations where people’s benefits from nature can only be reaped by means of complementary anthropogenic assets. For example, in the case of timber as a provisioning service, one can think not only of the machineries necessary to cut trees, but also the transport system (and roads) which makes the resource available to final users or producers in the intermediate sectors for further processing. Another example of a complementary anthropogenic asset are vessels, being essential for fishing offshore and in remote areas. Such transport means, machineries, and infrastructure are commonly clustered within the so-called manufactured or produced assets (United Nations 2009). Page 4 of 32
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
Moreover, produced capital is not the only anthropogenic asset utilized in the co-production of benefits by nature and society. The different kind of knowledge, skills and abilities embodied in individuals (i.e., human capital) also contribute significantly to this process. In this regard, one should think of knowledge in a broad sense, comprising indigenous and local knowledge systems as well as technical or scientific knowledge, including also formal and non-formal education (Figure 1). Furthermore, a third common category of capital assets is social capital, which comprises, for instance, social networks, interpersonal agreements or institutions (OECD 2001), the latter playing an important role for regulating the interaction between the supply of ecosystem services and their users.
10
15
20
25
30
35
The joint production of BES benefits thereby relies on the combination of the multiple anthropogenic assets available in the society and nature. For example, food or fiber crops heavily depend on ecosystem processes such as soil formation, nutrient cycling, or primary production (nature supply of benefits), as well as on social intervention of labor, knowledge of genetic variety selection and farming techniques (human capital); and machineries, storage facilities and transportation (produced capital). All these capital assets operate within an institutional framework regulating the interaction between ecosystem services and people. Consequently changes in institutions that alter people’s access to ecosystems can alter the availability of ecosystem services as much as biophysical changes. For example, if people are allowed to harvest wild foods on private land an ecosystem service, then many food products will flow, but if access is eliminated, the food production stops. Similarly, if infrastructure, such as a hiking trail, provides people with access to an area, then it may recreational ecosystem services is generated. However, in both these cases access may be available to some beneficiaries or groups of people and not others. For example, a forest may be accessible to people who have access to a car, but not to those who have no means of transportation to access the same forest. Similarly, a hiking trail may be accessible to able-bodied, but not to people with limited mobility (Reyers et al ., 2013). Furthermore, making nature available to its final beneficiaries is subject not only to the use of complementary anthropogenic assets available in the society, but also to technology. In some instances, a resource or potential BES is known to exist, however, the current state of technology does not allow an efficient access to it as the costs of exploiting this resource exceed the willingness to pay for it. In such a case lack of appropriate technology can be seen as a constraint on the availability of BES. In other cases, however, technology may lead to an overexploitation of the resource by the overuse of the resource, commonly articulated by substantial reductions in the prices of the goods and services resulting from nature (Giampietro and Mayumi 2008).
5.2.3 Identifying and quantifying drivers of BES and HWB changes
40
The flows of ecosystem services often accrue unevenly to various sectors of the society or beneficiaries, and are often influenced by direct anthropogenic drivers (habitat conversion, exploitation, climate change, pollution, population growth, species introduction, changes in income and wealth, international trade, etc), direct natural drivers (e.g., earthquakes, volcanic eruptions, etc), as well as indirect drivers of change in BES (institutions and governance systems, societal level of inequalities, cultural values and Page 5 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
practices, policies, technology, etc). The exact relationships, interactions, and consequences of these drivers on BES and HWB still needs to be further explored and modeled, as well as its feedback loops and thresholds or tipping points. 5
10
15
20
25
30
35
40
5.2.3.1 Direct anthropogenic drivers of BES and HWB changes Human demands for BES are pushing the limits of BES productivity and have consequent negative impacts on HWB. Modern production practices in combination with ballooning human populations and needs for BES often results to habitat conversions, inducing significant losses of BES. The conversion of natural forest into mono crop has for example been linked with those changes in the supply of BES. The conversion of natural forest ecosystem for palm oil, sugar cane, Jatropha spp, and other feedstocks in the tropical forest of Asia and Sub Saharan Africa is one of the most recent forms of Land Use Change and Conversions. Venter et al., (2009) observed that, a number of threatened mammal species are found within extensive areas of planned oil palm developments in Kalimantan, Borneo; which could consequentially lead to local extirpations of these threatened species. Similarly, it is projected that the 550,000 ha of planned sugar cane plantations on river deltas in Kenya and Tanzania could become a major threat to biodiversity in tropical wetlands (Sielhorst et al. 2008). In addition, modern production practices resulted to excessive use of nutrients (e.g., Nitrogen, Phosphorus, and Potassium) in agriculture production, which contributes to nutrient loading in river and marine ecosystems. Nitrogen loading pollutes river and marine ecosystem through eutrophication and impaired water quality - leading to collapse of aquatic ecosystems and seriously decreasing availability of productive or useable clean water (Seitzinger et al., 2002). Moreover, the great expansion and intensification of mining, fossil fuel extraction, and energy use disrupts landscape to varying degrees, which ultimately negatively impacts BES (Alcamo et al., 2005; Geist and Lambin, 2001). Explicit models on feedbacks of impact of anthropogenic activities on ecosystems services would be useful to inform decision making for management of ecosystems. Understanding and quantifying the limits or thresholds of anthropogenic impacts on BES and its negative consequences of HWB requires further scientific explorations, as well as its underlying mechanisms For instance, anthropogenic assets and wealth can have a varied effects on demand for BES in such s way that poor households under no regulation scenarios are likely to maximize the exploitation of BES in order to derive a livelihood (for instance, a fishery, a forest) and may therefore not care much about conservation. On the other hand, more wealthy households may be more interested in using BES for recreation - e.g., a marine park for diving and leisure, or a forest park for relaxation, and would therefore be willing to conserve the said BES. Brown et al. (2008) for instance note that poor persons prioritize the importance of provisioning services and their function of supporting livelihoods over other ecosystem services. Processes of globalization affect interactions between drivers of changes in ecosystem services, and they amplify the driving forces of BES collapse or over-exploitation by removing regional barriers and increasing the interdependency among people and between nations. For example, exchange rates and Page 6 of 32
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
international prices affect local prices and allocations of BES in such a way that destructive dollarearning shrimp farming maybe be favored in spite of the loss of coastal protection ES provided by the converted mangrove areas and thereby exposing local communities to risks of tsunamis and supertyphoons (Barbier et al. 2011). Cross boundary issues such as regulations and conflicts affect demand for BES. Uncertainties about the reliability of an BES, including availability of the resource for local consumers and traders, and inequalities in job opportunities and BES exploitation (e.g., for fishermen along fisheries) under no regulation scenarios can lead to overexploitation of the BES and its collapse (Abunge et al. 2013). Climate change is negatively impacting BES and consequently HWB in many parts of the world, by adding catastrophic events such as super-storms, droughts, flooding, coral reefs and forest die-backs, sea level rise, disease outbreaks, on an already overexploited ecosystems, and will surely put greater threats to the continued production of BES to meet demands for HWB. Modeling the exact impacts of climate change on BES and HWB will aid mitigation and adaptation policies that could address impending negative impacts of climate-driven decline in BES and HWB.
5.2.3.2 Direct natural drivers of BES and HWB changes
20
25
Natural environmental drivers such as tsunamis, hurricanes, cyclones, volcanic eruptions, and earthquakes, fires, and pest or disease outbreaks affect demands and flows of BES at different spatial scales - i.e., local, regional and global. Modeling the consequences of natural environmental drivers on BES production, demands, and flows can greatly help policies for reducing the vulnerability of communities that are greatly exposed or at risk to the major catastrophic impacts of such events. In addition, modeling the compound and interacting effects of natural drivers of BES and HWB changes in combination with anthropogenic drivers could help facilitate prioritization of resource allocations for maintaining BES and HWB.
5.2.3.3 Indirect drivers of BES and HWB changes
30
35
40
Some drivers to BES and HWB changes may not appear as direct and can be hidden and these can be challenging in modeling BES and HWB relationships. Such indirect drivers can include the following: prices and markets, technology development and adaptation, changes in local land use and land cover, species introductions and removals, and external inputs. For instance, fishing technologies (e.g. blast or cyanide fishing), and need to be incorporated in BES and HWB models (Barbier et al. 2011). Currently, the assignments of direct and indirect drivers to BES and HWB relationships my appear fuzzy in literature and cataloguing the categorization and effects of these various drivers of BES and HWB changes can be crucial in advancing BES and HWB modeling and science. Institutions (e.g., Government, NGOs and associated policy-legal frameworks) are meant to protect ecosystems such as forest, wetlands and marine parks from unsustainable exploitation (Abunge, et al., 2013). However, the same institutions can also act to promote BES exploitation through a policy favoring economic growth and prosperity. Norms instilled by collective management under common property resources also affect demand and utilization of BES. In many instances however, institutional failure Page 7 of 32
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
leads to negative conversion of BES, especially where local users have no way of organizing themselves into collective groups for community self regulation in the exploitation of BES. Thus, understanding how production and flow of BES and HWB is mediated by institutional factors is very important. In many parts of Asia, such as Thailand, policy failure - using input and output subsidies for shrimp farming lead to the massive destruction of mangrove ecosystems and thereby exposure of coastal communities to catastrophic storm and tsunami events (Barbier et al. , 2011; Barbier and Cox, 2004). Similarly, the institutional policy to increase the production of Nile Perch in Africa, shrimp in Asia, or corn in Latin America, to feed the international trade may lead to massive loss or conversion of biodiversity in the producing country and contribute to food or fuel security in importing countries, but not necessarily for the exporting country. Highlighting the key roles of indirect drivers of BES and HWB changes can help identify crucial leverage points for securing or sustaining BES and HWB of appropriate beneficiaries.
5.2.4 Linking BES and HWB relationships and understanding feedback loops, thresholds, and interactions 15
20
25
30
35
40
A seminal paper by Balvanera et al. (2006) demonstrated through a meta-analysis that increased biodiversity has a positive effect on supply of most ecosystem services. The study by Balvanera et al. (2006) reviewed 446 measures of biodiversity effects, mostly at plot scale. Similarly, using plot-scale data across the United Kingdom, Maskell et al. (2013) show that the positive relationships between biodiversity and ecosystem service supply are spatially heterogeneous and fundamentally constrained by landscape scale productivity gradients. The incorporation of different components or metrics of biodiversity in BES models are often limited by available data for many ecosystems (see chapter 2, 3, and 6 for detailed discussions on this topic), and therefore are often represented as proxies (e.g., percent good habitat cover, key functional groups, etc) (Vihervaara et al. 2010). Although several multi-species indices, community metrics, and community specialization indices have been proposed (Devictor et al. 2008, Tuomisto 2012; Gregory et al. 2010), they are still rarely used in ecosystem service models. To date, incorporating multiple species in most BES models is still challenging, and majority of existing BES studies treat various aspects of biodiversity and ecosystems as separate modules, often without necessarily incorporating interactions between modules (Villa et al. 2014). For example timber production and water production from a forest and watershed area are treated individually in a BES model without necessarily looking at their interactions and synergies (Villa et al. 2014). Also, in general, there is a biased representation of various ecosystems and taxonomic groups in existing BES models. To date, more BES modeling work have been completed on terrestrial or forest systems, watersheds, birds, and fisheries, than other less economically exploited or accessible ecosystems and taxonomic groups (Vihervaara 2010). Depending on the decision context (see Chapter 5 for elaboration on this topic), types of demands for BES, and production goals, the biodiversity metrics incorporated into the BES models may require estimates of species richness and evenness (index of high diversity) or simply presence of desired functional groups (e.g., extensive plantations of monoculture species or targeted species). For example, non-extractive services such as tourism and aesthetics often favor high biodiversity (i.e., requiring Page 8 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
indices of biodiversity), while extractive ecosystem services such as food production may favor low diversity and monoculture (i.e., requiring measures of biomass of targeted groups). However, the greater susceptibility of monoculture to crashes via disease outbreaks compared to high diversity culture systems should be compared to generate appropriate policies for sustainable food provisioning. 5
10
Currently, BES are modeled through a variety of approaches or methods (see chapter 2, 3, and 6 for detailed discussions on this topic). Major categories of these approaches include the following: (1) integrated modeling assessments (e.g., Fisher et al. 2011), (2) inventory of service production on landscape or BES flow to beneficiaries, (3) modeling functional relationships such as the relationships between land cover, land use change, and carbon sequestration (e.g., Bagstad et al. 2013), (4) probabilistic modeling (e.g., Bagstad et al., 2013), (5) explicit conceptualization of ecosystem services using Stochastic Adaptation of Path Attribution Network (SPAN) (e.g., Johnson et al., 2012), and (6) spatial biophysical modeling (e.g., Bryan and Crossman, 2013). In spite of the existence of a number of ecosystem tools at various spatial scales , the comparative studies of these tools are still lacking.
15
5.2.4.1 Gradual vs. abrupt BES and HWB changes
20
25
Social-ecological systems can shift from being organized around one set of processes to another. These regime shifts are persistent, often difficult to predict, and can occur relatively abruptly (Biggs et al 2012). Such changes usually result in substantial shifts in the ecosystem services produced by the socialecological system (Rocha et al 2015). For example, a clear low nutrient lake can shift to a turbid high nutrient lake, with declines in water quality, recreation, and fish populations or a forest can shift to a savannah, with shifts from forest products to fodder and grazing animal production. These shifts often occur due to the interaction of relatively slow changes in the structure and functioning of ecosystems and extreme events or shocks. For example, the combination of nutrient accumulation in benthic sediments with a extreme erosion event. The multi-scale and non-linear nature of these events they can be difficult to model.
5.2.4.2 Importance of cross-scale interactions in BES and HWB relationships 30
35
40
Cross-scale interactions can also complicate models of ecosystem services. While it is relatively straightforward to represent the constraining or driving impacts of large scale processes on local places, representing tele-connections between distant places, via social processes such as trade, conflict, or colonialism, or through biophysical processes, such as climatic feedbacks, species migration, hydrological flows can present major challenges to models. Some places, such as river deltas are known to be hot spots of connection, where social and biophysical, come together to link separate areas. While other types of cross-scale interaction, for example between moisture recycling, vegetation, fire, agriculture, and soil moisture holding capacity can shift in importance over time. For example, the US dust bowl was produced by the combination of financial regulation, market forces, climate variation, and farming practices in a dryland area with vulnerable soils (ref).
5.2.4.3 Connecting diverse knowledge systems for enhanced BES and HWB governance
Page 9 of 32
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
Diverse sources of knowledge (e.g., traditional, local, scientific, economic, and socio-political) are required in the assessment of status and management of BES and corresponding HWB (Villa et al. 2002; Carpenter et al. 2009; Lamarque et al. 2011; . However, to date, the integration and co-production of various knowledge systems in a decision-making framework to enhance management or governance of BES and HWB still needs evaluation and improvement in terms of the limitations, tools (e.g., for knowledge and information sharing), facilitation (e.g., power sharing), and interpretation (e.g., models developed within knowledge co-production framework that merges scientific and aboriginal knowledge) (Power 2010; Chan et al. 2006; Butler et al. 2013).
5.3 Different assessment & decision contexts assess different types of Ecosystem Benefits & Values
15
Policy, planning and management are three different types of decision making context. Policy involves the formulation of rules and regulations. Planning is a process of organizing, prioritizing and scheduling activities in order to achieve articulated goals. Management involves navigating the inevitable tensions, tradeoffs and opportunities that emerge from implementing any plan and policy.
20
Models and scenarios can improve structured decision making by transparently representing assumptions and thinking underpinning decisions, compressing and synthesizing complex information is an understandable way, and helping identify and explore new policies and unexpected outcomes. However, the value and utility of a model depends upon the decision context in which it is used.
25
30
35
40
Decision contexts is the social-ecological context in which a decision is made. Socially a decision context is defined by the broader social context of why a decision is being made, who is making the decision, and whether that decision maker or decision making body is considered to be legitimate. Ecologically a decision context is defined by what about ecosystem services and biodiversity is being decided, and whether the decision is a one-off decision, or part of a stream of interconnected decisions. Many decisions involving biodiversity, ecosystem services and human well-being are complex and morally fraught involving poorly understood processes, that involve conflicts of interests and values among different groups in society. Structured decision making can help improve such processes, but there is likely to be disagreement about what decision process is legitimate as well as what decisions should be made. For example, in the management of a coastal fishery industrial fishers, indigenous groups, environmental groups and governmental bodies may disagree over who makes decisions, how they are made, & made are the boundaries of decision making. The decision context will determine the scope of ecosystem service modelling and scenario analysis required. Important aspects of decisions contexts relative to ecosystem services include: political, temporal and geographical scale, characteristics of the beneficiaries, epistemologies of the decision makers, the decision dynamics (back-casting vs. forecasting) and the types of decisions to be made (e.g. identifying trade-offs; optimal investments; multi-criteria analyses; socio-political; experimental). In this section we briefly outline the major aspects and aims of IPBES regional and sub-regional assessments followed by discussion of the likely decisions contexts for the major aspects of assessments. Page 10 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
5.3.1 IPBES regional and subregional assessments
5
10
15
20
25
The IPBES regional and subregional assessments propose to assess (IPBES/3/6), at scales yet to be determined (e.g. national, regional, river basin or other), five major aspects of biodiversity and ecosystem services: 1. Trajectories of nature’s values: the values of nature’s benefits to people, including interrelationships between biodiversity, ecosystem functions and benefits to society, as well as the status, trends and future dynamics of ecosystem goods and services; 2. Trajectories of ecosystems: the status and trends of biodiversity and ecosystem services including the structural and functional diversity of ecosystems and genetic diversity; 3. Trajectories of drivers: the status and trends of indirect and direct drivers and the interrelations of such drivers; 4. Risks: future risks to drivers, biodiversity and ecosystems, ecosystem services and humanwellbeing under plausible socioeconomic futures. 5. Policy responses: the effectiveness of existing responses and alternative policy and management interventions, including the Strategic Plan for Biodiversity 2011-2020 and its Aichi Biodiversity Targets and the national biodiversity strategies and action plans developed under the CBD. The assessments will be completed for five regions (Africa, Americas, Asia-Pacific, Europe and Central Asia, and Open Oceans), with each regional assessment following a common structure but tailored to regional-specific contexts. The regional assessments will aim to answer policy relevant questions of i) the contribution of biodiversity and ecosystem services to economies, livelihoods and well-being; ii) the status and trends of that biodiversity and ecosystem services; iii) the pressures driving change in that biodiversity and ecosystem services, and; iv) possible interventions to ensure sustainability of the biodiversity and ecosystem services (IPBES/3/6/Add.1). The IPBES global assessment will then build on the regional and subregional assessments with processes established to ensure coherence between the two scales of assessment.
5.3.2 IPBES decision contexts 30
35
40
The decision contexts of IPBES are many and varied, they included the regional and national assessments, but which in turn will require many different approaches to modelling and scenario analyses. For example, the decision contexts for understanding trajectories of nature’s values are rooted in the social, geographical and economic sciences and will be defined primarily by the relevance of ecosystem service flows to beneficiaries. These analyses typically focus on geo-political boundaries at scales relevant to people and shaped by available demographic data. Decisions impacting substantially on beneficiaries will likely be made at coarse scale within socio-political contexts. This will require understanding, quantifying and mapping the flows of services to beneficiaries, an area of research only recently emerging (Syrbe and Walz 2012, Reyers et al. 2013, Bagstad et al. 2014). Recent concepts for linking beneficiaries to ecosystem services include quantifying service provisioning and benefitting areas and service connecting regions. The questions asked here may include identifying natural ecosystems of high scenic beauty and recreational value and the users of these areas (Palomo et al. 2013, Palomo et al. 2014). Also of question is the location of communities most vulnerable to climate change (who will then Page 11 of 32
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
be greatest beneficiaries of carbon sequestration and the climate regulation service) and to natural disasters such as flooding, landslides and cyclones (who will then be the greatest beneficiaries of flood regulation, erosion control and extreme event moderation ecosystem services, respectively). Of note is the recent work of Renaud et al. (2013) who explore how ecosystems have an important role in reducing risks associated with natural disasters, clearly demonstrating the value of ecosystems to people. Another emerging area of research is the impact of increasing urbanisation on the demand, supply and flow of ecosystem services from agro-ecosystems, and the subsequent risks with the increased disconnect between ecosystems and people (Cumming et al. 2014). The decision contexts for understanding trajectories of ecosystems are rooted in the ecological sciences. Decisions here are supported by biophysical models that aim to represent the processes that underlay the supply of ecosystem services and the changes to supply from changes in ecosystems and biodiversity. Decisions will often be location-specific and will involve identifying trade-offs in biodiversity, ecosystem and ecosystem service supply outcomes between alternative approaches to managing the land, water and biota. Important is establishing the relationships between elements of biota and physical systems and the supply of ecosystem services to provide evidence that management interventions will lead to beneficial outcomes. Recently ecological production functions have been suggested as a robust way to forecast the effect of human impacts on ecosystems and the supply of ecosystem services (Olander and Maltby 2014, Wong et al. 2014). According to Wong et al. (2014), ecological production functions are regression models that measure the statistical influence of marginal changes in ecosystem characteristics on final ecosystem services at a given location and time. A marginal change is the amount an output changes from an additional unit of input, all else held constant.
25
30
35
40
The questions asked related to the trajectories of ecosystems may include understanding the efficacy of land or water management interventions for improving the condition of ecosystems and the subsequent improvements in the supply of ecosystem services. The scale of these types of decisions will generally be small (e.g. plot, paddock, river reach, vegetation community), although may extend to landscapes if ecological connectivity is of interest, and the decision makers will collectively be highly diverse, consisting of many different types of land managers. Another decision context of assessments aims to understand drivers and risks to biodiversity, ecosystem services and human well-being, and the effectiveness of policy responses that mitigate risk. Decisions here will be improved by scenario analyses, potentially at a relatively coarse geographic and temporal scale and may involve any combination of trade-off analyses, optimisation, and multi-criteria analysis. Analyses will typically try to forecast the impact on and trade-offs to biodiversity and ecosystem service supply and demand from external influences, such as new policy and/or climate change (Nelson et al. 2013, Bryan et al. 2014). For example Bryan and Crossman (2013), using high resolution spatial data, modelled nearly 2,000 economic and biophysical scenarios to explore the land use changes, and subsequent impacts on the supply of ecosystem services, that may occur to the year 2050 in southern Australia following policy that creates markets for food, water, carbon and biodiversity. Using Page 12 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
comparable methods, but for the United Kingdom, Bateman et al. (2013) explore the potential land use changes and subsequent impacts on ecosystem service supply under six plausible future socio-economic scenarios that drive land use change. There are others who have done similar for other parts of the world, such as in the USA (Nelson et al. 2009) and Europe (Willemen et al. 2010, Willemen et al. 2012). 5 [Other decision contexts that need to be added Social; Access]
5.4 Types of models 10
15
20
25
5.4.1 What types of attributes differentiate ES models Modelling the impacts on beneficiaries resulting from change in biodiversity and ecosystems takes many different forms for many different purposes (Crossman et al. 2013b). Models of ecosystem services tend to fall into one of two categories: i) simpler proxy-based models of ecosystem services underpinned by land use and land cover, and ii) models which simulate biophysical processes and typically arrive at production functions and detailed system understanding (Kareiva et al. 2011; Crossman et al. 2013a; Maes et al. 2015). Most often modelled is the supply side of ecosystem services, that is, the dynamics of the flow of services from natural capital to people. Much less common is the modelling of beneficiary demands for ecosystem services, or how changes in human populations and demographics translate to changes in demand for natural capital and the flow of services. Chapter ?? describes how changes in beneficiaries impact on changes in demand and then potential changes to ecosystems and biodiversity so this will not be discussed here. In this section we summarise the key attributes of models of ecosystem service flows from the proxy to the process models, focussing on the supply side. We briefly describe the attributes, dynamics, scales, levels of complexity, and handling of uncertainty typically found in models. Different types of models can be related to different parts of the IPBES conceptual framework Figure 2.
Page 13 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 5.2 How different types of model connect to IPBES conceptual framework. (Draft – will be revised as section 4 is completed).
5
10
15
20
25
5.4.1.1 Proxy-based models At the simplest level these models are approximations of ecosystem service flows at a single point in time. Biodiversity (e.g. species distributions), land use, land cover and/or discrete elements of natural capital are usually used as proxies. For example, spatial data on perennial vegetation extent has been used to estimate the flow of ecosystem services such as moderation of extreme events (in combination with soil information, e.g. Chan et al. (2006), Schulp et al. (2012)) and carbon sequestration for climate regulation (in combination with carbon stocks, e.g. Nelson et al. (2009)). Soil and/or broader land cover data has also been used as proxy models for other regulating services such as soil fertility (Maes et al. 2012) and erosion prevention (Maes et al. 2012). Simpler proxy models have improved with the addition of complexity by disaggregating land use/cover data and combining with additional information (e.g. expert knowledge, higher spatial or temporal resolution data). Although still proxy-based, these types of models better account for spatial heterogeneity and may more accurately represent ecological structures and processes. A notable study where land cover data is complemented by a number of additional datasets is the study by Schulp et al. (2014) who modelled the production and consumption of wild foods in Europe. As a proxy for production, Schulp et al. (2014) used species distribution models to downscale coarse resolution species distribution data of important wild food species to high resolution land cover data. To model consumption, Schulp et al. (2014) used a mix of internet and literature searches, cookbook data and hunting statistics. A further example of the integration of expert opinion, land cover data and other empirical data is the matrix models that estimate the capacity (i.e. ability based on ecological condition and integrity) of a landscape to supply ecosystem services, pioneered by Burkhard et al. (2009). These Page 14 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
models have gained popularity as a pragmatic way to quantify spatio-temporal changes in supply of multiple ecosystem services under scenarios and drivers of environmental change, especially in data sparse locations (Kaiser et al. 2013), and to meet co-design, participatory and transdisciplinary needs inherent in ecosystem service assessments (Fish 2011; Jacobs et al. 2015). 5
10
15
20
25
The relative simplicity of proxy models means they require fewer resources and technical expertise, making them useful where ecosystem service data is poor and measurability is difficult. Their simplicity makes them very amenable to participatory processes. Proxy models are transferable, as done in the highly cited Costanza et al. (1997) study and their recent follow up (Costanza et al. 2014) who estimated the supply and value of the world's ecosystem services across a handful of broad global biomes. But the credibility of proxy models has been questioned because of their generalisation (Eigenbrod et al. 2010). Typically absent in proxy methods is system dynamics such as socio-ecological feedbacks, complex interactions, temporal changes, and inclusion of external drivers of change.
5.4.1.2 Process-based models These models aim to describe the ecosystem functions and biophysical processes that underlie the supply of services of benefit to people. These models estimate the flow of ecosystem services from natural capital with more realism than proxy-based methods. Process models can include socioecological feedbacks and interactions at fine scales, and therefore are very applicable to assess the changes to ecosystem services from changes to external drivers under a management, policy or climate scenario. Examples include the use of tree growth models, combined with stand management and spatially explicit soil and climate parameters to simulate carbon sequestration for measuring the climate regulation ecosystem service (Paul et al. 2013; Bryan et al. 2014). Hydrological process models have been used to link changes in land cover and land management to changes in the quantity of freshwater supply (Le Maitre et al. 2007) and the quality of freshwater (Keeler et al. 2012). Norton et al. (2012) integrated three complex process models to estimate the impact of alternative land management scenarios on freshwater quality.
30
35
With many of the process-based models of ecosystem service supply there is a long history in specific scientific disciplines which tends to be overlooked or not reported in the ecosystem services literature. For example hydrologists have for decades been modelling complex hydrological processes using detail time-series climate and stream gauge data, often at daily time-steps over 100+ years, to simulate catchment scale rainfall-runoff dynamics and the outcome of interventions such as land use change or dam construction (e.g. (CSIRO 2008)). Similarly, agronomists have built a number of crop yield simulation models using time-series climate data, soil parameters and crop management regimes which can be used to estimate the food production ecosystem service in agro-ecosystems. A prominent example is the Agricultural Production and Simulation Model (APSIM; (Keating et al. 2003)).
40 Process-based models are designed to replicate complex systems at fine scale and are often calibrated and validated against empirical time-series data to test for uncertainty. These types of models are good Page 15 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
for sensitivity analysis through multiple model runs, and, arguably are best for testing scenarios of management, climate and policy impact. However, they suffer from needing detailed technical expertise to implement, are data and time heavy, and are not easily transferred to other locations. 5
10
15
20
5.4.1.3 Middle ground: simple probabilistic models There has been recent interest in Bayesian probabilistic models that integrate expert knowledge with multiple data sources to model flow of ecosystem services (Haines-Young 2011; Landuyt et al. 2013). Although not themselves models that simulate biophysical processes, Bayesian models call or take outputs from biophysical models, and then integrate with probabilistic qualitative data often derived from expert knowledge about social systems. Integrating expert and stakeholder knowledge with quantitative data makes Bayesian models very useful for evaluating scenario impacts (Keshtkar et al. 2013; Fletcher et al. 2014) in situations of limited data availability and/or where there are participatory, co-design requirements. Being probabilistic, Bayesian models explicitly account for uncertainty. Bayesian models are therefore proposed as a robust way to bridge the gap between the more accurate but less transferable and participatory process models and the simple, transferable but heavily generalised proxy models (Landuyt et al. 2013). The technique of bayesian belief network have also been used to assess ecosystem services. Landuyt et al. (2013) provide a review of 47 such applications. This approach provides advantages of the ability to update and include additional data – which makes them used for applications with limited data -- and explicit treatment of uncertainty, but may be limited by the need for proprietary software, discretization of the data, and the absence of feedback loops. Similarly fuzzy cognitive maps combine an identification of causal links with probabilistic estimations of their impact. These models can be use to make qualitative scenarios more rigorous as well as elicit decision maker models (Kok 2009).
25
5.4.2 Description of major types of ES models [5 pages]
30
Models of ecosystem services can be classified as proxy of process based (Kareiva et al. 2011; Crossman et al. 2013a; Maes et al. 2015). We present models in three operational categories: general frameworks for modelling ecosystem services, general toolkits (which are rapid assessment tools), and other types of modelling approaches (which have been used in specific cases).
5.4.2.1 General models [a short introduction to these models e.g. a 1-3 paragraphs)
35
40
There are a number of general models that aim to provide a framework for assessing, mapping or valuing ecosystem services in particular cases. Below we list some of the tools that explicitly are designed to assess multiple services, that are documented, and have some degree of general availability. We focus on InVEST, Ecopath with EcoSim, and ARIES. We also mention several attempts at developing generalized ecosystem service modelling frameworks that have been used in multiple cases but aren’t generally available or well-documented. InVEST: Integrated Valuation of Ecosystem Services and Tradeoffs
Page 16 of 32
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
InVest is a well-developed and widely applied, suite of models for different types of ecosystem services, typically using the spatial extent and configuration habitat or land use as predictors of ES production. It has been continually developed and expanded by the Natural Capital Project since 2006. As of late 2014, the toolkit includes sixteen distinct InVEST models suited to terrestrial, freshwater, and marine ecosystems. InVEST models are based on production functions that define how an ecosystem's structure and function affect the flows and values of environmental services. The models account for both service supply (e.g. living habitats as buffers for storm waves) and the location and activities of people who benefit from services (e.g. location of people and infrastructure potentially affected by coastal storms) (Nelson et al. 2009, Sharp et al. 2014).
10
15
20
25
30
35
40
To be useful in situations where data is scarce, InVEST tools are relatively simple models with few input requirements. InVEST’s modular design provides an effective tool for exploring the likely outcomes of alternative management and climate scenarios and for evaluating trade-offs among sectors, services, and beneficiaries. These models are best suited for identifying spatial patterns in the provision and value of environmental services and trade-offs between management scenarios. With validation, these models can also provide useful estimates of the magnitude and value of services provided. Advantages of this approach are that it is transparent, open-source and freely accessible, with documentation and training available. The spatial extent of analyses is flexible, allowing users to address questions at the local, regional or global scale. The appropriate application scale is driven primarily by the quality and resolution of input data. Uncertainty in ES estimates produced by the InVEST models may be explored by performing sensitivity analyses on model inputs (e.g. Hamel and Guswa 2014). One model, carbon storage and sequestration, includes an automated uncertainty analysis in which users specify probability distributions for inputs and the model outputs include confidence intervals around carbon estimates. Feedbacks are not explicitly built into the model structure, but are taken into account during the process of project scoping, model building, and implementation. For example, models are often applied in a context of scenario assessment, in which stakeholders explore the consequences of expected changes on natural resources using one or more of the InVEST service models. These scenarios typically include a map of future land use and land cover or, for marine contexts, a map of future coastal/marine uses and habitats, and uncertainties and feedbacks in the social-ecological system should be considered and articulated into the formulation of scenarios. Based on 20 pilot demonstrations of InVEST in a diverse set of decision contexts, Ruckelshaus et al. (2013) have concluded that these simple production function models have been useful, with limitations appearing at the very small scale, and for specific future values. These models have been applied in multiple terrestrial, freshwater, and marine settings and in a range of decision contexts, including development and conservation planning, infrastructure permitting, climate adaptation planning, corporate sustainable sourcing, strategic environmental assessment, and designing payments for ecosystem services (PES). The application of InVEST for ES assessment is most effective when it is embedded within an iterative science-policy process that is broadly participatory (Rosenthal et al. 2014). Page 17 of 32
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
Following stakeholder consultations and scenario development, InVEST can estimate the amount and value of ecosystem services that are provided on the current landscape or under future scenarios. InVEST models are spatially explicit and produce results in either biophysical terms, whether absolute quantities or relative magnitudes (e.g., tons of sediment retained or % of change in sediment retention) or economic terms (e.g., the avoided treatment cost of the water affected by that change in sediment load). InVEST models run as stand-alone software tools, or older versions are available as script tools in the ArcGIS ArcToolBox environment. Users will need a mapping software such as QGIS or ArcGIS to view and synthesize results. Running InVEST effectively does not require knowledge of Python programming, but such skills will facilitate more complex analyses such as uncertainty assessments or optimization. The scope of the project and availability of data affect the amount of time and capacity required to apply InVEST. Typically it will take 1-3 people two months to a year to compile data and run one or more InVEST models. In our experience, the parts of the process requiring the most time include data collection, scenario development and iteration (i.e. re-running the models with better data and further stakeholder discussion to improve the usefulness of the model for decision-making). InVEST provides a framework that can be adapted to the needs of specific applications. For example , Guerry et al. (2012) used the INVEST approach on the west coast of Vancouver Island in British Columbia, Canada to consider multiple services -- shellfish aquaculture harvest, spatial extent of recreational kayaking, water quality, number of recreational homes, and habitat quality -- under baseline conditions and scenarios of industry expansion and conservation zoning. They found that the conservation zoning would increase the production of all services except for the number of recreational float homes, whereas the industry expansion scenario would increase recreational float homes and shellfish aquaculture, having negative effects on habitat and water quality (Guerry et al., 2012). They used a valuation approach for shellfish harvest, but not for the other services considered, and they found that stakeholders found using different currencies for valuing different ecosystem services to be an acceptable approach.
30
35
40
ARIES: ARtificial Intelligence for Ecosystem Services ARtificial Intelligence for Ecosystem Services (ARIES) is widely used process-based model framework that has been applied in the USA, Latin America, and Africa (Villa et al., 2014). It is spatially explicit and currently represents eight different ecosystem services. Initial conditions are set with a Bayesian network that feeds into non-Bayesian dynamic flow models, which include feedback. These flow models are selected from a model repository, and the modeling language is “ThinkLab”. A particular advantage of this approach is the flexibility to use alternative sets of models to assess a particular system. Additionally, ARIES includes the flow of benefits to beneficiaries, which is an important aspect of considering trade-offs (Ville et al., 2014). ARIES does not include valuation, however Sherrouse et al. (2014) have used ARIES together with the Social Values for Ecosystem Services (SolVES) tool, a GIS tool to map and quantify perceived social (nonmonetary) values, including biodiversity (Sherrouse et al. 2011). The SolVES tool is freely available, but requires the use of a proprietary GIS. Page 18 of 32
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
Ecopath with Ecosim (EwE) Ecopath with Ecosim (EwE) is a tropho-dynamic modeling approach that is developed to represent trophic energy flows in marine and aquatic ecosystems. EwE consists of three interlinked components (Christensen and Walters 2004). First, Ecopath describes a static mass-balanced snapshot of the stocks and flows of energy (usually biomass) in an ecosystem. In typical Ecopath models, the modelled foodweb is represented by functional groups that include one or multiple species with similar life history characteristics and trophic ecology and biomass removal by fishing is explicitly represented. Ecopath is described by two basic equations describing biomass production and consumption. Flows of biomass between functional groups are determined by data on diet composition. Out of 435 Ecopath models that are described in a database called EcoBase, 87% of the models were developed to answer questions regarding the functioning of the ecosystem, 64% to analyze fisheries, 34% to focus on particular species of interest, and 11% to consider environmental variability. Less than 10% of the models looked at issues related to MPA, pollution or aquaculture (Colléter et al. 2013). Second, Ecosim allows time dynamic simulation of ecosystems that are described by Ecopath. Ecosim is based on an Ecopath model to provide some of the initial-state Ecosim parameters. It uses a system of time-dependent differential equations from the baseline mass-balance Ecopath model to describe the changes in biomass and flow of biomass within the system over time, by accounting for changes in predation, consumption and fishing rates (Christensen et al., 2005; Pauly et al., 2000; Walters et al., 1997).Particularly, predator-prey interactions are controlled using an algorithm developed based on the “foraging arena theory”, through which spatial resource usages of predators and preys and their effects on their interactions are implicitly represented. It is primarily designed to explore fishing scenarios and their implications for the exploited ecosystems and fisheries catches. Ecosim also enables the representation of environmental forcing to and non-trophic interactions between function groups. Third, Ecospace allows spatial and time dynamic simulation of Ecopath modelled ecosystems. It allows users to explore the effects of spatial fisheries management policies such as Marine Protected Areas. Ecosim has been widely used to generate scenarios of changes or management of fishing effort on flows of ecosystem services from marine ecosystems through fishing. For example, EwE modeling was applied to explore the implications of limitation of beach seine fisheries on the wellbeing of coastal communities in Mombasa, Kenya, with a particularly focus on the poor group. Specifically, EwE provided expected ecological and fisheries responses of the Mombasa coral reef and seagrass ecosystem under a range fishing effort scenarios. The model represented trophic interactions between 56 functional groups and the effects of five fishing gears, including beach seine, fish trap, spear, hook and line, and net. Simulations were run to explore changes in fishing effort of these gear groups from now to 2030. Total catch, net present value of catches less fishing costs (discount rate = 11%) and total fish biomass were collected to provide indicators of food production, profitability and conservation as well as catch per unit effort by gear and by functional group. Functional groups were classified as high or low value based on fisheries monitoring data. The outputs from the EwE models were used to explore human wellbeing implications by using a rule-based ‘toy’ model to combine the key linkages of fish abundances, catches with wellbeing of individual stakeholders. The ‘toy’ model was used in a participatory workshop in which groups of stakeholders in the region was asked to explore ways to manage fishing effort of different Page 19 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
gear groups that would maximize the wellbeing of specific fishing gear groups or seafood traders (Daw et al. in review).
5
10
15
Integrated Systems Dynamics Models Integrated system dynamics models have also been used to translate biodiversity and ecosystem properties to ecosystem services and benefits, within the context of large-scale feedbacks between natural capital and human made capital. The earliest of these models was the Global Unified Metamodel of the Biosphere (GUMBO) model (Boumans et al., 2002). THE GUMBO model was used by Arbault et al. (2014) to consider life cycle analysis. The Multiscale Integrated Earth Systems Model (MIMES) builds on the GUMBO model using a spatially explicit approach and valuation methods for most ecosystem services (Boumans and Costanza, 2008). MIMES has been developed in Simile software, a commercial software package. Similarly, Fiksel et al. (2013) have used a systems approach to consider linkages among economy, society, and environment, where flows of ecosystem services provide value to both the economy and society. This approach has been implemented in the VENSIM software, which requires purchase of a license for commercial or government use. An advantage of systems dynamics models is that they are comprehensive, and represent feedbacks among sectors within each timestep. However, their complexity limits the usability by a wider group of stakeholders.
5.4.2.2 General Ecosystem Service Toolkits 20
25
30
There are several additional proxy-based or screening level tools and approaches available from a variety of sources that relate ecosystem state to ecosystem services (but do not include valuation). The Corporate Ecosystem Services Review (Hanson et al., 2012), developed by the World Resources Institute, is a structured methodology that helps businesses that interact with ecosystems to connect ecosystem health to business risks and opportunities. The Ecosystem Services Review uses a qualitative approach to consider the 27 ecosystem services given in the Millennium Ecosystem Assessment. TESSA is a toolkit that uses decision trees to guide users, through a process to rapidly prioritize ecosystem services for assessment, identify data needs, and communication approaches. It provides a template that users must adapt to specific cases (Phe et al 2013, Birch et al 2014). Co$ting Nature (Mulligan et al., 2010) models changes in four ecosystem services (carbon storage, water yield, nature-based tourism, and natural hazard mitigation) under scenarios of climate or landuse change. Similarly, the Land Utilization & Capability Indicator (LUCI) is a framework that considers services of production, carbon, flooding, erosion, sediment delivery, water quality, and habitat, based on GIS land and soil information (Jackson et al, 2013). These three approaches, ESR, Co$ting Nature, and LUCI, are compared in Bagstad et al. (2013).
35
5.4.2.3 General Green Accounting Approaches
40
Similar to accounting approaches focussed on ecosystem services there are simple accounting approaches for estimating how nature contributes to human wellbeing and inclusive wealth. None of these accounting frameworks succeeds in capturing all aspects of human wellbeing and all of these research areas are rapidly changing as researchers attempt to better define practical definitions and measures of human well-being, however different approaches provide complementary insights. These Page 20 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
accounting frames can be grouped in three categories: green accounting, composite indices, and subjective measures.
5
10
15
Green accounting approaches are usually conducted at the national level and represent environmental economics attempts to better incorporate environmental values in national accounts. These approaches include inclusive wealth, green GDP, and SEENA [need ref]. TEEB has also developed more case based green accounting frameworks for bringing ES into economic decision making [need ref]. Composite indices combine a set of different aspects into a single value to represent human wellbeing. Perhaps the most famous and widely applied of these is UNDP’s human develop index, which combine education, life expectancy, and income to approximate wellbeing based on Sen’s capabilities framework. Other examples of this type of indices include OECD’s better life index. [refs] A third approach is subjective measures, such as life satisfaction surveys. The most well known, widely used, and longest running of these is the world values survey. Such survey can be difficult to compare between groups, but local research can effectively combined subjective and objective measures of wellbeing in the context of ecosystem services (Abunge et al 2014).
5.4.2.4 Other types of approaches [this list needs to be further developed] 20
25
Participatory Scenario approaches [e.g. Peterson et al 2003 E&S + many recent examples] Scenario model linking approaches [e.g. Alcalmo] Expert ES matrices Urban ES in Capetown [Patrick O’Farrel et al XXX] Fisher et al. (2009) noted that simple overlays of maps of production function outputs can be used to identify where bundles of services will be produced on the landscape. This approach was used by Raudsepp-Hearne et al. (2010) in consideration of 12 ecosystem services in Quebec, Canada – they found six types of bundles, which they were able to link to areas on the landscape. Additionally, Raudsepp-Hearne et al. (2010) demonstrated tradeoffs between provisioning and regulating and cultural services, where regulating services were positively correlated with the diversity of services.
30
5.4.3 Comparing model types across decision contexts
35
40
Models and scenarios can improve structured decision making by transparently representing assumptions and thinking underpinning decisions, compressing and synthesizing complex information is an understandable way, and helping identify and explore new policies and unexpected outcomes. However, the value and utility of a model depends upon the decision context in which it is used. Decision making context can be defined by their ecological, social and decision context (Table 1). The complexity or simplicity of each of components what type of models better fit each of these contexts. Because model complexity increased geometrically with the number of variables included in a model increase in the number of variables explicitly modeled greatly increase the difficulty of creating, parameterizing, applying, analyzing and communicating a model. Consequently, decision context Page 21 of 32
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
complexity should not be met with a complex model, rather by the intelligent application of a set of simpler models that can address complementary aspects of complexity. Alternatively, a sequential process of modeling can potential iteratively reduce the complexity of the decision context by identifying key regions, variables, and decisions, by fostering data collection and synthesis, or by building trust, and enabling communication among different stakeholders. Table 5.1 Key ecological (green), social (red), and decision processes (blue) variables defining decision contexts. Variables
Simple
Complex
Geography/ecology
Homogenous
Diverse
Flows across landscape
Weakly connected
Strong interconnections
Connection
Weakly influenced by external world
Strongly influenced by external world
Governance system
Informal
Regulated/legal
Values
Homogenous
Diverse
Knowledge systems
Homogenous
Conflicting
Time period
Short term
Long term
Decision process
Unitary
Participatory
Objectives
Single objective
Multiple objective
Stakeholders
Unified
Diverse Contesting
Legitimacy
Accepted
Contested
5.4.3 Comparing model types across decision contexts 10
Models and scenarios can improve structured decision making by transparently representing assumptions and thinking underpinning decisions, compressing and synthesizing complex information is an understandable way, and helping identify and explore new policies and unexpected outcomes. However, the value and utility of a model depends upon the decision context defined by their ecological, social and decision context (Table 1).
15 [need to actually review modeling approaches presented in this section]
20
Most model development has focused on assessing ecosystem services, with minimal attention being focused on how these ecosystem services link to human wellbeing, especially of diverse groups of beneficiaries. Modelling the impact of ecological changes on human well-being is not well developed. This is partially because understanding of human wellbeing is poor, but especially due to the great need Page 22 of 32
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
for and historical lack of collaboration between people studying human wellbeing and those studying ecosystem services. Developing tools that better link human wellbeing and ecosystem services will require investment and trans-disciplinary collaboration of policy makers, with natural and social scientists to develop new frameworks, methods, and tools. Most modelling tools have been developed with government decision makers in mind, and there is a need for tools that are aimed to be used and adapted by other actors, and especially that can be adapted to fit with other knowledge systems. Particular issues that need more model development are: impact of ecological change on different groups of people, incorporation of different knowledge systems in modeling operation and practice, and adapting model communication for different decision contexts.
10
5.4.4 Methods for assessing, and communicating, uncertainty in modelling of consequences of change in biodiversity and ecosystems for nature’s benefits to people. 15
20
25
Decision making is shaped by two types of uncertainty, normative or decision uncertainty that is uncertainty about appropriate course of action and information uncertainty that is lack of information or knowledge about how the world works (Bark et al. 2013). Information uncertainty can be either statistical or systematic. Statistical uncertainty is the natural variation in a system which can be quantified through probability distributions, which can then be used to derive confidence intervals and risk profiles. Climate variability is an example of statistical uncertainty. Systematic uncertainty describes the lack of complete knowledge in a modelled system and its parameters and is a typical feature of complex socio-ecological systems that contain non-linear relationships, unpredictable stochastic behaviour, and unknown system conditions. Although systematic uncertainty cannot be quantified, it can be reduced to statistical uncertainty by collecting more data or improving system understanding. Bark et al. (2013) make it clear that in ecosystem service assessments, much uncertainty exists because of the complex physical and ecological systems that underpin supply of ecosystem services, plus the large uncertainty inherent in socio-economic systems that value/demand ecosystem services. In other words, ecosystems service assessments are a prime example of where there are deep levels of statistical and systemic uncertainty.
30
35
40
Schulp et al. (2014) clearly documents five sources of uncertainty in ecosystem services quantified across Europe. They identify uncertainty in the i) indicators, definitions and framework that classify ecosystem services; ii) level of process understanding which leads to different quantification methods; iii) purposes of quantification that influences the selection of indicators; iv) biophysical and socioeconomic input data, and v) models used to quantify ecosystem services. Robust communication of uncertainty has long occupied the IPCC. For the 5th Assessment Report (Mastrandrea et al. 2010), the IPCC used an elegant system that qualitatively describes the levels of confidence in reported findings, based on expert judgement, determined through evaluation of evidence and model agreement. They also used quantitative reporting of uncertainty that stems from statistical or modelling analyses, expert opinion, or other quantitative analyses. They describe Page 23 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
uncertainty using a likelihood scale to express a probabilistic estimate of the occurrence of a single event or outcome.
5
5.4.5 Data needs for model development, and for ongoing evaluation and calibration [to be done after evaluation of model assessment - to put in context]
10
15
20
Spatial data describing, ideally at high spatial and temporal resolution, the following: Land use, land cover Species presence and absence Physical attributes (e.g. soil, water, topography, geology, climate) Ecosystems (e.g. vegetation communities; biomes; primary production) Geo-political boundaries (e.g. coutry/state/local governments) Built infrastructure (e.g. cities and built-up areas; roads; dams) Protected areas and conservation zones (parks; green space; camping and recreational features) Demographic data (population characteristics) Economic data (land values, agricultural production and value) The above is primary data, but where possible if spatially explicit ecosystem service data is available, then even better. for example, the MAES project under EU FP7 has mapped many ecosystem services across the EU member states. The US EPA, through their EnviroAtlas, is in the process of doing similar. [Need to link to other knowledge systems here]
25
5.5 Synthesis & Research Frontiers
[2 pages]
[to be done after evaluation of model assessment]
5.5.1 A taxonomy of existing approaches [1 page] A synthesis of chapter – linking capacities to needs [to do after finishing 4.3] 30
5.5.2 Research frontiers [1 page] identify where are key gaps and potential areas for rapid progress [link to chpt 8?] [to do later] 35
Page 24 of 32
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
References
5
10
15
20
25
30
35
40
Abdallah, Saamah, Sam Thompson, and Nic Marks. 2008. “Estimating Worldwide Life Satisfaction.” Ecological Economics 65: 35–47. doi:10.1016/j.ecolecon.2007.11.009. Abunge, C., Coulthard, S. & Daw, T.M. (2013) Connecting marine ecosystem services to human well-being: insights from participatory well-being assessment in Kenya. Ambio, 42, 1010–21. Alcamo, J., van Vuuren, D., Cramer, W. (2005). Changes In Ecosystem Services And Their Drivers AcrossThe Scenarios. Ecosystems And Human Well-Being: Scenarios, Volume 2. Millennium Ecosystem Assessment, Chapter 9, pages 297-373. http://www.millenniumassessment.org/documents/document.333.aspx.pdf Bagstad, K. J., F. Villa, D. Batker, J. Harrison-Cox, B. Voigt, and G. W. Johnson. 2014. From theoretical to actual ecosystem services: mapping beneficiaries and spatial flows in ecosystem service assessments. Ecology and Society 19. Bagstad, K.J., Semmens, D.J., Waage, S. & Winthrop, R. (2013) A comparative assessment of decisionsupport tools for ecosystem services quantification and valuation. Ecosystem Services, 5, 27–39. Balvanera, P., A. B. Pfisterer, N. Buchmann, J. S. He, T. Nakashizuka, D. Raffaelli, and B. Schmid. 2006. Quantifying the evidence for biodiversity effects on ecosystem functioning and services. Ecology Letters 9:11461156. Bark, R. H., L. J. M. Peeters, R. E. Lester, C. A. Pollino, N. D. Crossman, and J. M. Kandulu. 2013. Understanding the sources of uncertainty to reduce the risks of undesirable outcomes in large-scale freshwater ecosystem restoration projects: An example from the Murray–Darling Basin, Australia. Environmental Science & Policy 33:97-108. Bateman, I. J., A. R. Harwood, G. M. Mace, R. T. Watson, D. J. Abson, B. Andrews, A. Binner, A. Crowe, B. H. Day, S. Dugdale, C. Fezzi, J. Foden, D. Hadley, R. Haines-Young, M. Hulme, A. Kontoleon, A. A. Lovett, P. Munday, U. Pascual, J. Paterson, G. Perino, A. Sen, G. Siriwardena, D. van Soest, and M. Termansen. 2013. Bringing Ecosystem Services into Economic Decision-Making: Land Use in the United Kingdom. Science 341:45-50. Bennett, E.M. (2003) Why global scenarios need ecology. FRONTIERS IN ECOLOGY AND THE ENVIRONMENT, 1, 322–329. Bennett, Elena M., Garry D. Peterson, and Line J. Gordon. 2009. “Understanding Relationships among Multiple Ecosystem Services.” Ecology Letters 12: 1394–1404. doi:10.1111/j.1461-0248.2009.01387.x. Brendan Fisher, et al., (2011). Measuring, modeling and mapping ecosystem services in the Eastern Arc Mountains of Tanzania. Progress in Physical Geography 35(5) 591-611 DOI: 0.1177/0309133311422968 Bryan B.A., Nolan M., Harwood T.D., Connor J.D., Navarro-Garcia J., King D., Summers D.M., Newth D., Cai Y., Grigg N., Harman I., Crossman N.D., Grundy M.J., Finnigan J.J., Ferrier S., Williams K.J., Wilson K.A., Law E.A. & Hatfield-Dodds S. (2014). Supply of carbon sequestration and biodiversity services from Australia's agricultural land under global change. Global Environmental Change, 28, 166-181. Bryan, B. A., and N. D. Crossman. 2013. Impact of multiple interacting financial incentives on land use change and the supply of ecosystem services. Ecosystem Services 4:60-72. Bryan, B. A., M. Nolan, T. D. Harwood, J. D. Connor, J. Navarro-Garcia, D. King, D. M. Summers, D. Newth, Y. Cai, N. Grigg, I. Harman, N. D. Crossman, M. J. Grundy, J. J. Finnigan, S. Ferrier, K. J. Williams, K. A. Wilson, E. A. Law, and S. Hatfield-Dodds. 2014. Supply of carbon sequestration and biodiversity services from Australia's agricultural land under global change. Global Environmental Change 28:166-181. Bryan, B. a., Nolan, M., Harwood, T.D., Connor, J.D., Navarro-Garcia, J., King, D., Summers, D.M., Newth, D., Cai, Y., Grigg, N., Harman, I., Crossman, N.D., Grundy, M.J., Finnigan, J.J., Ferrier, S., Williams, K.J., Wilson, K. a., Law, E. a. & Hatfield-Dodds, S. (2014) Supply of carbon sequestration and biodiversity services from Australia’s agricultural land under global change. Global Environmental Change, 28, 166–181.
Page 25 of 32
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Bryan, B. a., Raymond, C.M., Crossman, N.D. & Macdonald, D.H. (2010) Targeting the management of ecosystem services based on social values: Where, what, and how? Landscape and Urban Planning, 97, 111–122. Bryan, Brett Anthony, and Neville David Crossman. 2013. “Impact of Multiple Interacting Financial Incentives on Land Use Change and the Supply of Ecosystem Services.” Ecosystem Services 4 (June). Elsevier: 60– 72. doi:10.1016/j.ecoser.2013.03.004. Burkhard B., Kroll F., Müller F. & Windhorst W. (2009). Landscapes‘ Capacities to Provide Ecosystem Services - a Concept for Land-Cover Based Assessments. Landscape Online, 15, 1-22. Butler, Colin D., and Willis Oluoch-Kosura. 2006. “Linking Future Ecosystem Services and Future Human Well-Being.” Ecology and Society 11. doi:30. Butler, James R A, Grace Y. Wong, Daniel J. Metcalfe, Miroslav Honzák, Petina L. Pert, Nalini Rao, Martijn E. van Grieken, et al. 2013. “An Analysis of Trade-Offs between Multiple Ecosystem Services and Stakeholders Linked to Land Use and Water Quality Management in the Great Barrier Reef, Australia.” Agriculture, Ecosystems and Environment 180: 176–91. doi:10.1016/j.agee.2011.08.017. Cardinale, B.J., Duffy, J.E., Gonzalez, A., Hooper, D.U., Perrings, C., Venail, P., Narwani, A., Mace, G.M., Tilman, D., Wardle, D. a, Kinzig, A.P., Daily, G.C., Loreau, M., Grace, J.B., Larigauderie, A., Srivastava, D.S. & Naeem, S. (2012) Biodiversity loss and its impact on humanity. Nature, 486, 59–67. Carpenter, S.R., Defries, R., Dietz, T., Mooney, H.A., Polasky, S., Reid, W. V & Scholes, R.J. (2006) Millennium Ecosystem Assessment : Research Needs. Science, 314. Chan K.M.A., Shaw M.R., Cameron D.R., Underwood E.C. & Daily G.C. (2006). Conservation planning for ecosystem services. Plos Biology, 4, 2138-2152. Chan, K.M.A., Guerry, A.D., Balvanera, P., Klain, S., Satterfield, T., Basurto, X., Bostrom, A., Chuenpagdee, R., Gould, R., Halpern, B.S. & Hannahs, N. (2012a) Where are Cultural and Social in Ecosystem Services? A Framework for Constructive Engagement. BioScience, 62, 744–756. Chan, K.M.A., Satterfield, T. & Goldstein, J. (2012b) Rethinking ecosystem services to better address and navigate cultural values. ECOLOGICAL ECONOMICS, 74, 8–18. Chan, Kai M A, M. Rebecca Shaw, David R. Cameron, Emma C. Underwood, and Gretchen C. Daily. 2006. “Conservation Planning for Ecosystem Services.” PLoS Biology 4: 2138–52. doi:10.1371/journal.pbio.0040379. Christensen, V., Walters, C.J., 2004. Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling 172, 109–139. doi:10.1016/j.ecolmodel.2003.09.003 Christensen, V., Walters, C.J., Pauly, D., 2005. Ecopath with Ecosim: a user’s guide. Fisheries Centre, University of British Columbia, Vancouver, BC, Canada. Colléter M., Valls A., Guitton J., Morissette L., Arreguín-Sánchez F., Christensen V., Gascuel D., Pauly D. (2013) EcoBase: a repository solution to gather and communicate information from EwE models. Fisheries Centre Research Reports 21(1), 60p. Costanza R., d'Arge R., de Groot R., Farber S., Grasso M., Hannon B., Limburg K., Naeem S., ONeill R.V., Paruelo J., Raskin R.G., Sutton P. & vandenBelt M. (1997). The value of the world's ecosystem services and natural capital. Nature, 387, 253-260. Costanza R., de Groot R., Sutton P., van der Ploeg S., Anderson S.J., Kubiszewski I., Farber S. & Turner R.K. (2014). Changes in the global value of ecosystem services. Global Environmental Change, 26, 152-158. Costanza, Robert. 2000. “Social Goals and the Valuation of Ecosystem Services.” Ecosystems. doi:10.1007/s100210000002. Crossman N.D., Bryan B.A., de Groot R.S., Lin Y.-P. & Minang P.A. (2013a). Land science contributions to ecosystem services. Current Opinion in Environmental Sustainability, 5, 509-514. Crossman N.D., Burkhard B., Nedkov S., Willemen L., Petz K., Palomo I., Drakou E.G., Martín-Lopez B., McPhearson T., Boyanova K., Alkemade R., Egoh B., Dunbar M.B. & Maes J. (2013b). A blueprint for mapping and modelling ecosystem services. Ecosystem Services, 4, 4-14. Page 26 of 32
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
Crossman, N.D. & Bryan, B. a. (2009) Identifying cost-effective hotspots for restoring natural capital and enhancing landscape multifunctionality. Ecological Economics, 68, 654–668. Crossman, N.D., Burkhard, B., Nedkov, S., Willemen, L., Petz, K., Palomo, I., Drakou, E.G., Martín-Lopez, B., McPhearson, T., Boyanova, K., Alkemade, R., Egoh, B., Dunbar, M.B. & Maes, J. (2013) A blueprint for mapping and modelling ecosystem services. Ecosystem Services, 4, 4–14. CSIRO (2008). Water Availability in the Murray-Darling Basin. A report to the Australian Government from the CSIRO Murray-Darling Sustainable Yields Project. In. CSIRO Australia, p. 67. Cumming, G. S., A. Buerkert, E. M. Hoffmann, E. Schlecht, S. von Cramon-Taubadel, and T. Tscharntke. 2014. Implications of agricultural transitions and urbanization for ecosystem services. Nature 515:50-57. Daily, G.C., Polasky, S., Goldstein, J., Kareiva, P.M., Mooney, H. a, Pejchar, L., Ricketts, T.H., Salzman, J. & Shallenberger, R. (2009) Ecosystem services in decision making: time to deliver. Frontiers in Ecology and the Environment, 7, 21–28. Daniel, T.C., Muhar, A., Arnberger, A., Aznar, O., Boyd, J.W., Chan, K.M.A., Costanza, R., Elmqvist, T., Flint, C.G., Gobster, P.H., Gret-Regamey, A., Lave, R., Muhar, S., Penker, M., Ribe, R.G., Schauppenlehner, T., Sikor, T., Soloviy, I., Spierenburg, M., Taczanowska, K., Tam, J. & von der Dunk, A. (2012) Contributions of cultural services to the ecosystem services agenda. PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, 109, 8812–8819. Daw, T., Brown, K., Rosendo, S. & Pomeroy, R. (2011) Applying the ecosystem services concept to poverty alleviation: the need to disaggregate human well-being. Environmental Conservation, 38, 370–379. Devictor, Vincent, Romain Julliard, Joanne Clavel, Frédéric Jiguet, Alexandre Lee, and Denis Couvet. 2008. “Functional Biotic Homogenization of Bird Communities in Disturbed Landscapes.” Global Ecology and Biogeography 17: 252–61. doi:10.1111/j.1466-8238.2007.00364.x. Diener, Ed, Richard E. Lucas, and Shigehiro Oishi. 2009. “Subjective Well-Being: The Science of Happiness and Life Satisfaction.” In The Oxford Handbook of Positive Psychology, 63–73. doi:0.1093/oxfordhb/9780195187243.013.0017. Eigenbrod F., Armsworth P.R., Anderson B.J., Heinemeyer A., Gillings S., Roy D.B., Thomas C.D. & Gaston K.J. (2010). The impact of proxy-based methods on mapping the distribution of ecosystem services. Journal of Applied Ecology, 47, 377-385. El Serafy, Salah. 1997. “Green Accounting and Economic Policy.” Ecological Economics 21: 217–29. doi:10.1016/S0921-8009(96)00107-3. EU Biodiversity Strategy. (2013) Mapping and Assessment of Ecosystems and Their Services: An Analytical Framework for Ecosystem Assessments under Action 5 of the EU Biodiversity Strategy to 2020. Farley, Joshua. 2012. “Ecosystem Services: The Economics Debate.” Ecosystem Services 1: 40–49. doi:10.1016/j.ecoser.2012.07.002. Fish R.D. (2011). Environmental decision making and an ecosystems approach: Some challenges from the perspective of social science. Progress in Physical Geography, 35, 671-680. Fletcher P.J., Kelble C.R., Nuttle W.K. & Kiker G.A. (2014). Using the integrated ecosystem assessment framework to build consensus and transfer information to managers. Ecological Indicators, 44, 11-25. Gasper, Des. 2007. “Human Well-Being: Concepts and Conceptualizations.” In Human Well-Being. Concept and Measurement, 23–64. Geist, H., Lambin, E. F. (2001). What Drives Tropical Deforestation? A Meta-Analysis Of Proximate And Underlying Causes Of Deforestation Based On Subnational Case Study Evidence. LUCC report series No.4. Gregory, Richard D., and Arco van Strien. 2010. “Wild Bird Indicators: Using Composite Population Trends of Birds as Measures of Environmental Health.” Ornithological Science. doi:10.2326/osj.9.3.
Page 27 of 32
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Guerry, A. D., et al. (2012). "Modeling benefits from nature: using ecosystem services to inform coastal and marine spatial planning." International Journal of Biodiversity Science, Ecosystem Services & Management 8(1-2): 107-121. Haase, D., Larondelle, N., Andersson, E., Artmann, M., Borgström, S., Breuste, J., Gomez-Baggethun, E., Gren, Å., Hamstead, Z., Hansen, R., Kabisch, N., Kremer, P., Langemeyer, J., Rall, E.L., McPhearson, T., Pauleit, S., Qureshi, S., Schwarz, N., Voigt, A., Wurster, D. & Elmqvist, T. (2014) A quantitative review of urban ecosystem service assessments: concepts, models, and implementation. Ambio, 43, 413–33. Haines-Young R. (2011). Exploring ecosystem service issues across diverse knowledge domains using Bayesian Belief Networks. Progress in Physical Geography, 35, 681-699. Hamel, P. and A.J. Guswa. (2014). “Uncertainty analysis of a spatially-explicit annual water-balance model: case study of the Cape Fear Catchment, NC.” Hydrol. Earth Syst. Sci. Discuss. 11:11001-11036. Jacobs S., Burkhard B., Van Daele T., Staes J. & Schneiders A. (2015). ‘The Matrix Reloaded’: A review of expert knowledge use for mapping ecosystem services. Ecological Modelling, 295, 21-30. Johnson et al., (2012). Modelling Ecosystem Service Flows under Uncertainty with Stochastic SPAN. International Environmental Modelling and Software Society (iEMSs), International Congress on Environmental Modelling and Software: Managing Resources of a Limited Planet, Sixth Biennial Meeting, Leipzig, Germany, R. Seppelt, A.A. Voinov, S. Lange, D. Bankamp (Eds.) http://www.iemss.org/society/index.php/iemss-2012proceedings Johnson, G W, K J Bagstad, R R Snapp, and F Villa. 2012. “Service Path Attribution Networks (SPANs): A Network Flow Approach to Ecosystem Service Assessment.” International Journal of Agricultural and Environmental Information Systems 3: 54–71. doi:10.4018/jaeis.2012070104. Jopke, C., Kreyling, J., Maes, J. & Koellner, T. (2015) Interactions among ecosystem services across Europe: Bagplots and cumulative correlation coefficients reveal synergies, trade-offs, and regional patterns. Ecological Indicators, 49, 46–52. Kaiser G., Burkhard B., Römer H., Sangkaew S., Graterol R., Haitook T., Sterr H. & Sakuna-Schwartz D. (2013). Mapping tsunami impacts on land cover and related ecosystem service supply in Phang Nga, Thailand. Nat. Hazards Earth Syst. Sci., 13, 3095-3111. Kareiva P., Tallis H., Ricketts T.H., Daily G.C. & Polasky S. (eds.) (2011). Natural Capital: Theory and Practice of Mapping Ecosystem Services. Oxford University Press, Oxford, UK. Kasparian, Jérôme, and Antoine Rolland. 2012. “OECD’s ‘Better Life Index’: Can Any Country Be Well Ranked?.” Journal of Applied Statistics. doi:10.1080/02664763.2012.706265. Keating B.A., Carberry P.S., Hammer G.L., Probert M.E., Robertson M.J., Holzworth D., Huth N.I., Hargreaves J.N.G., Meinke H., Hochman Z., McLean G., Verburg K., Snow V., Dimes J.P., Silburn M., Wang E., Brown S., Bristow K.L., Asseng S., Chapman S., McCown R.L., Freebairn D.M. & Smith C.J. (2003). An overview of APSIM, a model designed for farming systems simulation. European Journal of Agronomy, 18, 267-288. Keeler B.L., Polasky S., Brauman K.A., Johnson K.A., Finlay J.C., O’Neill A., Kovacs K. & Dalzell B. (2012). Linking water quality and well-being for improved assessment and valuation of ecosystem services. Proceedings of the National Academy of Sciences. Keshtkar A.R., Salajegheh A., Sadoddin A. & Allan M.G. (2013). Application of Bayesian networks for sustainability assessment in catchment modeling and management (Case study: The Hablehrood river catchment). Ecological Modelling, 268, 48-54. Kok, K. (2009) The potential of Fuzzy Cognitive Maps for semi-quantitative scenario development, with an example from Brazil. Global Environmental Change, 19, 122–133. Kok, K., van Vliet, M., Bärlund, I., Dubel, A. & Sendzimir, J. (2011) Combining participative backcasting and exploratory scenario development: Experiences from the SCENES project. Technological Forecasting and Social Change, 78, 835–851. Page 28 of 32
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Kopmann, Angela, and Katrin Rehdanz. 2013. “A Human Well-Being Approach for Assessing the Value of Natural Land Areas.” Ecological Economics 93: 20–33. doi:10.1016/j.ecolecon.2013.04.014. Landuyt D., Broekx S., D'Hondt R., Engelen G., Aertsens J. & Goethals P.L.M. (2013). A review of Bayesian belief networks in ecosystem service modelling. Environmental Modelling & Software, 46, 1-11. Le Maitre D.C., Milton S.J., Jarmain C., Colvin C.A., Saayman I. & Vlok J.H.J. (2007). Linking ecosystem services and water resources: landscape-scale hydrology of the Little Karoo. Frontiers in Ecology and the Environment, 5, 261-270. Mace, G.M. (2014) Whose conservation? Science, 345, 1558–1560. Mace, Georgina M, Ken Norris, and Alastair H Fitter. 2012. “Biodiversity and Ecosystem Services: A Multilayered Relationship.” Trends in Ecology & Evolution 27 (1): 19–26. doi:10.1016/j.tree.2011.08.006. Maes J., Crossman N.D. & Burkhard B. (2015). Mapping ecosystem services. In: Handbook on Ecosystem Services (eds. Haines-Young R, Potschin M, Fish R & Turner K). Earthscan from Routledge UK. Maes J., Paracchini M.L., Zulian G., Dunbar M.B. & Alkemade R. (2012). Synergies and trade-offs between ecosystem service supply, biodiversity, and habitat conservation status in Europe. Biological Conservation, 155, 112. Maes, J., Egoh, B., Willemen, L., Liquete, C., Vihervaara, P., Schägner, J.P., Grizzetti, B., Drakou, E.G., Notte, A. La, Zulian, G., Bouraoui, F., Luisa Paracchini, M., Braat, L. & Bidoglio, G. (2012) Mapping ecosystem services for policy support and decision making in the European Union. Ecosystem Services, 1, 31–39. Malinga, R., Gordon, L.J., Lindborg, R. & Jewitt, G. (2013) Using Participatory Scenario Planning to Identify Ecosystem Services in. , 18. Martín-López, B., Iniesta-Arandia, I., García-Llorente, M., Palomo, I., Casado-Arzuaga, I., Amo, D.G. Del, Gómez-Baggethun, E., Oteros-Rozas, E., Palacios-Agundez, I., Willaarts, B., González, J. a, Santos-Martín, F., Onaindia, M., López-Santiago, C. & Montes, C. (2012) Uncovering ecosystem service bundles through social preferences. PloS one, 7, e38970. Maskell, L. C., A. Crowe, M. J. Dunbar, B. Emmett, P. Henrys, A. M. Keith, L. R. Norton, P. Scholefield, D. B. Clark, I. C. Simpson, and S. M. Smart. 2013. Exploring the ecological constraints to multiple ecosystem service delivery and biodiversity. Journal of Applied Ecology 50:561-571. Mastrandrea, M. D., C. B. Field, T. F. Stocker, O. Edenhofer, K. L. Ebi, D. J. Frame, H. Held, E. M. Kriegler, K.J., P. R. Matschoss, G.-K. Plattner, G. W. Yohe, and F. W. Zwiers. 2010. Guidance Note for Lead Authors of the IPCC Fifth Assessment Report on Consistent Treatment of Uncertainties. Intergovernmental Panel on Climate Change (IPCC). McAfee, Kathleen. 2012. “Nature in the Market-World: Ecosystem Services and Inequality.” Development. doi:10.1057/dev.2011.105. McGillivray, M., Clarke, M., 2006. Understanding Human Well-being. United Nations University, Tokyo-New York-Paris McGillivray, Mark, and Howard White. 1993. “Measuring Development? The UNDP’s Human Development Index.” Journal of International Development 5: 183–92. doi:10.1002/jid.3380050210. McMichael, a, Robert Scholes, and Manal Hefny. 2005. “Linking Ecosystem Services and Human WellBeing.” Ecosystems and Human Well-Being : Our Human Planet, 43–60. http://www.maweb.org/documents/document.341.aspx.pdf. McPhearson, T., Hamstead, Z. a & Kremer, P. (2014) Urban ecosystem services for resilience planning and management in New York City. Ambio, 43, 502–15. McShane, Thomas O., Paul D. Hirsch, Tran Chi Trung, Alexander N. Songorwa, Ann Kinzig, Bruno Monteferri, David Mutekanga, et al. 2011. “Hard Choices: Making Trade-Offs between Biodiversity Conservation and Human Well-Being.” Biological Conservation 144: 966–72. doi:10.1016/j.biocon.2010.04.038.
Page 29 of 32
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
MILLENNIUM ECOSYSTEM ASSESSMENT. 2005. Ecosystems and human well-being: Synthesis. Washington, DC: Island Press Morse, Stephen. 2003. “Greening the United Nation’s Human Development Index?” Sustainable Development 11: 183–98. doi:10.1002/sd.219. Nelson E., Mendoza G., Regetz J., Polasky S., Tallis H., Cameron D.R., Chan K.M.A., Daily G.C., Goldstein J., Kareiva P.M., Lonsdorf E., Naidoo R., Ricketts T.H. & Shaw M.R. (2009). Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Frontiers in Ecology and the Environment, 7, 4-11. Nelson, E. J., P. Kareiva, M. Ruckelshaus, K. Arkema, G. Geller, E. Girvetz, D. Goodrich, V. Matzek, M. Pinsky, W. Reid, M. Saunders, D. Semmens, and H. Tallis. 2013. Climate change's impact on key ecosystem services and the human well-being they support in the US. Frontiers in Ecology and the Environment 11:483-893. Nelson, E., G. Mendoza, J. Regetz, S. Polasky, H. Tallis, D. R. Cameron, K. M. A. Chan, G. C. Daily, J. Goldstein, P. M. Kareiva, E. Lonsdorf, R. Naidoo, T. H. Ricketts, and M. R. Shaw. 2009. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Frontiers in Ecology and the Environment 7:4-11. Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, Dr., Chan, K.M., Daily, G.C., Goldstein, J., Kareiva, P.M., Lonsdorf, E., Naidoo, R., Ricketts, T.H. & Shaw, Mr. (2009) Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Frontiers in Ecology and the Environment, 7, 4–11. Nelson, G.C., Janetos, A., Bennet, E. (2005). Drivers of change in ecosystem condition and services. In (Carpenter, S. R., Pingali, L. P., Bennett, M. E., and Zurek, M.B. eds) Scenarios Assessment of the Millennium Ecosystem Assessment. . Island Press. London. Chapter 7, pages 174-222. Norton L., Elliott J.A., Maberly S.C. & May L. (2012). Using models to bridge the gap between land use and algal blooms: An example from the Loweswater catchment, UK. Environmental Modelling & Software, 36, 64-75. Olander, L., and L. Maltby. 2014. Mainstreaming ecosystem services into decision making. Frontiers in Ecology and the Environment 12:539-539. Pagella, T.F. & Sinclair, F.L. (2014) Development and use of a typology of mapping tools to assess their fitness for supporting management of ecosystem service provision. Landscape Ecology, 29, 383–399. Palomo, I. & Montes, C. (2011) Participatory Scenario Planning for Protected Areas Management under the Ecosystem Services Framework : the Doñana Social-Ecological System in Southwestern Spain. Ecology and Society, 16, 23. Palomo, I., C. Montes, B. Martín-López, J. A. González, M. García-Llorente, P. Alcorlo, and M. R. G. Mora. 2014. Incorporating the Social–Ecological Approach in Protected Areas in the Anthropocene. Bioscience. Palomo, I., Martín-López, B., Potschin, M., Haines-Young, R. & Montes, C. (2013) National Parks, buffer zones and surrounding lands: Mapping ecosystem service flows. Ecosystem Services, 4, 104–116. Paul K.I., Reeson A., Polglase P., Crossman N., Freudenberger D. & Hawkins C. (2013). Economic and employment implications of a carbon market for integrated farm forestry and biodiverse environmental plantings. Land Use Policy, 30, 496-506. Pauly, D., Christensen, V., Walters, C., 2000. Ecopath, Ecosim, and Ecospace as tools for evaluating ecosystem impact of fisheries. ICES J. Mar. Sci. 57, 697–706. doi:10.1006/jmsc.2000.0726 Power, Alison G. 2010. “Ecosystem Services and Agriculture: Tradeoffs and Synergies.” Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 365: 2959–71. doi:10.1098/rstb.2010.0143. Raudsepp-Hearne, Ciara, Garry D. Peterson, Maria Tengö, Elena M. Bennett, Tim Holland, Karina Benessaiah, Graham K. MacDonald, and Laura Pfeifer. 2010. “Untangling the Environmentalist’s Paradox: Why Is Human Well-Being Increasing as Ecosystem Services Degrade?” BioScience. doi:10.1525/bio.2010.60.8.4. Page 30 of 32
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Renaud, F. G., K. Sudmeier-Rieux, and M. Estrella, editors. 2013. The Role of Ecosystems in Disaster Risk Reduction. United Nations University Press, Tokyo. Reyers, Belinda, Reinette Biggs, Graeme S Cumming, Thomas Elmqvist, Adam P Hejnowicz, and Stephen Polasky. 2013. “Getting the Measure of Ecosystem Services: A Social–ecological Approach.” Frontiers in Ecology and the Environment 11 (5): 268–73. doi:10.1890/120144. Ring, Irene, Bernd Hansjürgens, Thomas Elmqvist, Heidi Wittmer, and Pavan Sukhdev. 2010. “Challenges in Framing the Economics of Ecosystems and Biodiversity: The TEEB Initiative.” Current Opinion in Environmental Sustainability. doi:10.1016/j.cosust.2010.03.005. Ringold, Paul L., James Boyd, Dixon Landers, and Matt Weber. 2013. “What Data Should We Collect? A Framework for Identifying Indicators of Ecosystem Contributions to Human Well-Being.” Frontiers in Ecology and the Environment 11: 98–105. doi:10.1890/110156. Rosenthal, A., et al. (2014). "Process matters: a framework for conducting decision-relevant assessments of ecosystem services." International Journal of Biodiversity Science, Ecosystem Services & Management: 1-15. Ruckelshaus, M., McKenzie, E., Tallis, H., Guerry, A., Daily, G., Kareiva, P., Polasky, S., Ricketts, T., Bhagabati, N., Wood, S. a. & Bernhardt, J. (2013) Notes from the field: Lessons learned from using ecosystem service approaches to inform real-world decisions. Ecological Economics. Scholes, R., Reyers, B., Biggs, R., Spierenburg, M. & Duriappah, a. (2013) Multi-scale and cross-scale assessments of social–ecological systems and their ecosystem services. Current Opinion in Environmental Sustainability, 5, 16–25. Schulp C.J.E., Alkemade R., Klein Goldewijk K. & Petz K. (2012). Mapping ecosystem functions and services in Eastern Europe using global-scale data sets. International Journal of Biodiversity Science, Ecosystem Services & Management, 8, 1-13. Schulp C.J.E., Thuiller W. & Verburg P.H. (2014). Wild food in Europe: A synthesis of knowledge and data of terrestrial wild food as an ecosystem service. Ecological Economics, 105, 292-305. Schulp, C. J. E., B. Burkhard, J. Maes, J. Van Vliet, and P. H. Verburg. 2014. Uncertainties in Ecosystem Service Maps: A Comparison on the European Scale. PLoS ONE 9:e109643. Secretariat of the Convention on Biological Diversity. (2014) Global Biodiversity Outlook 4. Montréal,. Sharp, R., et al. (2014). InVEST User’s Guide. Stanford, CA, The Natural Capital Project, Stanford University. [url: http://ncp-dev.stanford.edu/~dataportal/invest-releases/documentation/current_release] Sielhorst, S., Molenaar,J.W. & Offermans,D. (2008) Biofuels in Africa. An assessment of risks and benefits for African wetlands. Commissioned by Wetlands International Stiglitz, Sen and Fitoussi 2009. Report by the Commission on the Measurement of Economic Performance and Social Progress. Paris: Commission on the Measurement of Economic Performance and Social Progress. Sukhdev, Pavan, Joshua Bishop, patrick ten Brink, Haripriya Gundimeda, Katia Karousakis, Pushpam Kumar, Carsten Neßhöver, et al. 2009. TEEB - The Economics of Ecosystems & Biodiversity: Climate Issues Update. Source. Sutherland, W.J., Gardner, T. a., Haider, L.J. & Dicks, L. V. (2013) How can local and traditional knowledge be effectively incorporated into international assessments? Oryx, 48, 1–2. Syrbe, R.-U., and U. Walz. 2012. Spatial indicators for the assessment of ecosystem services: Providing, benefiting and connecting areas and landscape metrics. Ecological Indicators 21:80-88. Taylor, P., Martínez-harms, M.J. & Balvanera, P. (2012) Methods for mapping ecosystem service supply: a review. International Journal of Biodiversity Science, Ecosystem Services & Management, 37–41. Tengö, M., Brondizio, E.S., Elmqvist, T., Malmer, P. & Spierenburg, M. (2014) Connecting Diverse Knowledge Systems for Enhanced Ecosystem Governance: The Multiple Evidence Base Approach. Ambio, 579–591. Tuomisto, Hanna. 2012. “An Updated Consumer’s Guide to Evenness and Related Indices.” Oikos 121: 1203–18. doi:10.1111/j.1600-0706.2011.19897.x.
Page 31 of 32
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
Venter, O., Meijaard,E., Possingham,H., Dennis,R., Sheil,D., Wich,S., Hovani,L. & Wilson,K. (2009) Carbon payments as a safeguard for threatened tropical mammals. Conservation Letters, 2, 123-129. Vervoort, J.M., Kok, K., Beers, P.-J., Van Lammeren, R. & Janssen, R. (2012) Combining analytic and experiential communication in participatory scenario development. Landscape and Urban Planning, 107, 203–213. Vihervaara, P, M Ronka, and M Walls. 2010. “Trends in Ecosystem Service Research: Early Steps and Current Drivers.” Ambio 39: 314–24. doi:DOI 10.1007/s13280-010-0048-x. Villa, Ferdinando, Kenneth J. Bagstad, Brian Voigt, Gary W Johnson, Rosimeiry Portela, Miroslav Honzák, and David Batker. 2014. “A Methodology for Adaptable and Robust Ecosystem Services Assessment.” PloS One 9 (3): e91001. doi:10.1371/journal.pone.0091001. Walters, C., Christensen, V., Pauly, D., 1997. Structuring dynamic models of exploited ecosystems from trophic mass-balance assessments. Reviews in Fish Biology and Fisheries 7, 139–172. doi:10.1023/A:1018479526149 Willemen, L., A. Veldkamp, P. H. Verburg, L. Hein, and R. Leemans. 2012. A multi-scale modelling approach for analysing landscape service dynamics. Journal of Environmental Management 100:86-95. Willemen, L., E. G. Drakou, M. B. Dunbar, P. Mayaux, and B. N. Egoh. 2013. Safeguarding ecosystem services and livelihoods: Understanding the impact of conservation strategies on benefit flows to society. Ecosystem Services 4:95-103. Willemen, L., L. Hein, and P. H. Verburg. 2010. Evaluating the impact of regional development policies on future landscape services. Ecological Economics 69:2244-2254. Wong, C. P., B. Jiang, A. P. Kinzig, K. N. Lee, and Z. Ouyang. 2014. Linking ecosystem characteristics to final ecosystem services for public policy. Ecology Letters:n/a-n/a.
Page 32 of 32
First Order Draft
6
IPBES Deliverable 3c
DO NOT CITE
Linking and harmonizing scenarios and models across scales and domains Coordinating Lead Authors:William W. L. Cheung and Carlo Rondinini Lead Authors: Ram Avtar , Marjan van den Belt, Thomas Hickler, Jean Paul Metzger, JörnScharlemann, Ximena Velez-Liendo, Tianxiang Yue ) Contributing Authors: Rob Alkemade, Mike Harfoot, Qi-Quan LI, Xiao-Fang Sun, Angela Wilkinson, Felix Eigenbrod
5
10
Key messages
15
Decision makers either at local or global level require knowledge about their actions within a domain (social, economic or ecological components/element) and across interconnected domains. Models and scenarios integrate knowledge and techniques across temporal and spatial scales, as well as among dynamic societal economic and natural systems to address these complex challenges and guide decisions.
20
The integration of models and scenarios requires compatibility; such process is referred to as ‘harmonization’. Harmonization allows comparisons across models and scenarios, which is a necessary step to understand the uncertainty around the possible outcomes of the complex interactions between drivers, biodiversity and ecosystem services. Processes of harmonization include upscaling/downscaling across space and time, and/or standardization of model metrics. To integrate between domains, we need to improve our understanding of the direct and indirect causalities between biodiversity and ecosystem services.
25 Different models or scenarios are linked through feeding outputs of one model as input to another model. Models and scenarios can also link through the use of multi-dimensional matrix to provide a more holistic assessment. 30
35
40
It is important to identify and, if possible, to standardize indicators (biodiversity, ecosystem services and human wellbeing) to facilitate linking and harmonizing models and scenarios across different domains. Furthermore, it is equally important to recognize the potential uncertainties associated with the lack of understanding of the processes that generate ecosystem functions and services and how comparing models and scenarios could improve their utility for decision makers through understanding their strengths and weaknesses. We need to build communities of multi-disciplinary researchers and practitioners to harmonize and link across models, scales and domains.
6.1
Importance of linking and harmonizing models and scenarios
Models and scenarios are important tools to understand and communicate effects of natural and human drivers on biodiversity and ecosystem services (Ch. 1, 2, 3, 4). Models are abstractions of
Page 1 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
reality, while scenarios are plausible descriptions of how the future may develop based on a coherent and internally consistent set of assumptions about key driving forces (e.g., rate of technological change, prices) and relationships (see Chapter 1: 1.2.4 & 1.2.5; IPCC 2014: Glossary). The temporal, spatial, and social organization scales that modelling and scenario assessments focus on are generally context-specific. However, biodiversity and ecosystem services and their drivers are interconnected, and span across scales, geographical and disciplinary domains, or elements (the terms “element” and “component” are also used in the IPBES framework, IPBES – see Chapter 1; here, element, component and domain are referred to interchangeably). Thus, linking models and scenarios at different scales and across domains is an important step in advancing our understanding of how the human subsystem may sustainably operate within planetary boundaries (Rockström et al. 2009; Mace et al. 2012). Models are applied to study ecosystem stocks and flows and the implications on biodiversity and management in a social-ecological system e.g., a model linking fish population dynamics (Nature) with the provision of food biomass for fisheries (Nature’s benefits to people). In contrast, scenarios describe the future changes of the system or a subset of the system given the assumption of direct and indirect drivers while projection focuses on potential future evolution of a quantity or set of quantities conditional under a scenario, often computed with the aid of a model, e.g. the projection of change in mean species abundance in a region following a given change in temperature and precipitations under a global greenhouse gas emission scenario. Decision makers, from individuals to global institutions, are unlikely to have knowledge about the entirety of impacts of their chosen actions within a domain and across multiple, interconnected domains. To help identify the (positive and negative) impacts of an action across interconnected elements, models/scenarios can be used to understand the interactions and feedbacks when linking multiple elements/domains of social, economic and natural systems. Action impacts affect individual elements in different and often unexpected ways across spatial and temporal scales, as well as potentially affecting multiple elements. For example, damming a river impacts fish up- and downstream of the dam (migration barrier; spatial impacts), immediately and in the longer term (altered water flow, sediment accumulation in reservoir; temporal impacts), and impacts fish, aquatic and terrestrial plants, and people (multiple organizational scales). Models and scenarios that integrate feedbacks and tradeoffs across temporal and spatial scales and among dynamic social, economic and natural systems can help address complex environmental challenges and guide decision making (Carpenter et al. 2006). To link models or scenarios, they need to be made compatible or consistent with one another; such process is referred to as ‘harmonization’. Harmonization also enables comparisons across models and scenarios, which is a necessary step to understand the uncertainty around the possible outcomes of the complex interactions between drivers, biodiversity and ecosystem services. For some decision contexts, multiple models/scenarios exist, or could be developed, that provides different outputs or advice to decision makers. Model or scenario outputs may differ because they (a) use different input data (e.g. different biophysical layers for species distribution modelling); (b) produce different outputs (e.g. carbon ecosystem service models can output carbon stocks or carbon sequestration); (c) were developed to address subtly different questions for different audiences (e.g. composition and function of biodiversity, or the four different IPCC RCP scenarios); (d) represent different components/aspects/elements within the model/scenario (e.g. process-based biodiversity
Page 2 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
models may incorporate metabolism, reproduction, growth, dispersal, mortality); (e) cover different spatial and/or temporal scales; or (f) use different methodologies/techniques to build model/scenario (e.g. quantitative, qualitative, inductive, deductive, statistical, process-based). 5
10
15
20
The main reason to link and harmonize across models and scenarios is to aid decision makers who are often unable to evaluate multiple and potentially contradictory outputs. Harmonizing model/scenario inputs and outputs enables their intercomparison. In the Coupled- Model Intercomparison Project (CMIP, part of IPCC assessment reports), for example, climate models explore the same sets of greenhouse gas emission scenarios and other forcings to produce common outputs such as annual mean atmospheric temperature (Taylor et al. 2011). Recently, also the climate impact modelling community attempted to compare projected impacts across different sectors (e.g. agriculture, hydrology, carbon cycling and biome shifts) using one common set of driving variables and modelling protocol within the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP, Schellnhuber et al. 2011). Additional sectors will be included in the next phase, also including impacts on biodiversity (https://www.pik-potsdam.de/research/climate-impacts-andvulnerabilities/research/rd2-cross-cutting-activities/isi-mip). Comparing among models helps to identify differences and their causes, and ultimately assess the model/scenario quality. Models/scenarios can be harmonized in multiple ways, by using: standardized inputs (e.g. all Integrated Assessment Models used in IPCC AR5 use the same harmonized land use data, (Hurtt et al. 2011), agreed output metrics, evaluation or benchmarking against common observational data sets (e.g. GCMs to be included in IPCC need to be able to hind cast historic temperature trends, derived from multiple sources), or specifying the key components that need to be represented in the model/scenario.
25
30
35
40
45
Figure 6.1: Relationship between decision stakes and system uncertainty (Funtowicz and Ravetz, 1993). Linking and harmonizing models/scenarios may not be appropriate in every decision context, for instance, when the causality of links across elements is poorly understood. Further, linking too many components either statically or dynamically may create complex models complex and become unhelpful for decision making. Thus, the amount of linkage among models/scenarios needs to be tailored to the decision at stake (Figure 6.1). Harmonizing scenarios may reduce the variability of potential future conditions taken into account in a decision. By a priori standardising inputs and components as well as ensuring validation against a standard dataset, models/scenarios that are projecting the “unknown unknowns” (low frequency, high impact events) will be excluded (Levin 2003) (e.g., the 2008 financial crisis or abrupt climate change). Although standardising likely reduces the uncertainty around estimates (by removing outliers), many models/scenarios will give similar
Page 3 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
outputs which may be more precise but not necessarily accurate and therefore less policy relevant. Furthermore, some issues cannot be reduced to a few metrics, e.g. for biodiversity multiple output metrics are required to assess impacts fully (e.g. species richness, abundance and diversity) (Schipper et al. in prep.). 5
10
This chapter aims to: a) summarize existing approaches and initiatives that link and harmonize models and scenarios across scales and domains; b) identify knowledge gaps and propose possible ways for IPBES to undertake cross-scale/domain linkages and harmonization to assess biodiversity and ecosystem services; c) discuss relevance to policy-making. Models for biodiversity and ecosystem services run at a wide variety of time scales (from seasons to years) depending on the elements/domains and processes that they represent (see section 6.2). For scenarios, our discussion focuses on both short (10 – 15 years) and long (multi-decadal) time scales, and present case studies selected across a wide variety of domains and applications to showcase approaches to tackling complex issues.
15
6.2. Approaches to linking and harmonizing models and scenarios of biodiversity and ecosystem services 20
25
Models or scenarios related to biodiversity and ecosystem services that are developed for different spatial and temporal scales, biomes and/or organizational domains are linked or harmonized through different means (Table 6.1). Specifically, different models are linked together (also referred to as coupling) through feeding outputs of one model as input to another model. Models and scenarios can also link through combining outputs qualitatively (e.g., through narratives or description of storylines) or quantitatively through the use of multi-dimensional matrix. Models and scenarios describing different components of biodiversity and ecosystem services are also combined quantitatively or qualitatively to provide a more holistic assessment. Processes of harmonization include upscaling and downscaling models across space and time, standardization of modelling methods, standardization of model and scenario metrics, and benchmarking. Table 6.1: Summary of different approaches to linking and harmonizing models and scenarios.
Linking
Harmonizing
Approaches
Model
Scenario
Input-output, one-way coupling (offline)
x
Input-output, two-way coupling (dynamic feedbacks and interactions among elements/domains)
x
Combining outputs qualitatively using multi-dimensional matrix
x
x
Cross-scale (upscaling, downscaling, across dimensions of biodiversity, ecosystems, ecosystem services, human and society)
x
x
Standardization of methods
x
Standardization of metrics (input and/or output)
x
Benchmarking
x
x
30
Page 4 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
6.3 Linking models and scenarios of biodiversity and ecosystem services 6.3.1 Model coupling through input-output 5
10
15
20
25
30
35
40
Models representing different components of the social-ecological system related to biodiversity and ecosystem services and their drivers are linked through either one-way (off-line) or two-way (dynamic) coupling. In both cases, outputs from one model feed into another model as inputs. However, in the case of one-way coupling, the linking is unidirectional. For example, in modelling the effects on changes in ocean conditions (temperature, primary productivity, oxygen level and acidity, and the resulting species range shifts) on marine ecosystems in the Northeast Pacific coasts, Ainsworth et al. (2011) took simulated changes in ocean conditions from coupled-ocean- atmospheric earth system model and projected range shifts from species distribution models as inputs (forcing factors) in trophodynamic foodweb models to simulate the effects of multiple CO 2 related drivers on marine ecosystems and fisheries yields. For example, Visconti et al. (2015) used scenarios of climate and land use change to project species distributions into the future, and assembled these projections into policy-relevant indicators of biodiversity change (Box 6.1). In contrast, two-way coupling includes feedbacks of inputs-outputs between models. For example, a marine ecosystem model, Atlantis (Fulton et al. 2011), links model components describing ocean biogeochemistry, lowertrophic level ecosystem, upper-trophic level ecosystem and human activities (with a focus on fishing) in which outputs from the components mutually affecting one another directly or indirectly over space and time. In the terrestrial environment, one-way linking has been extended to complex chains of models and scenarios. The choice of coupling methods depends on the dynamics of the modelled systems and the objectives of the models. One-way coupling is simpler relative to dynamic two-way coupling as the responses of the modeled system are more predictable (e.g., predicting changes in species distributions driven by climate model outputs). On the other hand, non-linear system dynamics and feedback between model domains cannot be directly revealed with models that are linked by one- way coupling. Two-way coupling is more realistic for understanding social-ecological systems where feedbacks and resulting non-linear responses are common. However, it is more technically difficult, particularly if components operate at different temporal and spatial time-scales. The model responses are also less predictable and often result in large internal variability. Regarding potential linkages between models representing different aspects of biodiversity (from the genetic to the ecosystem level [Chapter 3]) and models or modelling frameworks adopted to develop ecosystem service scenarios (Chapter 4), only a limited fraction of the available biophysical model types has been extensively used in ecosystem service scenario work (Table 6.2). Ecosystem service scenarios have often been based on rather simple proxies or indicators for biodiversity and the associated ecosystem services (Chapter 4.1.1). The state-of-the-art in biophysical modelling has often not been used. Proxy-based approaches, however, have the advantage that they are easier to handle and amendable to participatory processes. Furthermore, quantitative model results and qualitative expert knowledge can be integrated (Chapters 4.1.1 and 4.1.3).
Page 5 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Table 6.2: Examples of major model types and models that can be used to project future changes in biodiversity (from the species to the ecosystem level), main associated ecosystem services, and models or modelling frameworks that have been or can be linked to the outputs of biodiversity models. The ecosystem service model categories have been adopted from Crossmann et al. (2013). Biodiversity models
Inputs (most common)
Outputs
Examples
Ecosystem services
Ecosystem service scenario examples
Species distribution models
Climate, land cover, soil types, ocean biophysics and biogeochemistry
Species ranges
BIOMOD, Maxent, Aquamap, DBEM
Provisioning1 Cultural and
Exploited marine species (Cheung et al. 2010) Wild food provisioning (Schulpet al. 2014)
Expert-knowledgebased species-habitat relationship
Vegetation/ land cover types
Local occupancy of particular species
Igleciaet al. (2012)
Cultural and Amenity
-
Community-based biodiversity change models Abundance models
Multiple environmental layers, including, e.g., land cover and use, climate and vegetation type
Mean species abundance relative to undisturbed ecosystem or changes in community types, groups of species and community characteristics Species population dynamics and/or abundances
GLOBIO3, Ferrier and Guisan (2006) GLOBIO, PREDICTS
Provisioning1 Cultural and Amenety2
Fish yields (Blanchard et al. 2012)
Population dynamics models
Multiple environmental layers
Population dynamics of individual species
KramerSchadtet al. (2005)
Cultural and Amenity
-
Richness models
Multiple environmental Species richness layers (at smaller scales) or available habitat area (species-area relationship) or as for species distribution models (stacking results from individual species) Area of available habitat
Ferrier and Provisioning1 Guisan (2006), van Vuurenet al. (2006) Rahbek et al. (2007), Algar et al. (2009), Calabrese et al. (2014)
-
Functional trait models
Climate, soil types, atmospheric CO2
JEDI, aDGVM2
-
Trait composition/diversity
Amenity2
Provisioning1 Regulation3 Habitat4
Landscape models
Forestry models
Dynamic Global Vegetation Models
Climate, soil types, land use
Landscape-level land cover and vegetation structure
Landclim
Climate, soil types, land use
Timber yield, C cylce, habitat quality
3-PG, EFISCEN
Provisioning1
Climate, land cover and use, soil types, atmospheric chemistry (e.g. CO2 concentration, nitrogen deposition)
Biome distribution, vegetation structure, plant functional type diversity, NPP, C, N, P and water cycles
LPJ, SDGVM, MC1
Provisioning1
Provisioning1 Regulation3 Habitat4
Regulation2
Habitat for biodiversity (Bryan and Crossmann 2013) Carbon sequestration (Bryan and Crossmann 2013, Paul et al. 2013, Bryan et al. 2014) -
Regulation 2 Habitat3
Page 6 of 34
First Order Draft Ecosystem or biogeochemistryprocess models
Agricultural models
Hydrology models
1
IPBES Deliverable 3c
DO NOT CITE
Climate, soil types, land cover, atmospheric composition (e.g. CO2, nitrogen deposition), Climate, land cover, soil types, atmospheric chemistry, ocean biogeochemistry and foodweb Climate, crop management, soil types, atmospheric chemistry
NPP, C, N, P and water cycles
Century, TEM, TOPAZ, Ecopath with Ecosim, Atlantis
Provisioning Regulation and provision services
Fish yields (Christensen et al.2015) Carbon cycle (Wenzel et al. 2014)
Crop yields
GEPIC, APSIM, LPJmL
Provisioning
Food production (Keating et al. 2003, Bryan and Crossmann 2013)
Climate, land cover, soil types
Water cycle
WaterGap2,
Provisioning Regulating
Fresh water supply (Bryan and Crossmann 2013)
Food, water, raw materials, genetic, medicinal and ornamental resources
2
Aesthetic information, opportunities for recreation and tourism, inspiration for culture, art and design, spiritual experience, information for cognitive development 3 Air quality, climate, moderation of extreme events, water flows, waste treatment, erosion prevention, maintenance of soil fertility, pollination, biological control 4 Maintenance of life cycles and genetic diversity, cultural and amenity services
5
10
15
20
25
References for biodiversity models or modelling frameworks: BIOMOD (Thuiller et al. 2009), Maxent(Phillips and Dudík 2008), GLOBIO (Alkemade et al. 2009), JEDI (Pavlick et al. 2013), aDGVM2 (Scheiter et al. 2013), Landclim (Schumacher and Bugmann 2006), 3-PG (Landsberg and Waring 1997), EFISCEN (Schelhaaset al. 2007), LPJ (Sitch et al. 2003), SDGVM (Woodward and Lomas 2004), MC1 (Gonzalez et al. 2010), Century (Parton et al. 2010), TEM (McGuire et al. 1992) Ecopath with Ecosim (Christensen and Walters 2004), Atlantis (Fulton et al. 2011), GEPIC (Liu et al. 2007), APSIM (Keating et al. 2003), LPJmL (Bondeau et al. 2007), WaterGap2 (Alcamo et al. 2003). Species distribution modelling techniques have only rarely been used to estimate current and potential future ecosystem services (e.g. provisioning of wild foods, Schulp et al. 2014, see also Chapter 4), but mainly for projecting future changes in ecosystem service supplies. Hanewinkel et al. (2013) used species distribution models to project future range shifts for major tree species and associated changes in economic revenues across Europe. In the marine realm, species distribution models have been applied to project future changes in fisheries catch potential (Cheung et al. 2010). Expert-knowledge-based species-habitat relationships (Chapter 3.3.2.5) and population dynamics models have, to our knowledge, only been used to estimate the probability of occupancy, distribution or population dynamics of particular species (in the examples in Table 6.2 for birds [Chapter 3.3.2.5], amphibians (Ficetola et al. 2015), mammals (Rondinini et al. 2011) and Eurasian lynx [Lynx lynx]) but without in-depth consideration of associated ecosystem services. Communitybased biodiversity models (Chapter 3.3.2.6) simulate more general biodiversity patterns, and the GLOBIO3, which simulates mean species abundance relative to their abundance in undisturbed ecosystems, has also been implemented within the Integrated Assessment Model IMAGE3.0 (Box 6.2). Species richness patterns can be simulated with a variety of approaches (Chapter 3). As in the case of functional trait models, the results of such models have not been interpreted in terms of ecosystem service supplies (see also Verkerk et al. 2011, Boettcher et al. 2012). Dynamic Global Vegetation Models (DGVMs) simulate a number of ecosystem functions that represent ecosystem services (e.g. carbon storage) or are closely linked to these (e.g. vegetation type – provisioning of Page 7 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
habitat), and the results have been interpreted in terms of ecosystem service supplies (e.g. Doherty et al. 2009). The DGVM LPJmL (which includes major global crop types) has also been implemented in IMAGE3.0, but we are not aware of any DGVM application within a more social-science-based ecosystem service scenario framework. This partly reflects a scale mismatch; most ecosystem service scenario work concerns smaller scales. DGVMs that are adapted to the regional scale (e.g. Hickler et al. 2012, Seiler et al. 2014) would be more suitable for smaller scale ecosystem service scenario development. Crop and hydrology models have been widely used in ecosystem service scenario work (Table 6.2), but these models don’t focus on biodiversity, even though hydrology models increasingly account for the effects of vegetation or plant functional type composition on hydrological cycling (e.g. Rost et al. 2008). The question of whether biodiversity and ecosystem service models should be directly linked depends on the research objectives and societal demands. Nevertheless we think that not all useful linkages have been utilized and more direct linkages have great potential. The ARIES ecosystem service modelling framework (Chapter 4.2.1) presents an important advance as it allows linking a variety of models in a very flexible framework (Villa et al. 2014). Also among biodiversity or biophysical models based on natural science (first column in Table 6.2), not all potential links have been utilized. Most species distribution models, for example, use climatic variables, land cover, and in rare cases, soil types as input, but a biome shift could influence the occurrence of species much more than the a change in climate. In spite of this obvious link, projected changes in biome distribution or vegetation structure, as simulated, by DGVMs, have only rarely been used in species distribution models (Linder et al. 2012). Most examples above concern one-way coupling between models. Two-way coupling or full socioecological systems modelling has only rarely been achieved. The rare examples include Integrated Assessment Models (IAMs), such as IMAGE3.0 (Box 6.2), which only represent very general system characteristics and are of limited use for regional or local policy making or stakeholders. Integrated assessment models combine components (sub-models) representing the future development of human societies, including major sectors such as energy use, industrial development, land use, that are important for making projections about the future of human and natural ecosystems (Harfoot et al. 2013). Currently, the main applications of IAMs are on modelling climate change and effects of climate mitigation. In most IAMs, their sub-models, including both natural and human subsystems, are linked although dynamic linkages are not commonly represented in most IAMs (Harfoot et al. 2013). An example of natural systems sub-models in an IAM is the linkages between hydrological models providing inputs regarding water and nutrient supply into terrestrial vegetation models. For human systems sub-models, it includes, for example, component representing the energy sectors that capture the demand and supply of energy that are links to industrial development, population demands and commodity prices. There are also components that link natural-human systems such as food production, linking vegetation, land-use with societal demand, energy sources (particularly from bio-energy crops) and commodity prices. IAM provides a framework for linking models to represent complex social-ecological systems, however, there are existing gaps in apply IAMs to address questions related to biodiversity and ecosystem services (Harfootet al. 2013).Currently, representation of biodiversity in IAMs is only limited to terrestrial vegetation and existing models focus strongly on climate change-related questions. Thus, using the IAM framework to address a broader range of questions related to Page 8 of 34
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
biodiversity may require further works in incorporating model components that represent more ecological processes e.g., population dynamics or biogeography of important groups of animals. Also, existing IAMs are structured differently and often designed to address different questions, contributing to a major source of uncertainties with the outputs and scenarios generated by IAMs. Model intercomparisons allow readers to identify the levels and causes of these differences (see section 6.5).
6.3.2. Combining models and scenarios through multi-dimensional matrix 10
15
20
25
30
Models that are complementary in representing different domains and scales in relation to assessing biodiversity and ecosystem services can be integrated by arranging the model outputs or scenario narratives in a multi-dimensional matrix. Different aspects or scales of each domain of the models or scenarios are arranged in one axis, with the outputs of interest being described in each element of the matrix (Figure 6.2). Thus, collectively, each element contains outputs from different models and scenarios representing different domains or scales, providing more comprehensive description of biodiversity and ecosystem services.
Figure 6.2: A hypothetical example using a multi-dimensional matrix to combine outputs and projections from different models related to marine biodiversity and fisheries yield. The (sub-) models include coupled oceanatmospheric global climate models projecting changes in ocean conditions, ecological models of fish population dynamics and marine biodiversity, models representing fishing behaviour and models projecting the economy. The matrix provides a way to summarize and integrate outputs from different models to develop narratives and scenarios that account for different aspects of the social-ecological system.
The narratives generated by models and scenarios representing different domains can be integrated to describe potential changes in social-ecological systems or a subset of the systems. In some cases, this relies on expert knowledge to synthesize information from different models and scenarios. This approach is particularly useful in participatory modelling and scenario development exercises in which experts or developers of models or scenarios on particularly domains and/or scales are contributing to development of more holistic assessments.
Page 9 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
6.4 Harmonizing models and scenarios of biodiversity and ecosystem services 6.4.1. Harmonizing models across scales 5
10
15
20
25
Harmonizing models to assess the status and trends, and project future changes in biodiversity and ecosystem services require synthesizing biophysical and socio-economic data and results that are available at different spatial scales. Biodiversity dynamics are determined and controlled by their spatial heterogeneity at various levels (continental, landscape and habitat) and their temporal change (Goetze et al. 2014) (Figure 6.3). Particularly, spatial patterns of biodiversity, species distribution and ecosystem services, such as stocks and flows of water and other living resources, as well as their natural and anthropogenic direct and indirect drivers of change are scale- dependent (Levin 1992), require data that have large spatial extension and for an extended period of time. Moreover, local actions affect the environment globally, and vice-versa, and because the success of global scenario projections will depend on the congruence of scenarios and goals planned at more local scales (Cash et al. 2006). On the other hand, to protect native biodiversity, local conservation actions must be established, but often the required data at a fine-grained resolution is not available (Bombi et al. 2012, Fernandes et al. 2014). Scaling is a widely used method in environmental science to modify the prediction of phenomena at different scales from its initial record or model. Scaling can be done in two different directions: upscaling information from local, fine-grained resolution to global, coarse-grained resolution, or viceversa, downscaling the information. Upscaling usually leads to an increase in the extent, and decrease in the resolution, while downscaling increases the resolution of the data, while losing the extent. In both directions, predictions are associated with errors and uncertainty, which are explored in the next section (6.5). Environmental studies involve dynamic modeling with data sources at various spatial and temporal scales which need to be integrated. Furthermore the output of the model often needs post-processing treatment to modify its scale to fit it to the resolution required for policy decisions. Each component of the research cycle has its own temporal or spatial scale. Scaling methods refers to the transfer of information on spatial and temporal scales.
30
35
40
Scaling is inevitable when modeling biodiversity and ecosystem services, and is necessary for both global and local ecological and environmental planning and adaptive management (Holling 1978). There is frequently a mismatch between availability and the scale of data, model outputs or scenarios descriptions that are needed for biodiversity and ecosystem services assessment at global and regional scales. Available data, models or scenarios are often obtained from sporadic studies, unevenly distributed spatially, cover a short period of time (e.g., snapshot samples), and collected with inconsistent methodologies. Thus, understanding how patterns revealed from these models and scenarios change across scales and how to transfer information among scales, generally referred to as "scaling", is crucial to integrate different components of the social-ecological systems that operate at different spatial scales (Wu and Li 2006). However, scaling between different systems can become complex, or even unnecessary, because of factors such as scale variance and abrupt changes in scaling properties. Furthermore, nonlinear interactions among different components may make the scaling difficult.
Page 10 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 6.3: Schematic illustrations of different scales and levels that are critical in understanding and responding to human-environment interactions (redrawn from Cash et al. 2006).
5
10
15
Spatial scale can be defined considering two main dimensions: grain and extent. Grain refers to the resolution of the dataset, while extent is the size of the observation. More specifically, the spatial grain is the size of the sampling unit and the temporal grain is the frequency of data observation. The extent refers to how large is an area (spatial extent) or period of time (temporal extent). For example, with an Enhanced Thematic Mapper Plus sensor, the spatial grain is 30 meters (for bands 1 to 5), the temporal grain is 16 days (the satellite makes an image of the same place each 16 days), the spatial extent corresponds to a track of 183 km wide, and the temporal extent is the duration of the study (for example, few days, one season, several years). Within the framework of IPBES, we will refer to three spatial extents (global, regional, and national/subnational/local scales) and to short (around 10 years) and long-term (e.g. several decades) temporal extents or time series data. Global extent considers the whole planet, regional extent encompasses several countries, specific ocean basin or large marine ecosystem (for example, the South African region is composed by South Africa, Mozambique, Zimbabwe, Namibia and Botswana), national to local scales refer to space in a same country.
20
25
Spatial and temporal resolution of models interacts with one another. Models with coarse spatial resolution usually do not resolve processes that operate at fine temporal scale, while models at a more local scale will require more fine-grained spatial and temporal resolution data. This is partially due to an intrinsic relationship between space and time, which makes processes acting in more local scales more dynamic (local fast changes are more likely), while those operating at larger spatial scales require larger temporal observation. For example, global scale population dynamic models of fishes do not resolve fine scale behaviuor shift of individuals because of changing local ecological or environmental conditions. Page 11 of 34
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
Understanding the dynamics of the ecosystems and biodiversity requires modelling across scales (spatial and temporal) and both have their own importance. Several models have been proposed and adopted and provide knowledge about potential of ecosystems responses to global change in spatial-temporal dimension (Dietze et al. 2011). Some of these models are static, with snapshot changes (e.g. Computable General Equilibrium), linear with projected changes overtime (e.g. VISIT, Integrated Valuation of Environmental Services and Tradeoffs (InVEST)) and system based models (e.g. World3, Global Unified Meta-model of the Biosphere (GUMBO), Multi-scale Integrated Modeling of Ecosystem Services (MIMES). The accuracy of these models depends on the availability of up-to-date data sets. Considering the fact that ecosystems are temporally- interact), most of the models do not take into consideration different time scales (Dietze et al. 2011). According to Dietze et al. (2011), the performances of ecosystems models vary across time scales, with more recorded uncertainty in annual and diurnal cycles than the daily. But the tendency of looking at the ecosystems and biodiversity at a fine, high temporal scales would also prevent a broad understanding of the changes, heterogeneities and interactions with various environmental and socio-economic factors over large time-frame, hence the need for linking and harmonizing scenarios and models across time. This is necessary for planning both at local and global scale.
6.4.1.1. Spatial scale 20
25
30
35
40
45
Downscaling When fine grained- resolution data is not available, downscaling is a common technique to provide information particularly for local conservation issues or management needs, such as establishing priority conservation areas (Rondinini et al. 2005, Bombi et al. 2012, Fernandes et al. 2014). With the impossibility or high cost for obtaining fine-grained resolution data, downscaling approaches are a possible cost-effective alternative. Furthermore, downscaling is also relevant to incorporate projections of climate models into local conservation. Downscaling is then necessary to align the scales of conservation with the scales of climate change projections (Wiens and Bachelet 2010), allowing, for example, the combination of IPCC climate simulations and scenarios with more regional socio-economic scenarios (Walz et al. 2014). There is a long history of developing downscaling methods for climate data that provides valuable experience for downscaling of biodiversity and ecosystem services models and scenarios (Box 6. 3). The main methods can be categorized into dynamic and statistical approaches. There are different techniques for downscaling, most of them based on statistical relationships between biological data and environmental attributes. Some of these techniques use hierarchical models (Keil and Jetz 2014, Keil et al. 2013), projecting the relationship between coarse- grain species and environmental data to a fine-grain scale using fine-grain environmental (predictor) variables. This method was used with success for downscaling exploited fishes and invertebrates’ distributions in Western Australia (Cheung et al. 2012). A similar approach was used by Barwell et al. (2014) to downscale a coarse-grained (> 100 km2) Odonata atlas data to a more fine-grain (25km2, 4km2 and 1 km2) local scale in mainland Britain. Ten different downscaling models were used for 38 species. Results suggest reasonable estimates of fine-grain occupancy, with varying errors according to species traits. Particularly, high dispersal ability was associated with relatively poor downscaling predictions (Barwell et al. 2014). Furthermore, recent studies showed high predictive performance, including similar results to field observations when downscaling models. For example, Fernandes et al. (2014) estimate the distribution of alien invader species, Keil et al. (2013) for bird species, Sandø et al. Page 12 of 34
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
(2014) with marine ecosystem conditions and Bombi et al. (2012) with sardinian reptiles occurrences. Hierarchical Bayesian modelling frameworks are generating better fine-grain estimation than other more conventional methods (direct approach, point sampling), which usually over predict occurrences (Keil et al. 2013). Those predictions may be even more improved when combined with macroecological relationships (e.g. scale area relationships). Upscaling Environmental problems or the consequences of human activities usually encompass broad spatial and temporal scales, which need global assessments and policy actions. For this reason, it is often necessary to upscale high-quality fine-resolution data to broader scales. Upscaling consists of transferring information from a broad scale to a high resolution scale (Flint and Flint, 2012). Upscaling methods seem more intuitive involving mean values extrapolated over a larger time period or space. Nevertheless aliasing effect can occur when the scaling ratio is poorly sampled and close to the frequency limit of the phenomena. In that case, artifacts may be created and contribute to scaling uncertainties (6.5). Most upscaling approaches use satellite imagery and combines statistical and image processing analyses, with simulation models, and field-based observations (Zhang et al. 2007, Chen et al. 2010, Fu et al. 2014). This has been used to estimate net ecosystem exchange or carbon dioxide fluxes from flux towers to landscape and regional scales (Fu et al. 2014), or to upscale leaf area index (LAI) in terrestrial ecosystems from Arctic landscapes. In this last case, a simple exponential relationship between LAI and the Normalized Difference Vegetation Index (NDVI) obtained with a LANDSAT image was used to upscale LAI values (Williams et al. 2008). This relationship was valid at different scales, with similar prediction errors, and thus can be used, according to the authors, for estimating and upscaling carbon cycling. Other methods have been developed to upscale gross ecosystem production (GEP) from leaf or stand levels to larger regions (ca. 12 km2) taking into account the tree canopy structure (Hilker et al. 2008),
30
35
40
45
using Light Detection and Ranging (LiDAR) images. Results showed a high correlation (r2 between 0.75 and 0.91, p< 0.05) between estimated and measured ecosystem productions. A good fit between upscaled estimated value and field measurements were also obtained with net primary productivity in China, showing that the integration of field data with remote sensing though an ecosystem model can generate reliable data (Zhang et al. 2007). Upscaled estimations can even outcome in better results than those obtained from coarse- grained resolution images, such as MODIS data (Fu et al. 2014), possibly because they can integrate the variability observed at finer scales in the coarse-scale evaluation. Similar results were obtained by Hay et al. (1997) when upscaling forest stand characteristics with image resampling techniques. They showed that appropriately upscaled imagery can represent a more accurate estimation than an image obtained at the upscaled resolution. Cross-scaling Cross-scale interactions make multi-scale modelling approaches particularly important in order not only to consider the different factors that impact species distribution or ecological processes at different scales, but also to integrate their interactions across scales. Particularly, due to the complexity and the uncertainty involved in both up-scaling and downscaling, an intermediate but
Page 13 of 34
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
integrated framework or model that can provide solutions to the problem of scaling would be necessary. Indeed, Boscolo and Metzger (2009) showed that multi-scale models (which considered pattern-process relationships at different extents in a unique model) always performed better than single-scale model to predict the occurrences of bird species in a tropical forest, probably because extinction and recolonization processes that control species occurrences simultaneously act at different scales.
6.4.1.2. Temporal scale 10
15
Time is an important characteristic of the ecosystems and biodiversity, which helps to understand processes, fluctuations and the distribution of species at a specific period. Temporal scale is also important to monitor the shifting of priorities from one ecosystem service to another ecosystem service and also changes with the change in supply of and demand for ecosystem services. Therefore, while studying trade-offs it is important to look into the temporal dimension because the provision of one ecosystem service is reduced as a consequence of another ecosystem services. Tradeoffs at temporal scale allows policy makers to understand the long-term effects of one ecosystem service over another and have enormous impact on future ecosystem services (Rodriguez et al. 2006). Some models are developed at a large temporal scale while some exists within the short time scale (at higher time resolution). Linking and transforming between scales are necessary for holistic understanding of the ecosystem.
20
6.4.1.3. Organizational scale (social aggregation and biodiversity levels)
25
30
35
40
45
Any form of modelling and scenario development and usage stems from and affects a social, organizational context. The desire to understand and make better decisions with regard to this relationship between the human and natural systems causes us to consider linkages between models and scenario for and from models. Following an ecological economics convention, the human sub-system is roughly comprised of social, human and built capital and embedded within the natural system (i.e. natural capital, ecosystems, nature) (Costanza et al., 1997a). However, this is merely one perspective; the underlying (often deeply engrained and unspoken) assumptions and mental models people hold about this relationship drives the type of models and scenarios individuals and their social/cultural contexts will accept and develop (Hamilton, 2011) (see also IPBES Deliverable 3d on ‘diverse values and valuation’). As discussed earlier in this chapter, issues of scale relating to space and time needs to be ‘fit for purpose’ in an organizational context. As models refer ‘any abstract reflection of reality’ (including mental models) we confine this section to ecosystem services models; i.e. the models that stem from the desire to highlight and make visible the benefits and well-being people derive from natural capital. An ecosystem-services approach can be considered a strength-based, organizing principle linking natural and human systems (Costanza et al., 1997b; Daily, 1997; Millennium Ecosystem Assessment, 2005; Braat and de Groot, 2012). As a trans-disciplinary approach, the ecosystem services concept and its associated tools, has undergone a rapid evolution (including the use of model building) since the 90’s. Economic ‘benefit transfer methodology’ was effectively used to highlight value from ecosystems that is not visible in the market and therefore often neglected. For example, Costanza et al. (1997, 2014) calculated an annual flow of value of ecosystem services derived from the stock of global natural
Page 14 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
capital to conservatively be twice that of global Gross Domestic Product (GDP). In addition, these values were spatially displayed on a global map. While the authors never claimed robustness or accuracy, this global ecosystem service value in relation to GDP, drew attention from policymakers, business and a wider audience and sparked a fierce debate. In fact, Costanza et al. (1997) laid out the many challenges that would have to be overcome for this value to be identified and captured for management purposes. Since then, numerous data bases have been developed to support valuation of ecosystem services, for example Earth Economics (http://www.eartheconomics.org/). For example, the draft chapter 3 on ‘Ecosystem Services’ of the 1st World Oceans Assessment identified 14 databases that directly or indirectly are used in conjunction with and ecosystem services approach. Given that the marine ecosystems are lagging in application of ecosystem services approaches, the vast amount of databases and information available for other ecosystem (often unconnected) can be assessed. While basic value transfer assumes that value per ecosystem type remains constant (e.g. Costanza et al. 1997b, 2006), expert modified value transfers adjust values for local conditions of ecosystems (e.g., Batker et al. 2010) and Natural Capital Project (http://www.naturalcapitalproject.org). Involvement of stakeholders and their different mental models / interests in understanding local dynamics and non-spatial trade-offs between bundles of ecosystem services has been explored though systems thinking and system dynamics e.g Mediated Modelling (van den Belt et al. 2012 ). Values may also be adjusted based on statistical models of spatial and other dependencies (meta regression analysis (e.g. de Groot et al. 2012). This rapid development currently sees spatially explicit dynamic modelling frameworks at multiple scales (e.g. Boumans et al., 2002; Boumans and Costanza, 2007; van den Belt 2009; Boumans and McNally, 2012; Altman et al., 2014). Furthermore, an ecosystem services approach is also inherently multi-scale as ecosystem services can be classified according to their spatial characteristics (Costanza 2008): (1) At a global level, climate regulation, carbon sequestration and storage as well as cultural or existence values don’t depend on people’s proximity, whereas (2) local proximity is relevant for disturbance regulation / storm protection, waste treatment, pollination, biological control and habitat. (3) A directional flow characterises water regulation/ flood protection, water supply, sediment retention/erosion control or nutrient regulation. (4) a point of use is relevant for soil formation, food/forest production and other raw material and finally (5) some ecosystem services and the benefits/values derived from them are related to the manner in which users move in space (and time), e.g. genetics resources, recreational potential and cultural values. Bundling of ecosystem services and variations for possible tradeoffs exist and require pluralism (including multiple modelling approaches), which need to acknowledge values (or participation ) of stakeholders. Due to its complexity the scope for approaches aiming for optimization are limited. Co-evolutionary approaches that acknowledge humans as one species dependent and impacting on ecosystems are more appropriate under high uncertainty (and limited predictability) (e.g. IHOPE http://ihopenet.org/). Up- and down- scaling along a social organizational scale requires awareness of humanly imposed boundaries and conventions, which often are not following an ecosystem logic (O’Brien and Vickerman 2013). Even the distinction between surface and groundwater leads to different spatial extent. In addition, it is not unusual for governing bodies to be guided by multiple ways in which their constitutions are divided in space; e.g. Auckland Council identified 30 different ways in which space is divided for water management (including water supply, water treatment, storm water, river/coastal and ground water protection and various values from interest groups such a Maori culture). Page 15 of 34
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
While most decision making time frames are relatively short term (e.g. election and budgetary cycles), the collective human impact is changing long term ecological trends (e.g. IPCC scenarios of climate change, IHOPE, World3), with local spatial variations. Following an ecosystem services classification using the characteristics of ES provides one option for harmonizing across time scales (Costanza 2008, Table 6.3). Table 6.3: Marine ecosystem services and their spatial scales (Costanza 2008).
1. Global-non proximal (does not depend on proximity) Climate sequestration (NEP) Carbon storage Cultural/Existence value 2. Local Proximal (depends on proximity) Disturbance regulation/storm protection Waste treatment Pollination Biological control Habitat/Refugia 3. Directional flow-related: flow from point of production to point of use Water regulation/flood protection Water supply Sediment regulation/erosion control Nutrient regulation 4. In situ (point of use) Soil formation Food production/non-timber forest products Raw materials 5. User movement related: flow of people to unique Genetic resources Recreation potential Cultural/aesthetic 10
15
Understanding variations in the ecosystems species and the influence of external factors such as climate change and socioeconomics at different temporal scales require the use of geospatial data. For example, when considering time from a (spatial) data perspective leads to a basic tradeoffs between temporal and spatial resolution. It means if we are looking for high spatial resolution satellite data (e.g. Landsat 30m spatial resolution with repeat cycle 16 days) then we need to compromise with the temporal resolution (MODIS 500m spatial resolution and 1 day repeat
Page 16 of 34
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
cycle).This is particularly relevant for in-situ ecosystems services (soil formation, food production). However, directional flow related (from point of production to point of use, e.g. Water regulation/flood protection, Water supply, Sediment regulation/erosion control and Nutrient regulation) has a time dimension of seconds to days, as in the case of hydrology models, to multiple decades in the case of land use/cover change (UK NEA). Ecosystem services which do not depend on proximity of people to a spatial denomination (e.g. Climate regulation) can span centuries. Multi-scale scenario comes in our theoretical/practical framework with intention to critically evaluate the stronger aspects of linkage between single scale scenarios. Although multi-scale environmental scenarios area more valuable than single scale in terms of broadening the aspects and perspectives; immature knowledge about different attributes about the system may also end up with disastrous or ineffective outcome. Again in multi-scale scenarios, we do have two types of coupling one is loose and other one is tightly coupled. It is being found that loosely coupled scenarios where we have a common focal point but each scenarios developed independently has better output in terms of maintaining the consistency of the credibility and importance of different scenarios to final end users. Henceforth, loosely linked scenarios tends to allow for the convergence of different issues for better output of viewpoint on one hand but also it serve to keep intact the robustness of findings of overall exercise (Biggs et al., 2006). The linkages between scales are useful to study about multiscale scenarios and it can provide information about the number of scales at which scenarios are developed and the connection between the scales. According to Biggs et al., 2006 the scenarios can be divided into three types (a) single-scale scenario exercise (b) loosely linked scenarios and (c) cross-scale scenarios that are tightly coupled across two or more scales. In the case of loosely linked multiscale scenarios, links may be established up front or after scenario development and have varying degree of flexibility. In the case of tightly couples cross-scale scenarios, links are usually established up front and reinforced by an iterative process of downscaling and upscaling, with greater emphasis on downscaling because researcher and policy makers has more interest on how downscaling institutional and economic drivers affect regions.
30
6.4.2. Standardization of scenarios
35
40
Scenarios that are related to biodiversity and ecosystem services are produced from a number of international, regional and national assessments. Each of them was developed for a specific set of objectives, such as greenhouse gas emission, sustainable development of the society. Most of them employed different methodologies in developing the scenarios, even between different iterations of the assessment. For example, various sets of scenarios for different global environmental assessments, including the Global Scenario Group (GSG)’s work on Great Transitions (Raskinetet al. 2002, 2005), the IPCC Special Report on Emission Scenarios (SRES) (Nakicenovicet al. 2000) and UNEP’s Third Global Environmental Outlook (GEO3) (UNEP 2002) and the World Water Vision work (Cosgrove and Rijsberman 2000, van Vuurenet al. 2012). Moreover, the IPCC develops a new set of socio-economic scenarios (Shared Socio-economic Pathway, SSP) for their assessments.
45
Page 17 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
5
10
15
20
25
Figure 6.4: A suggested mapping of the different scenarios used by the IPCC (based on O’Neill et al. 2013).
Available literature on standardizing and harmonizing scenarios for global environmental assessments suggested three main steps: (1) identify and discuss the application of the scenarios and their main characteristics (2) compare the key assumptions and storylines behind the scenarios, and (3) compare the trends observed in the main scenario methodology in relation to policy making (van Vuuren et al. 2012). For example, application and characteristics of scenario can be identified and categorized by “scenario family”, i.e., a set of scenarios in the literature that seem to share a very similar scenario storyline, logic and underlying assumptions. Some of the key elements in which these scenarios differ include risk-perception and resulting policy actions to environmental change, spatial scale of drivers and systems, trends relative to the past, and degree of cooperation in the society (van Vuuren et al. 2012) (see Table 6.4 for a case study that applies to five sets of scenarios for global environmental assessments).
Page 18 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Table 6.4: Key assumptions and approximate mappings of scenario types (from van Vuuren et al. 2012).
5
10
15
20
6.4.3 Benchmarking of models Benchmarking is the process of systematically comparing sets of model predictions against measured data in order to evaluate model performance, and represents a structured way of using empirical data to inform biodiversity and ecosystem services models (McCarthy et al. 2012). It is common practice in fields other than ecology: for example, Global Circulation Models included in IPCC need to be able to hindcast historic temperature trends, derived from multiple sources. In biodiversity and ecosystem services science, benchmarking has been used less frequently. Luo et al. (2012) proposed a framework for benchmarking land models, which are models that predict future states of ecosystem and climate. Land models account for biophysical processes (exchanges of water and energy) and biogeochemical cycles of carbon, nitrogen, and trace gases (e.g. Wang et al., 2010; Zaehle et al., 2010), as well as vegetation dynamics and disturbances (Sitch et al., 2003; Thonicke et al., 2010). The framework proposed by Luo and colleagues as part of the International Land Model Benchmarking (ILAMB) project includes 1) targeted aspects of model performance to be evaluated, (2) a set of benchmarks as defined references to test model performance, (3) metrics to measure and compare performance skills among models, and (4) model improvement. Williams et al. (2009) developed methods to benchmark land surface models (which simulate terrestrial biosphere exchanges of matter and energy) with primary data provided by FLUXNET, which is an international network of sites that measure the land surface exchanges of carbon, water and energy.
6.5 Uncertainty in linking and harmonizing models and scenarios 25
6.5.1 Cascade of uncertainty from models linking biodiversity and ecosystem services 30
Uncertainty in model projections for any time horizon and spatial scale arises from three sources: (1) internal variability, (2) model uncertainty and (3) scenario uncertainty (Figure 6.5). Internal variability is caused by natural physical, ecological and social processes that are intrinsic to systems. It arises even in the absence of any human drivers. Model uncertainty is comprised of two sub-categories: Page 19 of 34
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
parameter and structural uncertainty (Tebaldi and Knutti, 2007). Parameter uncertainty relates to the specific parameter values used in the formulae that determine the behavior of the models for biodiversity and ecosystem services (Tebaldi and Knutti, 2007; Knuttiet al., 2010). Structural uncertainty relates to different ways in which ecological, social and economics interactions can be mathematically represented. Scenario uncertainty relates to the many possible futures that may happen due to differences in the natural and/or anthropogenic forcing that drive the model simulation. Uncertainties from one model component can carry into another components, and in some cases, magnified when they are linked.
10 Figure 6.5: Cascade of uncertainties of linking biodiversity and ecosystem services models.
15
The linking of models across spatial and temporal scales and domains may potentially enlarge the envelope of uncertainty from the different types of uncertainties.To assess confidence in model projections, we suggest three possible tiers of evaluation: (1) consistency with mean spatial patterns and temporal patterns across the scale of interest; (2) consistency with past observed responses to variability; and (3) consistency with observed long-term trends attributable to specific drivers. Representative model metrics are needed to evaluate different aspects of model projections.
20
25
30
The first tier of evaluation is suggested to be applied to all biodiversity and ecosystem services projections so that implausible projections can be identified. Model simulations that do not reproduce the broad mean spatial and temporal patterns of changes suggest that the models may not sufficiently represent the biophysical and social-economic components that are important for the aspects of biodiversity or ecosystem services and scale of interest. Also, data are generally available for such broad-scale evaluation (Table 3). This is consistent with the description of Overland et al.(2011) of a coarse "culling" of simulations if they are in very stark disagreement observations. Limited availability of data sets may make it impossible to evaluate models for at all three tiers. In particular, data is challenged by issues of consistency between timeframe and spatial scales and confounding effects of multiple human pressures such as climate and fishing (e.g., ref) These limitations of available data should not prevent application of the models and deem all model
Page 20 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
projection unreliable, as projections also gain credibility through their reliance on robust ecological and physiological principles [ref]. It should, however, temper interpretation of results.
6.5.2 Scaling errors and uncertainty 5
Downscale and upscale predictions can differ from observed values for different reasons, including nonlinearity in scaling properties (or in the functional relationships between processes and environmental variables; Jarvis 1995), species spatial aggregation or patchiness and cross-scale interactions. As a consequence, the higher the order of magnitude of scaling predictions, the higher is the risks of propagating errors (Jarvis 1995).
10
15
20
25
The first potential source of error and uncertainty in extrapolating pattern across scales is scale variance. It is reasonable to think that the larger the extent (with a constant grain) or the grain (with a constant extent), the larger will be the heterogeneity observed, and the higher will be the probability of having discontinuity in the properties of the observed object with the change in scale. For example, temperature and precipitation in mountain areas are highly affected by the heterogeneity of the relief, which is unevenly distributed spatially. As a consequence of this local aggregated heterogeneity, there are sharp differences in mountain areas between original (field- based) temperature and precipitation values and downscaled data sets provided by global climatic models (Karynand Williams 2010). Those abrupt changes or transition zones in system properties delimit “domains of scale”. Inside each domain, scale variance is linear (i.e., the functional relationships among studied components are constant), while in between domains, properties change abruptly. Scaling inside a same domain of scale, or for objects or systems that present scale invariance (such as fractal systems), is usually simple and can be done with linear regression functions. For example, it is well known that the size and frequency of disturbances are inversely related (e.g., large-scale disturbances are less frequent than small-scale disturbances), and this can be easily represented by a power-law function. However, scaling between two or more domains of scale, where non-linear relationships occur, may be much more challenging to apprehend with simple mathematical models, and thus can lead to significant errors propagation. For some authors, extrapolations across domains of scale are not recommended (Wiens 1989).
30
35
40
45
A second significant source of scaling errors is related to species’distributional features, such as species spatial aggregation, which may bias estimates if the scaling process is nonlinear (Stoyet al. 2009). Errors can be particularly more severe when projecting the location of species, and not only a global assessment of the distributional area. Indeed, downscaling usually tend to overestimate species distributions (Sardà-Palomeraet al. 2012). However, precise information on species distribution at a local level is crucial for local decision making (Franklin et al. 2013), such as for identifying biodiversity hotspots (Sardà-Palomeraet al. 2012). In those cases, a more complex framework, combining niche and spatial models within a spatially explicit approach is necessary to reduce errors when modeling species locations (Azaeleet al. 2012). Different techniques have been proposed to deal with species' spatial aggregation, such as the scale transition theory (Melbourne andChesson 2006) and the shot noise Cox processes (SNCP), which allow a better prediction of population estimates at fine scales starting from coarser ones (Azaeleet al. 2012). Another additional source of error is related to cross-scale interactions, when processes interact at different spatial or temporal scales. Errors and uncertainty are thus inherent to any scaling procedure. Carbon flux from woody debris, for example, is simultaneously affected by climate,
Page 21 of 34
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
site environment and species-based variation in wood quality (Weedonet al. 2009), i.e. by processes that occur at different spatial and temporal scales. As a consequence, any upscaling or downscaling framework will need to consider the interactions among those processes to properly model carbon dynamics. In this sense, it could be useful to consider in global carbon-climate models species traits that regulate wood and decomposition characteristics at a more local (plot) scale (Weedonet al. 2009).To reduce this problem, it is first crucial to identify “domains of scale” and theirs respective scaling thresholds, which should reflect fundamental shifts in underlying processes that regulate the studied system (Wu and Li 2006), and to deal with caution with any extrapolation across domains. It is also necessary to identify cross-scale interactions and to develop multiple- scaled models that allow integrating those interactions across scales. Ground observations and global models on coarse spatial resolutions are very important sources for data of simulating changes of biodiversity and ecosystem services (Box 2). However, too sparsely distributed ground observations are often unable to satisfy the data requirements of most ecosystem change studies; one major problem is how to estimate values for locations where primary data is not available by interpolation. Many global models are difficult to be used at regional and local levels because their spatial resolutions are too coarse; to improve the poor performance of the global models at local and regional scales, we have to develop downscaling approaches. The inappropriate interpolation and downscaling methods might cause too big uncertainty to be applied to the issues that have accuracy requirements. The ground observation data from local areas and correct uses of these data are critical important for the interpolation and downscaling results with a higher accuracy.
6.6. Conclusions 25
30
35
40
Because of the complexity of systems related to assessing current status and trends and developing future scenarios for biodiversity and ecosystem services, it is often necessaryto link models or scenarios representing different components of the relevant social-ecological systems. Also, there are multiple parallel efforts in modelling and assessing biodiversity and ecosystem services for different domains and at different scales. Models and scenarios that integrate feedback and tradeoffs across temporal and spatial scales and among dynamic societal economic and natural systems can address complex challenges and guide decision making. However, the question of whether biodiversity and ecosystem service models should be directly linked depends on the research objectives and societal demands. Nevertheless we think that not all useful linkages have been utilized and more direct linkages have great potential To facilitate the development of methods for linking and harmonizing scenarios and models, we need to build communities of multi-disciplinary researchers and practitioners to support such research and development. The rapidly growing communities of model intercomparison projects facilitate the development of harmonization of models and cultivate a community to make advancement in the long-term. However, existing intercomparison projects are sectoral focus e.g., for carbon cycling, agriculture or fisheries. A major need in models and scenarios for biodiversity and ecosystem services is strengthening the linkages between the biophysical and human domains. There are increasing efforts in this area such as IAMs and ecosystem services assessments. However, more extensive development and application of these approaches should be encouraged to accelerate the state-ofthe-art in linking models and scenarios across social and natural domains.
45
Page 22 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Box 6.1 Using scenarios of global change to project species distributions and biodiversity trends into the future 5
10
15
20
25
30
Visconti et al. (2015) projected trends of ca. 400 species of terrestrial large mammals in two widely used indicators of population abundance (the Living Planet Index; LPI) and extinction risk (the Red List Index; RLI) under different climate and land-use change scenarios. These two complementary indicators have been adopted by the CBD to measure progress towards global biodiversity targets. The impact of climate change on species' geographic range was quantified by fitting bioclimatic envelope models to the present-day species’ distributions, and projecting these under future climate associated with two scenarios of socio-economic development until 2050. The two scenarios, developed for the Rio+20 conference held in Rio in 2002 represent business-as-usual production and consumption patterns and rates, or reduced consumption (PBL 2012). For each socio-economic scenario, three species responses to climate change were tested: 1) species cannot disperse into new climatically suitable areas; 2) species can expand their distributions each generation by a median dispersal distance estimated using statistical models; or 3). Species adapt locally (their geographic ranges are not affected by climate change). Projected species ranges were further assessed for compatibility with species’ fine-scale ecological requirements with habitat suitability models (Rondinini et al. 2011, Visconti et al. 2011) based on species’ land cover and altitudinal preferences and sensitivity to human disturbance. These models were applied to projected land-use maps from the IMAGE model (Bouwman et al. 2006) under each scenario, to quantify for each species the extent of suitable habitat (ESH). The distribution projected under each climate change scenario was taken as the extent of occurrence (EOO). The ESH was treated as the maximum potential value of area of occupancy (AOO). The number of mature individuals of a species was estimated by multiplying the AOO by population density from observed and modelled data. These parameters were applied to Red List criteria to evaluate each species’ Red List category for each year under each scenario, from which the overall RLI was calculated following Butchart et al. (2007) (Fig. B6.1). The uncertainty around the proportion of mature individuals and proportion of suitable habitat occupied (AOO/ESH) was incorporated into RLI projections by randomly sampling these parameters from a distribution with intervals gathered from the literature and performing a Monte Carlo simulation. Estimates of mature individuals for each species and each year were used to generate the LPI for each scenario following Collen et al. (2009). The methodology was validated through hind-casting species distributions and biodiversity indicators from 1970.
Page 23 of 34
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
Figure B6.1: Spatial patterns of trends in Red List Index. (A,B) Bivariate plot showing spatial pattern in species richness and trends in the Red List Index (d-RLI) between 2010 and 2050 under the business-as-usual scenario, with land use and climate change and assuming maximum dispersal (A) and no dispersal (B). (C-D) Relative improvements in d-RLI for the reduced impact scenario relative to business-as-usual for year 2050 under maximum dispersal (C) and no dispersal (D). Areas in white contain fewer than 5 species per grid cell modeled in 2010.
Testing these on terrestrial carnivore and ungulate species, Visconti and colleagues found that both indicators decline steadily, and by 2050, under a business-as-usual scenario, the LPI declines by 18-35% while extinction risk increases for 8-23% of the species, depending on assumptions about species responses to climate change. Business-as-usual will therefore fail CBD target 12 of improving the conservation status of known threatened species. An alternative sustainable development scenario reduces both extinction risk and population losses compared with Business-asusual and could lead to population increases.
Box 6.2: Integrated assessment model - The IMAGE 3.0 - GLOBIO Framework 20
25
30
The IMAGE integrated assessment modelling framework has been developed to understand how global, long-term environmental change and sustainability problems develop over time, driven by human activities, such as economic development and population growth (Fig. B6.2). Similar to other integrated assessment models, IMAGE can be used to identify problems of global environmental change, and to advise on possible response strategies. Earlier versions of the IMAGE model have been used to support various international assessments, including IPCC assessments, UNEP’s Global Environment Outlooks, OECD’s Environmental Outlooks and the Millennium Ecosystem Assessment. Moreover, the model has been extensively used in the scientific literature. IMAGE assesses the impacts of socio-economic drivers on the physical environment, such as climate change, land-use change and pollution, and these provide inputs to the GLOBIO model to help evaluate impacts on biodiversity. GLOBIO was developed to provide information to policy makers at the international level on current biodiversity status and future trends (Alkemadeet al.
Page 24 of 34
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
2009). The model delivers quantified results on the impact of environmental drivers and potential policy options on biodiversity. Potential trends in biodiversity are addressed in future scenarios, including the expected outcome in the absence of additional policies to prevent biodiversity loss. GLOBIO delivers output in terms of MSA (species abundance relative to the natural state of original species), land cover and land use (high resolution land use and land use intensity based on GLC2000 and IMAGE), SRI (species richness index) and Wilderness area.
10
15
20
25 Figure B6.2: Framework of the IMAGE3.0 – GLOBIO Integrated Assessment Model.
30
35
40
45
Box 6.3 A case study of interpolating data from meteorological stations Lloyd (2005) compared the performance of different interpolation methods, including a moving window regression (MWR), inverse distance weighting (IDW) and Kriging, which demonstrated that methods using elevation as secondary data performed better than others because of the relationships between climate factors, such as temperature, precipitation and evaporation, to elevation. Thiessen polygons (TP), IDW, Spline and ordinary Kriging (OK) were used to interpolate thirteen widely scattered rainfall stations and their daily time series into gridded rainfall surfaces over the 1950–1992 period in a West African catchment; assessment of the interpolation methods using reference point data indicated that interpolations using the IDW and OK were more efficient than TP and, to a lesser extent, Spline (Ruellandet al. 2008). Different interpolation models in a GIS environment were used to generate precipitation surfaces for the northwestern Himalayan Mountains and upper Indus plains of Pakistan; this precipitation simulation using a regional climate model (PRECIS) showed that OK, using elevation as secondary data, provided the best results especially for the monsoon months (Ashiqet al. 2010). Three interpolation approaches, IDW, Spline and Co-kriging, were used to interpolate monthly mean temperature, seasonal mean temperature, and annual mean temperature in the eastern part of India; it was found that Spline was preferred to Kriging and IDW because it was faster and easier to use (Samantaet al. 2012).
Page 25 of 34
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
A combination of interpolation methods applied through statistical transfer functions (STFs) is an efficient approach to improve the estimation error of climatic variables for locations where primary data is not available. It has been determined that the statistical relationship between mean annual temperature (MAT) and its environmental determinants is the same no matter where the measurement takes place (Yue, 2011). But a simple ‘global’ model cannot explain the relationship between mean annual precipitation (MAP) and its environmental variables. The MAP relationship changes across space as a function of topographic structure across the landscape. In other words, MAT exhibits spatial stationarity and its statistical transfer function can be expressed by an Ordinary Least Squares regression (OLS), while MAP exhibits spatial non-stationarity and therefore its statistical transfer functions have to be formulated through Geographically Weighted Regression (GWR). The introduction of spatial non-stationarity analyses into the interpolation of meteorological stations has greatly improved the interpolated climate surfaces. For instance, IDW was applied to interpolation of MAT and MAP in China, taking a digital elevation model (DEM) as secondary data; mean absolute errors of the interpolated MAT and MAP were respectively 1.51 °C and 102.23 mm (Shang et al., 2001; Pan et al., 2004). The mean relative error of the interpolated MAT is decreased by 6% because a statistical transfer function was combined into IDW and the one of MAP decreased by 3% due to the combination of Geographically Weighted Regression for spatial non-stationarity issues with IDW; in addition, when a method for high accuracy surface modelling (HASM) is used, which has much better performance comparing with the classical methods such as IDW, Kriging and Spline (Haber, 2012; Jorgensen, 2011), the accuracies of the interpolated MAT and MAP have been increased by 1% and 3% respectively (Yue et al., 2013).
25
Page 26 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
References
5
10
15
20
25
30
35
40
45
50
Ainsworth, C.H., Samhouri, J.F., Busch, D.S., Cheung, W.W.L., Dunne, J., Okey, T.A., 2011. Potential impacts of climate change on Northeast Pacific marine foodwebs and fisheries. ICES Journal of Marine Science, 68, 1217-1229. Alcamo J, Doll P, Henrichs T, Kaspar F, Lehner B, Rosch T, Siebert S (2003) Development and testing of the WaterGAP 2 global model of water use and availability. Hydrological Sciences Journal 48:317-337. Alkemade R, van Oorschot M, Miles L, Nellemann C, Bakkenes M, ten Brink B (2009) GLOBIO3: A Framework to Investigate Options for Reducing Global Terrestrial Biodiversity Loss. Ecosystems 12:374-390. Algar, A.C., Kharouba, H.M., Young, E.R., Kerr, J.T., 2009. Predicting the future of species diversity: macroecological theory, climate change, and direct tests of alternative forecasting methods. Ecography, 32, 2233. Altman, I., Boumans, R., Roman, J., Gopal, S. & Kaufman, L. 2014.An Ecosystem Accounting Framework for Marine Ecosystem-Based Management. In: Fogarty, M. J. & McCarthy, J. J. (eds.) Marine Ecosystem-based Management. Harvard University Press. Anderson M. C., William P. Kustas, and John M. Norman, 2003. Upscaling and Downscaling—A Regional View of the Soil–Plant–Atmosphere ContinuumAgron. J. 95:1408–1423 Ashiq MW, Zhao CY, Ni, J, Akhtar M (2010) GIS-based high-resolution spatial interpolation of precipitation in mountain–plain areas of Upper Pakistan for regional climate change impact studies. Theoretical and Applied Climatology 99: 239-253. Azaele, S., Cornell, S.J., Kunin, W.E., 2012. Downscaling species occupancy from coarse spatial scales. Ecological Applications, 22, 1004-1014. Barwell, L.J., Azaele, S., Kunin, W.E., Isaac, N.J.B., Richardson, D., 2014. Can coarse-grain patterns in insect atlas data predict local occupancy? Diversity and Distributions, 20, 895-907. Batker, D., Kocian, M., McFadden, J. & Schmidt, R. 2010. Valuing The Puget Sound Basin: Revealing Our Best Investments. Biggs, R., Raudsepp-Hearne, C., Atkinson-Palombo, C., Bohensky, E., Boyd, E., Cundill, G., Fox, H., Ingram, S., Kok, K., Spehar, S., 2007. Linking futures across scales: a dialog on multiscale scenarios. Ecology and Society, 12, 17. Blanchard, J., Jennings, S., Holmes, R., Harle, J., Merino, G., Allen, J., Holt, J., Dulvy, N.K., Barange, M., 2012. Potential consequences of climate change for primary production and fish production in large marine ecosystems. Philosophical Transactions of the Royal Society B: Biological Sciences, 367(1605), 2979-2989. Böttcher, H., Verkerk, P.J., Gusti, M., HavlÍk, P., Grassi, G., 2012. Projection of the future EU forest CO2sink as affected by recent bioenergy policies using two advanced forest management models. GCB Bioenergy, 4, 773-783. Bombi, P., Salvi, D., Bologna, M.A., Pettorelli, N., 2012. Cross-scale predictions allow the identification of local conservation priorities from atlas data. Animal Conservation, 15, 378-387. Bondeau, A., Smith, P.C., Zaehle, S., Schaphoff, S., Lucht, W., Cramer, W., Gerten, D., Lotze-Campen, H., MÜLler, C., Reichstein, M., Smith, B., 2007. Modelling the role of agriculture for the 20th century global terrestrial carbon balance. Glob Chang Biol, 13, 679-706. Boumans, R. &Costanza, R. 2007. The multiscale integrated Earth systems model (MIMES): the dynamics, modelling and valuation of ecosystem services. In: Leslie, R. N. (ed.) GWS Issues in Global Water System Research, Global Assessments: Bridging Scales and Linking to Policy. Germany: The Global Water System Project. Boumans, R., Costanza, R., Farley, J., Wilson, M. A., Portela, R., Rotmans, J., Villa, F. & Grasso, M. 2002.Modeling the dynamics of the integrated earth system and the value of global ecosystem services using the GUMBO model. Ecological Economics, 41, 529-560. Bouwman A., Kram T., Klein Goldewijk K. (2006) Integrated modelling of global environmental change: an overview of Image 2.4. Netherlands Environmental Assessment Agency Bilthoven. Boumans, R. & McNally, D. 2012. MIMES: multiscale, integrated models for assessing ecosystem services. Vermont, USA: Afordable Futures LLC. Available: http://www.afordablefutures.com/about-us [Accessed 13th September 2012 2012].
Page 27 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Chen, B., Ge, Q., Fu, D., Yu, G., Sun, X., Wang, S., Wang, H., 2010. A data-model fusion approach for upscaling gross ecosystem productivity to the landscape scale based on remote sensing and flux footprint modelling. Biogeosciences, 7, 2943-2958. Braat, L. C. & de Groot, R. 2012. The ecosystem services agenda:bridging the worlds of natural science and economics, conservation and development, and public and private policy. Ecosystem Services, 1, 415. Brashares JS et al. (2004) Bushmeat Hunting, Wildlife Declines, and Fish Supply in West Africa. Science 306: 1180-1183. Bryan BA, Crossman ND (2013) Impact of multiple interacting financial incentives on land use change and the supply of ecosystem services. Ecosystem Services 4:60-72. doi:http://dx.doi.org/10.1016/j.ecoser.2013.03.004. Bryan BA, Nolan M, Harwood TD, Connor JD, Navarro-Garcia J, King D, Summers DM, Newth D, Cai Y, Grigg N, Harman I, Crossman ND, Grundy MJ, Finnigan JJ, Ferrier S, Williams KJ, Wilson KA, Law EA, HatfieldDodds S (2014) Supply of carbon sequestration and biodiversity services from Australia's agricultural land under global change. Global Environmental Change-Human and Policy Dimensions 28:166-181. Butchart S.H.M., Akçakaya H.R., Chanson J. et al. (2007) Improvements to the red list index. PLoS ONE 2: e140 Calabrese JM, Certain G, Kraan C, Dormann CF (2014) Stacking species distribution models and adjusting bias by linking them to macroecological models. Global Ecology and Biogeography 23:99-112. Carpenter S. R., Bennett E. M., Peterson G. D., 2006. Scenarios for Ecosystem Services: An Overview. Ecology and Society 11 (1): 29. Cash, D.W., Adger, W.N., Berkes, F., Garden, P., Lebel, L., Olsson, P., Pritchard, L., Young, O., 2006. Scale and cross-scale dynamics: governance and information in a multilevel world. Ecology and Society, 11, 8. Charles SP, Bates BC, Whetton PH, Hughes JP (1999) Validation of downscaling models for changed climate conditions: case study of southwestern Australia. Climate Research 12: 1-14. Chen, B., Ge, Q., Fu, D., Yu, G., Sun, X., Wang, S., Wang, H., 2010. A data-model fusion approach for upscaling gross ecosystem productivity to the landscape scale based on remote sensing and flux footprint modelling. Biogeosciences, 7, 2943-2958. Cheung, W.W.L., Lam, V.W.Y., Sarmiento, J.L., Kearney, K., Watson, R.E.G., Zeller, D., Pauly, D., 2010. Large-scale redistribution of maximum fisheries catch potential in the global ocean under climate change. Global Change Biology, 16, 24-35. Cheung, W.W.L., Meeuwig, J.J., Feng, M., Harvey, E., Lam, V., Langolis, T., Slawinski, D., Sun, C., Pauly, D., 2012. Climate change induced tropicalization of marine communities in Western Australia. Marine and Freshwater Research, 63, 415-427. Christensen, V., Walters, C.J., 2004.Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling, 172, 109-139. Christensen, V., Coll, M., Buszowski, J., Cheung, W.W.L., Frölicher, T., Steenbeek, J., Stock, C., Watson, R.A., and Walters, C.J. 2015. The global oceans is an ecosystem: stimulating marine life and fisheries. Global Ecology and Biogeography. Collen B., Loh J., Whitmee S., McRAE L., Amin R., Baillie J. (2009) Monitoring change in vertebrate abundance: the Living Planet Index. Conservation Biology 23: 317-327. Cosgrove, W.J., Rijsberman, F.R., 2014. World water vision: making water everybody's business: Routledge. Costanza, R. 2008. Ecosystem services: Multiple classification systems are needed. Biological Conservation, 141, 350-352. Available: DOI DOI 10.1016/j.biocon.2007.12.020. Costanza, R., Cumberland, J., Daly, H., Goodland, R. &Norgaard, R. 1997a.An Introduction to Ecological Economics. Florida, USA, CRC Press. Costanza, R., dArge, R., deGroot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., Oneill, R. V., Paruelo, J., Raskin, R. G., Sutton, P. & van den Belt, M. 1997. The value of the world's ecosystem services and natural capital. Nature, 387, 253-260. Available: DOI Page 28 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
10.1038/387253a0. Costanza, R., de Groot, R., Sutton, P. C., van der Ploeg, S., Anderson, S., Kubiszewski, I., Farber, S. & Turner, K. 2014.Changes in the global value of ecosystem services. Global Environmental Change Part A: Humand and Policy Dimensions, 26. Costanza, R., Wilson, M., Troy, A., Voinov, A., Liu, S. &D’Agostino, J. 2006.The Value of New Jersey’s Ecosystem Services and Natural Capital. Burlington, Vermont, USA: Gund Institute for Ecological Economics. Costanza, R., dArge, R., deGroot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., Oneill, R. V., Paruelo, J., Raskin, R. G., Sutton, P. & van den Belt, M. 1997b. The value of the world's ecosystem services and natural capital. Nature, 387, 253-260. Available: DOI 10.1038/387253a0.Crossman, N.D., Burkhard, B., Nedkov, S., Willemen, L., Petz, K., Palomo, I., Drakou, E.G., Martín-Lopez, B., McPhearson, T., Boyanova, K., Alkemade, R., Egoh, B., Dunbar, M.B. andMaes, J. (2013) A blueprint for mapping and modelling ecosystem services. Ecosystem Services, 4, 4-14. Daily, G. C. 1997. Nature's services : societal dependence on natural ecosystems. Washington, DC, Island Press. De Groot, R., Brander, L., van der Ploeg, S., Costanza, R., Bernard, F., Braat, L., Christy, M., Crossman, N., Ghermandi, A., Hein, L., Hussain, S., Kumar, P., McVittie, A., Portella, R., Roderiguez, L. C., ten Brink, P. & van Beukeringen, P. 2012. Global estimates of the value of ecosystems and their services in monetary units. Ecosystem Services, 1, 11. Dibike Y, Gachon P, St-Hilaire A, Ouarda TBMJ, Nguyen VTV (2008) Uncertainty analysis of statistically downscaled temperature and precipitation regimes in Northern Canada. Theoretical and applied Climatology 91(1): 149-170. Dietze M. C., et al. 2011. Characterizing the performance of ecosystem models across time scales: A spectral analysis of the North American Carbon Program site-level synthesis. JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 116. Doherty RM, Sitch S, Smith B, Lewis SL, Thornton PK (2009) Implications of future climate and atmospheric CO2 content for regional biogeochemistry, biogeography and ecosystem services across East Africa. Global Change Biology 16:617-640. Fernandes, R.F., Vicente, J.R., Georges, D., Alves, P., Thuiller, W., Honrado, J.P., 2014. A novel downscaling approach to predict plant invasions and improve local conservation actions. Biological Invasions, 16, 2577-2590. Ferrier S, Guisan A (2006) Spatial modelling of biodiversity at the community level. Journal of Applied Ecology 43:393-404. Ficetola, G.F., Rondinini, C., Bonardi, A., Baisero, D., Padoa-Schioppa, E. (2014) Habitat availability for amphibians and extinction threat: a global analysis. Diversity and Distributions, doi: 10.1111/ddi.12296. Flint L. E., and Flint A. L., 2012.Downscaling future climate scenarios to fine scales for hydrologic and ecological modeling and analysis. Ecological Processes 1:2. Franklin, J., Davis, F.W., Ikegami, M., Syphard, A.D., Flint, L.E., Flint, A.L., Hannah, L., 2013. Modeling plant species distributions under future climates: how fine scale do climate projections need to be? Glob Chang Biol, 19, 473-483. Fu, D., Chen, B., Zhang, H., Wang, J., Black, T.A., Amiro, B.D., Bohrer, G., Bolstad, P., Coulter, R., Rahman, A.F., Dunn, A., McCaughey, J.H., Meyers, T., Verma, S., 2014. Estimating landscape net ecosystem exchange at high spatial–temporal resolution based on Landsat data, an improved upscaling model framework, and eddy covariance flux measurements. Remote Sensing of Environment, 141, 90-104. Fulton, E.A., Link, J.S., Kaplan, I.C., Savina‐Rolland, M., Johnson, P., Ainsworth, C., Horne, P., Gorton, R., Gamble, R.J., Smith, A.D., 2011. Lessons in modelling and management of marine ecosystems: the Atlantis experience. Fish and Fisheries, 12, 171-188. Funtowicz, S.O., Ravetz, J.R., 1993. Science for the post-normal age. Futures 25, 739–755. Gachon P, Dibike Y (2007) Temperature change signals in northern Canada: convergence of statistical downscaling results using two driving GCMs. International Journal of Climatology 27: 1623-1641. Goetze D., Karlowski U., Porembski S., Tockner K., Watve A., Riede K., 2014. In book: Biodiversity: Structure and Function, from Encyclopedia of Life Support Systems (EOLSS), Edition: online [http://www.eolss.net], Chapter: Spatial and temporal dimensions of biodiversity dynamics, Editors: UNESCO, Page 29 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Eolss Publishers, Oxford, UK. Gonzalez P, Neilson RP, Lenihan JM, Drapek RJ (2010) Global patterns in the vulnerability of ecosystems to vegetation shifts due to climate change. Global Ecology and Biogeography 19:755-768. Grotch SL, MacCracken MC (1991) The use of general circulation models to predict regional climate change. Journal of Climate 4: 286-303. Haber W (2012) Book review. Journal of Geographical Science 22 (3): 574. Hamilton, L. C. 2011. Education, politics and opinions about climate change evidence for interaction effects. Climatic Change, 104, 231-242. Hanewinkel M, Cullmann DA, Mart-Jan Schelhaas, Nabuurs G-J, Zimmermann NE (2013) Climate change may cause severe loss in the economic value of European forest land. Nature Climate Change 3:203-207. Harding KJ, Snyder PK,Liess S (2013) Use of dynamical downscaling to improve the simulation of CentralU.S. warm season precipitation in CMIP5 models. Journal of Geophysical Research: Atmospheres 118: 12522–12536. Harfoot, M., Tittensor, D.P., Newbold, T., McInerny, G., Smith, M.J., Scharlemann, J.P.W., 2014. Integrated assessment models for ecologists: the present and the future. Global Ecology and Biogeography, 23, 124-143 Hay, G., Niernann, K., Goodenough, D., 1997. Spatial thresholds, image-objects, and upscaling: a multiscale evaluation. Remote Sensing of Environment, 62, 1-19. HewitsonBC, CraneRG (1996) Climate downscaling: techniques andapplication. Climate Research 7: 8595. Hickler T, Vohland K, Feehan J, Miller PA, Smith B, Costa L, Giesecke T, Fronzek S, Carter TR, Cramer W, Kühn I, Sykes MT (2012) Projecting the future distribution of European potential natural vegetation zones with a generalized, tree species-based dynamic vegetation model. Global Ecology and Biogeography 21:50-63. Hilker, T., Coops, N.C., Hall, F.G., Black, T.A., Chen, B., Krishnan, P., Wulder, M.A., Sellers, P.J., Middleton, E.M., Huemmrich, K.F., 2008. A modeling approach for upscaling gross ecosystem production to the landscape scale using remote sensing data. Journal of Geophysical Research, 113. Holling, C.S., 1978. Adaptive environmental assessment and management. Adaptive environmental assessment and management. Hurtt GC et al. (2011) Harmonization of land-use scenarios for the period1500–2100: 600 years of global gridded annual land-usetransitions, wood harvest, and resulting secondary lands. Climatic Change 109:117–161. Iglecia MN, Collazo JA, McKerrow AJ (2012) Use of Occupancy Models to Evaluate Expert Knowledgebased Species-Habitat Relationships. Avian Conservation and Ecology 7(2): 5. doi:10.5751/ace-00551-070205 Jarvis, P.G. Scaling processes and problems (1995) Plant, Cell and Environment, 18 (10), pp. 1079-1089. Jorgensen SE (2011) Book review. Ecological Modelling 222: 3300. Karyn, T., Williams, J.W. Globally downscaled climate projections for assessing the conservation impacts of climate change (2010) Ecological Applications, 20 (2), pp. 554-565. Keating BA, Carberry PS, Hammer GL, Probert ME, Robertson MJ, Holzworth D, Huth NI, Hargreaves JNG, Meinke H, Hochman Z, McLean G, Verburg K, Snow V, Dimes JP, Silburn M, Wang E, Brown S, Bristow KL, Asseng S, Chapman S, McCown RL, Freebairn DM, Smith CJ (2003) An overview of APSIM, a model designed for farming systems simulation. European Journal of Agronomy 18:267-288. Keil, P., Belmaker, J., Wilson, A.M., Unitt, P., Jetz, W. (2013) Downscaling of species distribution models: A hierarchical approach Methods in Ecology and Evolution, 4 (1), pp. 82-94. Keil, P., Jetz, W., 2014. Downscaling the environmental associations and spatial patterns of species richness. Ecological Applications, 24, 823-831. Khalili M, Van Nguyen VH, Gachon P (2013) A statistical approach to multi-site multivariate downscaling of daily extreme temperature series. International Journal of Climatology 33(1): 15-32. Kramer-Schadt S, Revilla E, Wiegand T (2005) Lynx reintroductions in fragmented landscapes of Germany: Projects with a future or misunderstood wildlife conservation? Biological Conservation 125:169-182. Landsberg JJ, Waring RH (1997) A generalised model of forest productivity using simplified concepts of radiation-use efficiency, carbon balance and partitioning. Forest Ecology and Management 95:209-228. Page 30 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Laprise R (2008) Regional climate modelling. Journal of Computational Physics 227 (7): 3641-3666. Levin, S.A. (1992) The problem of pattern and scale in ecology: the Robert H. MacArthur Award Lecture. Ecology 73(6): 1943-1967. Levin, S., 2003. Complex adaptive systems: exploring the known, the unknown and the unknowable. Bulletin of the American Mathematical Society, 40, 3-19. Linder HP, Bykova O, Dyke J, Etienne RS, Hickler T, Kühn I, Marion G, Ohlemüller R, Schymanski SJ, Singer A (2012) Biotic modifiers, environmental modulation and species distribution models. Journal of Biogeography 39:2179-2190. Liu, J., Williams, J.R., Zehnder, A.J.B., Yang, H., 2007. GEPIC – modelling wheat yield and crop water productivity with high resolution on a global scale. Agricultural Systems, 94, 478-493. Lloyd CD (2005) Assessing the effect of integrating elevation data into the estimation of monthly precipitation in Great Britain. Journal of Hydrology 308: 128-150. Luo Y, Randerson J, Abramowitz G, Bacour C, Blyth E, Carvalhais N, Ciais P, Dalmonech D, Fisher J, Fisher R et al. 2012. A framework of benchmarking land models. Biogeosciences Discussions 9: 1899–1944. Mace GM, Norris K, Fitter AH (2012) Biodiversity and ecosystem services: a multilayered relationship. Trends in Ecology and Evolution 27: 19-26. May RM, Levin SA, Sugihara G (2008) Complex systems: Ecology for bankers. Nature 451, 893-895. McCarthy HR, Luo Y, Wullschleger SD (2012). Integrating empirical–modeling approaches to improve understanding of terrestrial ecology processes. New Phytologist 195: 523-525. McGuire AD, Melillo JM, Joyce LA, Kicklighter DW, Grace AL, Moore III B, Vorosmarty CJ (1992) Interactions between carbon and nitrogen dynamics in estimating net primary productivity for potential vegetation in North America. Global Biogeochemical Cycles 6: 101-124. Melbourne, B.A., Chesson, P., 2006. The scale transition: scaling up population dynamics with field data. Ecology, 87, 1478-1488. Millennium Ecosystem Assessment 2005. Ecosystems and human well-being : synthesis. Washington, DC, Island Press. Nakićenoić N, Swart R (2000) Special report on emission scenarios. A special report of working group III of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK Noss RF (1990) Indicators for monitoring biodiversity: a hierarchical approach. Conservation Biology 4: 355-364. Overland, J.E., Wang, M., Bond, N.A., Walsh, J.E., Kattsov, V.M., Chapman, W.L., 2011. Considerations in the Selection of Global Climate Models for Regional Climate Projections: The Arctic as a Case Study*. Journal of Climate, 24, 1583-1597. Pan YZ, Gong DY, Deng L, Li J, Gao J (2004) Smart distance searching based and DEM informed interpolation of surface air temperature in China. ActaGeographicaSinica 59: 366–374 (in Chinese) Parton WJ, Hanson PJ, Swanston C, Torn M, Trumbore SE, Riley W, Kelly R (2010) ForCent model development and testing using the Enriched Background Isotope Study experiment. Journal of Geophysical Research –Biogeosciences 115. doi:G04001 Paul KI, Reeson A, Polglase P, Crossman N, Freudenberger D, Hawkins C (2013) Economic and employment implications of a carbon market for integrated farm forestry and biodiverse environmental plantings. Land Use Policy 30:496-506. Pavlick R, Drewry DT, Bohn K, Reu B, Kleidon A (2013) The Jena Diversity-Dynamic Global Vegetation Model (JeDi-DGVM): a diverse approach to representing terrestrial biogeography and biogeochemistry based on plant functional trade-offs. Biogeosciences 10:4137-4177. PBL. (2012) Roads from Rio+20: Pathways to achieve global sustainability goals by 2050, The Hague, The Nederlands. Phillips SJ, Dudík M (2008) Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography 31:161-175. Pielke Sr. RA, Wilby RL. 2012. Regional climate downscaling: what's the point? Eos, Transactions American Geophysical Union 93(5): 52-53. Prudhomme C, Davies H (2009) Assessing uncertainties in climate change impact analyses on the Page 31 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
river flow regimes in the UK, Part 1: baseline climate. Climatic Change 93: 177-195 Rahbek C, Gotelli NJ, Colwell RK, Entsminger GL, Rangel T, Graves GR (2007) Predicting continental- scale patterns of bird species richness with spatially explicit models. Proceedings of the Royal Society B-Biological Sciences 274:165-174. Raisanen J (2007) How reliable are climate models? Tellus A 59: 2-29. Randall, D.A., R.A. Wood, S. Bony, R. Colman, T. Fichefet, J. Fyfe, V. Kattsov, A. Pitman, J. Shukla, J. Srinivasan, R.J. Stouffer, A. Sumi and K.E. Taylor (2007) Climate models and their evaluation. In Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK and New York, NY, USA. Raskin, P., Banuri, T., Gallopin, G., Gutman, P., Hammond, A., Kates, R., Swart, R., 2002. Great transition. Umbrüche und Übergänge auf dem Weg zu einer planetarischen Gesellschaft, Materialien Soziale Ökologie, 20. Raskin, P.D., 2005. Global scenarios: background review for the Millennium Ecosystem Assessment. Ecosystems, 8, 133-142. Rockström, J., W. Steffen, K. Noone, Å. Persson, F.S. Chapin, III, E.F. Lambin, T.M. Lenton, M. Scheffer, C. Folke, H.J. Schellnhuber, B. Nykvist, C.A. de Wit, T. Hughes, S. van der Leeuw, H. Rodhe, S. Sörlin, P.K. Snyder, R. Costanza, U. Svedin, M. Falkenmark, L. Karlberg, R.W. Corell, V.J. Fabry, J. Hansen, B. Walker, D. Liverman, K. Richardson, P. Crutzen, and J.A. Foley, 2009: A safe operating space for humanity. Nature, 461, 472475, doi:10.1038/461472a. Rodríguez, J. P., Beard Jr, T. D., Bennett, E. M., Cumming, G. S., Cork, S. J., Agard, J., Dobson, A. P. and Peterson, G. D. 2006. Trade-offs across space, time, and ecosystem services. Ecology and Society, 11. Rondinini, C., Stuart, S., & Boitani, L. (2005). Habitat suitability models and the shortfall in conservation planning for African vertebrates. Conservation Biology, 19(5), 1488–1497. Rondinini, C., di Marco, M., Chiozza, F., Santulli, G., Baisero, D., Visconti, P., … Boitani, L. (2011). Global habitat suitability models of terrestrial mammals. Philosophical Transactions of the Royal Society B: Biological Sciences, 366(1578), 2633–2641. Rost, S., Gerten, D., Bondeau, A., Lucht, W., Rohwer, J., Schaphoff, S., 2008. Agricultural green and blue water consumption and its influence on the global water system. Water Resources Research, 44(9). Ruelland D, Ardoin-Bardin S, Billen G, Servat E (2008) Sensitivity of a lumped and semi-distributed hydrological model to several methods of rainfall interpolation on a large basin in West Africa. Journal of Hydrology 361: 96-117. Samanta S, Pal DK, Lohar D, Pal B (2012) Interpolation of climate variables and temperature modeling. Theoretical and Applied Climatology 107: 35-45. Sandø, A.B., Melsom, A., Budgell, W.P., 2014. Downscaling IPCC control run and future scenario with focus on the Barents Sea. Ocean Dynamics, 64, 927-949. Sardà-Palomera, F., Brotons, L., Villero, D., Sierdsema, H., Newson, S.E., Jiguet, F., 2012. Mapping from heterogeneous biodiversity monitoring data sources. Biodiversity and Conservation, 21, 2927-2948. Scheiter S, Langan L, Higgins SI (2013) Next-generation dynamic global vegetation models: learning from community ecology. New Phytologist 198:957-969. Schelhaas M-J, Eggers J, Lindner M, Nabuurs GJ, Päivinen R, Schuck A, Verkerk PJ, Werf DC, van der Zudin S (2007) Model documentation for the European Forest Information Scenario model (EFISCEN 3.1.3). Alterra report 1559. Alterra and European Forest Institute, Wageningen and Joensuu, Wageningen. p 118 Schellnhuber HJ, Frieler K, Kabat P (2014) The elephant, the blind, and the intersectoral intercomparison of climate impacts. Proceedings of the National Academy of Sciences 111:3225-3227. Schulp CJE, Thuiller W, Verburg PH (2014) Wild food in Europe: A synthesis of knowledge and data of terrestrial wild food as an ecosystem service. Ecological Economics 105:292-305. Schumacher S, Bugmann H (2006) The relative importance of climatic effects, wildfires and management for future forest landscape dynamics in the Swiss Alps. Global Change Biology 12:1435-1450. Seiler C, Hutjes RWA, Kruijt B, Quispe J, Anez S, Arora VK, Melton JR, Hickler T, Kabat P (2014) Modeling forest dynamics along climate gradients in Bolivia. Journal of Geophysical Research –Biogeosciences 119:758-775. Page 32 of 34
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Shang ZB, Guo Q, Yang DA (2001) Spatial pattern analysis of annual precipitation with climate information system of China. ActaEcologicaSinica 21: 689–694 (in Chinese). Sitch S, Smith B, Prentice IC, Arneth A, Bondeau A, Cramer W, Kaplan J, Levis S, Lucht W, Sykes M, Thonicke K, Venevsky S (2003) Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ Dynamic Global Vegetation Model. Global Change Biology 9:161-185. Sitch, S., Huntingford, C., Gedney, N., Levy, P. E., Lomas, M., Piao, S. L., Betts, R., Ciais, P., Cox, P., Friedlingstein, P., Jones, C. D., Prentice, I. C., and Woodward, F. I.: Evaluation of the terrestrial carbon cycle, future plant geography and climate-carbon cycle feedbacks using five Dynamic Global Vegetation Models (DGVMs), Glob. Change Biol., 14, 2015–2039, 2008. Stoy, P.C., Williams, M., Disney, M., Prieto-Blanco, A., Huntley, B., Baxter, R., Lewis, P., 2009. Upscaling as ecological information transfer: a simple framework with application to Arctic ecosystem carbon exchange. Landscape Ecology, 24, 971-986. Taylor, K.E., Stouffer, R.J., Meehl, G.A., 2011. An Overview of CMIP5 and the Experiment Design. Bulletin of the American Meteorological Society, 93, 485-498. Tebaldi, C., Knutti, R., 2007. The use of the multi-model ensemble in probabilistic climate projections.Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 365, 2053-2075.Knutti, R., Furrer, R., Tebaldi, C., Cermak, J., Meehl, G.A., 2010. Challenges in Combining Projections from Multiple Climate Models.Journal of Climate, 23, 2739-2758. O'Brien, S. &Vickerman, S. 2013. From principles to practice developing a vision and policy framework to make ecosystem management a reality. The Laws of Nature: Managing Ecosystems for A Sustainable Future. Thomas CD et al. (2004) Extinction risk from climate change. Nature 427: 145-148. Thonicke, K., Spessa, A., Prentice, I. C., Harrison, S. P., Dong, L., and Carmona-Moreno, C.: The influence of vegetation, fire spread and fire behaviour on biomass burning and trace gas emissions: results from a processbased model, Biogeosciences, 7, 1991–2011, doi:10.5194/bg-7-1991-2010, 2010. Thuiller W, Lafourcade B, Engler R, Araujo MB (2009) BIOMOD - a platform for ensemble forecasting of species distributions. Ecography 32:369-373. United Nations Environmental Program 2002. Global Environmental Outlook 3. Past, Present and Future Perspectives. Earthscan Publication Ltd, London. Van den Belt, M., McCallion, A., Wairepo, S., Hardy, D., Hale, L. & Berry, M. 2012. Mediated Modelling of Coastal Ecosystem Services: A case study of TeAwanui Tauranga Harbour, ManaakiTahoMoana project. Van Vuuren, D.P., Weyant, J., de la Chesnaye, F., 2006. Multi-gas scenarios to stabilize radiative forcing.Energy Economics, 28, 102-120. Van Vuuren, D.P., Kok, M.T.J., Girod, B., Lucas, P.L., de Vries, B., 2012. Scenarios in Global Environmental Assessments: Key characteristics and lessons for future use. Global Environmental Change, 22, 884-895. Van den Belt , M. Multi-scale Integrated Modeling for Sustainable Adaptive Systems. System Dynamics Society Conference, 27-31 July 2009 Albuquerque, New Mexico. Verkerk PJ, Anttila P, Eggers J, Lindner M, Asikainen A (2011) The realisable potential supply of woody biomass from forests in the European Union. Forest Ecology and Management, 261:2007-2015. Villa F, Bagstad KJ, Voigt B, Johnson GW, Portela R, Honzak M, Batker D (2014) A Methodology for Adaptable and Robust Ecosystem Services Assessment. PLoS ONE 9. Visconti P., Pressey R.L., Giorgini D. et al. (2011) Future hotspots of terrestrial mammal loss. Philosophical Transactions of the Royal Society B: Biological Sciences 366: 2693-2702. Visconti, P., Bakkenes, M., Baisero, D., Brooks, T., Butchart, S. H. M., Joppa, L., … Rondinini, C. (2015). Projecting global biodiversity indicators under future development scenarios. Conservation Letters doi:10.1111/conl.12159. Von Storch H, Zorita E, Cubash U (1993) Downscaling of global climate change estimates to regional scales: an application to Iberian rainfall in wintertime. Journal of Climate 6: 1161-1671. Walz, A., Braendle, J.M., Lang, D.J., Brand, F., Briner, S., Elkin, C., Hirschi, C., Huber, R., Lischke, H., Schmatz, D.R., 2014. Experience from downscaling IPCC-SRES scenarios to specific national- level focus scenarios for ecosystem service management. Technological Forecasting and Social Change, 86, 21-32. Wang YP, Law RM, Pak B (2010) A global model of carbon, nitrogen and phosphorus cycles for the Page 33 of 34
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
terrestrial biosphere. Biogeosciences, 7: 2261–2282. Wenzel, S., Cox, P.M., Eyring, V., Friedlingstein, P., 2014. Emergent constraints on climate‐carbon cycle feedbacks in the CMIP5 Earth system models. Journal of Geophysical Research: Biogeosciences. Weedon, J.T., Cornwell, W.K., Cornelissen, J.H., Zanne, A.E., Wirth, C., Coomes, D.A., 2009. Global meta‐analysis of wood decomposition rates: a role for trait variation among tree species? Ecology Letters, 12, 45-56. Wiens, J.A., 1989. Spatial Scaling in Ecology. Functional Ecology, 3, 385-397. Wiens, J. A., & Bachelet, D. (2010). Matching the multiple scales of conservation with the multiple scales of climate change. Conservation Biology,24(1), 51-62. Wilby RL, Dawson CW, Barrow EM (2002) SDSM-a decision support tool for the assessment of regional climate change impact. Environmental Modelling and Software 17: 147-159. Williams, M., Bell, R., Spadavecchia, L., Street, L.E., Van Wijk, M.T., 2008. Upscaling leaf area index in an Arctic landscape through multiscale observations. Glob Chang Biol, 14, 1517-1530. Williams M, Richardson AD, Reichstein M, Stoy PC, Peylin P, Verbeeck H, Carvalhais N, Jung M, Hollinger DY, Kattge J et al. 2009. Improving land surface models with FLUXNET data. Biogesciences 6: 1341–1359. Woodward FI, Lomas MR (2004) Vegetation dynamics – simulating responses to climatic change. Biological Reviews 79:643-670. Wu & Li 2006 Wu, J., Li, H. Concepts of scale and scaling (2006) Scaling and Uncertainty Analysis in Ecology: Methods and Applications, pp. 3-15. Xu CY (1999) From GCMs to river flow: a review of downscaling methods and hydrologic modelling approaches. Progress in Physical Geography 23 (2): 229-249. Yue TX (2011) Surface Modelling: High Accuracy and High Speed Methods. CRC Press, New York Yue TX, Zhao N, Ramsey RD,Wang CL, Fan ZM, Chen CF, Lu YM, Li BL (2013) Climate change trend in China, with improved accuracy.Climatic Change120: 137-151. Yue TX (2015) Principles and Methods of Earth Surface Simulation. Science Press, Beijing (in Chinese) Zaehle S, Friedlingstein P, Friend AD (2010) Terrestrial nitrogen feedbacks may accelerate future climate change, Geophys. Res. Lett., 37: L01401. Zhang, N., Yu, Z., Yu, G., Wu, J., 2006. Scaling up ecosystem productivity from patch to landscape: a case study of Changbai Mountain Nature Reserve, China. Landscape Ecology, 22, 303-315 Zhou LS, Sun H, Shen YQ, Deng JZ, Shi YL (1981) Comprehensive Agricultural Planning of China. China AgriculturalPress: Beijing (in Chinese) Zorita E, von Storch H (1999) Theanalog method as a simple statistical downscaling technique: comparison with morecomplicated methods. Journal of Climate 12: 2474-2489.
Page 34 of 34
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
7 Building capacity for developing, interpreting and using scenarios and models 5
Coordinating Lead Authors: Carolyn Lundquist, Khaled Allam Harhash Lead Authors: Dolors Armenteras, Nakul Chettri, James Mwang’ombe Mwamodenyi, Vasyl Prydatko, Sandra Acebey Quiroga, Andriambolantsoa Rasolohery Contributing Authors: Kamaljit Sangha, Rosario Gomez, Fernando Santos-Martín, Shaun Awatere
Key message: 10
Regional, sub-regional and national similarities and differences currently exist in the capacity for scenario development and modelling for biodiversity and ecosystem services, and reflect historical, environmental, cultural, and economic factors.
7.1 Introduction 15
20
25
30
Previous chapters have introduced the methodologies for scenario analysis and modelling of biodiversity and ecosystem services, discussing a wide range of tools and methodologies that can be used to support assessment and decision-making needs within the IPBES platform. This chapter reviews the underlying capacity required to support scenario analysis and modelling across a broad range of spatial scales (global, regional, sub-regional, and local), and decision-making contexts. Key aspects of the technical capacity required for scenario development and modelling include strengthening human resources and infrastructure; improving regional and national access to, and inter-operability of, quality standardized data‐sets, including developing methods for better incorporation of local data and knowledge, and developing synergies with existing Millennium Ecosystem Assessments (MEAs) for data and scenario sharing; and improving access to, and guAidelines for, user-friendly software tools for scenario analysis, modelling and decision‐support systems. Capacity building for scenario analysis and modelling also must consider capacity to support development of effective strategies for mainstreaming scenario processes at different geographic scales to allow their integration into participatory approaches, decision‐making processes and public awareness across different policy, planning and management contexts. While regional, sub-regional and national similarities and differences currently exist in the capacity for scenario development and modelling for biodiversity and ecosystem services, this chapter presents a path forward to balancing human resources, infrastructure and data accessibility to enable scenario development and modelling at regional, sub-regional and national scales.
7.2 Understanding regional and cultural similarities and differences in perspectives on, and capacity for, scenario analysis and modelling 35
The UNDP defines capacity development for environmental sustainability as ‘the process through which individuals, organizations and societies obtain, strengthen and maintain their capabilities to set and achieve their own development objectives over time. Components of capacity include skills, systems, structures, processes, values, resources and powers that together, confer a range of political, managerial and technical capabilities’ (UNDP 2011). Within a framework of scenario analysis and Page 1 of 42
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
modelling of biodiversity and ecosystem services, this capacity includes the human resources and technical capacity required to support scenario analysis and modelling across a broad range of spatial scales (global, regional, sub-regional, and local), and decision-making contexts (Figure 7.1). Key aspects of the technical capacity required for scenario development and modelling include strengthening human resources and infrastructure; improving regional and national access to, and interoperability of, quality standardised data‐sets, including developing methods for better incorporation of local data and knowledge, and developing synergies with existing MEAs for data and scenario sharing; and improving access to, and guidelines for, user-friendly software tools for scenario analysis, modelling and decision‐support systems. Capacity building for scenario analysis and modelling also must consider capacity to support development of effective strategies for mainstreaming scenario processes at different geographic scales to allow their integration into participatory approaches, decision‐making processes and public awareness across different policy, planning and management contexts.
Figure 7.1 Capacity building requirements for scenario analysis of modelling of biodiversity and ecosystem services.
7.2.1 Regional, national, cultural, and thematic similarities and differences in capacity for scenario analysis and modelling 20
25
Regional, sub-regional and national similarities and differences currently exist in the capacity for scenario development and modelling for biodiversity and ecosystem services. These differences are a reflection of political history, environmental variability, information and communications technology, economic capacity, population size and education, and many other factors. Understanding the context behind these existing differences can be used to both alleviate differences, and enhance the uptake of Biodiversity and Ecosystem Services (BES) modelling and scenario analyses into environmental decisionmaking. Page 2 of 42
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
Differences in capacity are most noticeable when comparing infrastructure to support scenario analysis and modelling across nations and regions. Significant differences are apparent when comparing economic investment priorities by different governments, which include prioritization of research on biodiversity and ecosystem services (Figure 7.2). In many countries, innovations in BES models are supported by government funding to academic and research institutions or through direct funding to government ministries to develop and implement management solutions. However dependence on external organizations (e.g., environmental non-governmental organizations) to provide technical and financial resources is common in many nations with smaller economies, with resulting challenges with long-term viability and uptake by local stakeholders (Morrison et al. 2010; Horigue et al. 2012, Mills et al. 2014). Technical capacity often correlates with financial resources, and includes aspects ranging from support and contributions of information toward global datasets (see e.g., section 7.3), human capacity to collect and translate information into geospatial datasets suitable for modelling and scenario analysis, and support participation in collaborative projects, training and networking. Information and communication technology differs widely across nations and regions (Figure 7.2), and underpins access to web-based or cloud-based global datasets, training resources, and software tools. Technical capacity also includes the human skills required to support engaging with stakeholders, translating stakeholder objectives into models and scenarios, skills in software and geospatial datasets, and skills in communicating model results and engaging with the public, with stakeholders and with decision and policy makers (see section 7.8). Disparities in authorship of scientific papers on BES models are apparent, with over 50% of peerreviewed resources authored from one of three countries (USA, England, Australia), reflecting disparities in investment in both human and technological capacity driving innovation in BES modelling (Figure 7.2). Other factors also influence existing capacity for BES models and scenario analysis. Thematic biases are apparent, with BES models and scenarios more commonly used to support decision-making in terrestrial ecosystems compared to marine and freshwater ecosystems (FRB 2013). Socio-economic drivers also result in differing capacity across topical issues, with model capacity biased toward resource based modelling (e.g., fisheries, forestry, agriculture), with less capacity supporting models without direct underlying economic gain. Increased understanding and integrating of ecosystem service concepts is reversing this trend and international commitments to platforms such as IPCC and IPBES are resulting in models that are more holistic, and include modelling of environmental (e.g., water quality), climate (e.g. coastal inundation, sea level rise, ocean acidification) and cultural and community objectives. Cultural differences exist in recognition of the importance and uses of scenarios and models into planning process. These include biases from lack of cross-cultural engagement and understanding, but also reflect spatial scales of local or traditional management practices, customary and participatory decision-making, and oral knowledge and data gathering. Cultural frameworks also guide taboos about types of management and decision-making frameworks that are acceptable, data collection and data sharing. The separation of people and nature (or lack thereof) translates to local and national priorities for institutional priorities, and communication, coordination, and collaboration within and between government institutions. Finally, external drivers can influence the use of BES scenarios and modelling. Political agendas which vary on temporal scales of political terms can provide impetus for new innovations and decision-making, but also can cause reversals of existing decisions and environmental commitments (e.g., Australia’s 2014 Page 3 of 42
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
decision to repeal its carbon tax, and resulting changes in institutional support for climate-related research). National and regional environmental policies often have topical biases (e.g., biases toward terrestrial over marine and aquatic policies) that drive funding, data collection, and decision-making. Similarly, non-governmental organizations have research priorities that result in biases in research agendas, e.g., focus on protected area implementation rather than sustainable agriculture or water quality. While we have demonstrated the existence of regional, sub-regional and national similarities and differences in the capacity for scenario development and modelling for biodiversity and ecosystem services, our emphasis throughout the remainder of this chapter will be on identifying key capacity requirements in human resources, infrastructure and data accessibility, and processes to enhance and strengthen capacity at regional, national and local scales to enable scenario development and modelling of biodiversity and ecosystem services.
A
B
15
Page 4 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
C
5
10
Figure 7.2 Regional differences in capacity to support BES modeling and scenario analysis. (A. Peer-reviewed publications of scientific and technical journal articles based on search of ISI Web of Science citation database for all years (1900-current) for nationality of authors of publications with TOPIC: (ecosystem service*) OR TOPIC: (biodiversity*) AND TOPIC: (model* OR scenario*); B. Research and development expenditure (% of GDP). Current and capital expenditures (both public and private) on creative work undertaken systematically to increase knowledge, including knowledge of humanity, culture, and society, and the use of knowledge for new applications. Data source: United Nations Educational, Scientific, and Cultural Organization (UNESCO) Institute for Statistics http://databank.worldbank.org/data/views/reports/tableview.aspx C. Internet users: people with access to the world wide network. Data source: World Bank/World Development Indicators (http://data.worldbank.org/indicator/IT.NET.USER)).
7.3 Capacity to enhance human resources and skill base 15
Key message: Various human resource and skills base required for scenario and modeling for biodiversity and ecosystem services are not evenly spread across the regions and several strategies can be employed to improve the prevailing situation.
20
25
7.3.1 Human resource and skill base required for BES scenarios and modelling BES scenarios and modelling requires human resource skilled in various fields. These include ecological modellers, social scientists, economists, lawyers, GIS specialists, policy analysts and programmers as well skills in data collection such ecologists and taxonomists that collect data related to flora and fauna, soil scientists and other experts. In addition, facilitators experienced in participatory approaches are necessary considering the fact that local and traditional knowledge or stakeholder input is increasingly being sought to provide data useful in creating “storylines”.
7.3.1.1 Human resources to enhance technological aspects of BES scenarios and models 30
In a scenario and modelling for biodiversity and ecosystem services exercise, various experts are required ranging from; scientists and/or practitioners with expertise in measuring the conditions and trends in the ecosystem services of interest and relevant human well-being indicators; to social and political analysts with local knowledge and experience to work on relevant responses. A range of skills is required to support BES scenarios and modelling, which includes: Page 5 of 42
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
Data collection, Model development, Model (and software) use and analysis, Model interpretation and translation for policy makers, Communicating results and findings. Scientists and practitioners with a deep understanding of the different components of a system and how they are interconnected may effectively inform complex and nuanced narratives about possible futures of a region. Linked, multiscale scenarios addressing ecological dynamics may require the participation of scientists with specific technical skills.
10
7.3.1.2 Human resources to enhance integration of scenarios and models into decisionmaking
15
20
25
30
Scenario analysis plays an important function where it provides an approach and opportunity to reflect on and think through the possible implications of alternative decisions in a structured manner. Simply, a scenario avails a platform/basis that allows decision units (governments, agencies) to reflect on how changes in their respective context (that is, developments beyond their immediate spheres of influence) may affect their decisions. Scenario analysis is increasingly being used to explore the potential future implications of different approaches for sustaining ecosystem services in the face of growing pressures. For example, the Millennium Ecosystem Assessment (MA) offers four global scenarios based on the implications of different assumptions regarding approaches toward governance and economic development (regionalized versus globalized) and toward ecosystem service management (reactive versus proactive) (see MA 2005a, Carpenter et. al. 2006). Communication expertise is crucial in dissemination of results of a scenario and modelling exercise. The dissemination of results should be guided by clear communication goals which define the specific target audiences that will determine the appropriate means of communication. Target audiences are defined by their profession and areas of focus/interest which will influence the content and style of materials used. Full reports are useful as reference documents but the content and conclusions must be synthesized into short and specific messages that will resonate with the target audience. The main messages are usually not simply a summary of all the information produced but rather a more strategic culling of the points most relevant to each audience, presented in a way that promotes the credibility of the findings. This means backing up important statements with data and examples and using easy-tounderstand graphs, illustrations, and tables. Care must be taken to ensure that outputs developed for communication relevant to policy makers is not policy prescriptive but present information relevant to policy by ensuring inclusion of information most relevant to the choices being faced by the audiences.
35
7.3.1.3 Human resources to enhance integration incorporation of indigenous and local knowledge
40
A wide range of qualitative and quantitative participatory methods have been used to facilitate engagement of stakeholders in scenario development. These include workshops; scenario-based stakeholder engagement; facilitated discussion and ranking (Tompkins et. al. 2008); cooperative discourse (e.g. Renn 2006); multi-criteria evaluation (e.g. Madlener et. al. 2007; Kowalski et. al. 2013); Page 6 of 42
First Order Draft
5
10
15
20
IPBES Deliverable 3c
DO NOT CITE
conceptual system modelling (e.g. Magnuszewski et. al. 2005); and mediated or dynamic systems modelling (Bousquet et. al. 2002; van den Belt 2004; Castella et. al. 2005). Reed et al. (2013) in a summary analysis suggests the following steps as necessary to facilitate effective stakeholder participation in scenario development: I. Define context (biophysical, socio-economic and political) and establish a basis for stakeholder engagement in scenario development; II. Systematically identify and represent relevant stakeholders in the process; III. Define clear objectives for scenario development with solutions including spatial and temporal boundaries; IV. Select relevant participatory methods for scenario development: a. To set up relevant construction of scenarios, b. To evaluate and select scenarios for further investigation, c. To support decision-making based on scenarios. The incorporation of local and indigenous knowledge is recognized as a critical aspect of participatory processes, particularly for qualitative ecosystem service values that cannot easily be modelled. In addition to building relevance, credibility and legitimacy, local communities/stakeholders have particular and unique knowledge about the ecosystems where they live and work, as well as about their own associated well-being. Often the links between ecosystem services and human well-being are complex and obscured and must be teased out with contextual knowledge. In addition, published data about ecosystems and societies in many parts of the world are scarce and can be much enhanced with local knowledge.
7.3.2 Current state of skills in scenario and modelling for BES distribution across the regions 25
30
The use of scenario and modelling for biodiversity and ecosystem services has mainly been undertaken in the developed countries, with limited exercises in the developing countries often with technical and financial support from developed countries. A large number of these BES exercises are related to past and ongoing assessments such as the MA (MA 2005b). In addition, several processes are ongoing in Europe such as the PEER (Partnership for European Environmental Research; see Maes 2012). The bulk of relevant literature is authored by researchers from or based in developed countries (Figure 7.2). This clearly indicates that the current distribution of skills on scenario and modelling of biodiversity and ecosystem services is concentrated in developed countries. In particular, human resources and skills in both African and Southeast Asian countries are more limited due to limiting training opportunities, infrastructure, and financial resources.
35
7.3.3 Current mechanisms and opportunities for mobilization and enhancement of human capacity to perform scenarios and modelling for BES
40
Training courses are an important part of building human capacity to support BES models and scenarios analysis. Many global programs exist, led by IGOs and NGOs. For example, UNEP-WCMC provides opportunities to partners for human resource and skills development through a wide range of training courses, workshops, internships and collaborative projects. Partnerships and initiatives are convened Page 7 of 42
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
that enable participants to share skills and learn from each other. This also includes hosting of staff seconded from other institutions enabling learning and sharing of expertise. Further, UNEP-WCMC relates closely with a number of European and North American Universities through teaching and research supervision. UNEP-WCMC partners with the University of Cambridge to train the next generation of conservationists through an MPhil in Conservation Leadership, supervising student consultancy and placement projects, and providing programme oversight and direction as part of the Steering Committee (www.unep-wcmc.org/expertise). There are efforts to train practitioners in some of the tools (such as InVEST) in scenario and modelling through short courses through projects such as CHIESA (Climate Change Impacts on Ecosystem Services and Food Security in Eastern Africa) under ICIPE (International Centre for Insect Physiology and Ecology) with coordination among four universities in Africa and Europe. Other opportunities for capacity building include developing a “fellows program” for young scientists to partner with more senior or experienced scientists.
7.4 Infrastructure to support BES modelling and scenario analysis Key messages:
20
Capacity building can take advantage of the exponential development of e-society and e-infrastructure related to biodiversity and ecosystem services, and to build-up actively corporative e-platforms, georeferencing of archived and ‘civil data’, and to conduct improving new e-geosociety.
7.4.1 Institutional infrastructure within other international agreements
25
30
35
40
Institutional infrastructure underpins all international MEAs, such CBD COP decisions and strategic goals, and expected contributions to Aichi Targets 1, 4, 5, 6, 9, 11, 12, 13, 18 and 19 (GBIO, 2012). Key priority capacity building criteria and requirements are similarly required to support the IPBES platform as summarized in working documents IPBES/1/INF/10 and IPBES/2/INF/13, and underpinning IPBES work programme objectives (Deliverable 1a,b, 2a, 3b, and 4c). Together the numerous international agreements incorporating biodiversity and ecosystem services function as a complex mechanism, and in which modern mobiles’ applications act as new gearwheel. It pays attention to the spectacular development of e-society and e-infrastructure related to α-, β-, and γbiodiversity and to respective research and conservation that take place on level of billions of motivated mobile- and internet users. Database contents are varied, ranging from information on habitat and ecosystems to individual species distributions such as locations of plant species or marine mammals. In 1990s basic BES modelling and scenario analysis infrastructure required: enthusiasts and trained staff, geospatial datasets, field data collection, data processing, and decision-making support and networking especially at the national level. Very often protected areas were the primary regions of interest for BES models. The first scenarios of 1990s were largely the result of expert evaluation, quantitative models, and models were typically for present time values, and rarely incorporated long-range, future forecasting. Currently (with an eye to Archi Targets 2020) BES modelling infrastructures now also requires trans-national cooperation and data collection, trained staff in geospatial and modelling software and dedicated specialists to translate complex BES models into decision-making. Many of these Page 8 of 42
First Order Draft
5
10
15
20
25
30
35
40
IPBES Deliverable 3c
DO NOT CITE
aspects rely on integrated and compatible data management structures at national, regional and international scales to allow comparison within and between nations, and minimize duplication of efforts to support data management, collection and analysis between MEAs. Global experience of previous years (examples) - MDG7, CBD and SBSTTA, CITES, CMS, ‘Treaty’, Ramsar, WHC, IPPC, WCMC, GBIF, WB and GEF projects, EEA, GLOBIO, FP7, consortiums and many others - has formed a unique library of scales, behaviours and evolution of networks important for the biodiversity modelling & scenario development and respective infrastructures analyses, and has allowed a clear comparison of advantages and disadvantages. There is indeed a diversity of targets, members, involved institutions, levels of biodiversity data management, maps, tools, DPSIR- indicators and IFSs, and models/scenarios. Together with biodiversity data sources and global-, regional-, local-, semi-private- and private platforms it forms much bigger net for BM&SD (examples): MEAs (general – 43, inland water pollution – 24, fisheries – 24, flora and fauna – 36, marine pollution – 59, atmospheric pollution – 23, nuclear -22, miscellaneous – 15) and respective DBs; national BD-resources (MEA-, PA-, SoE-, annual-statistic- and cadastres related); FAO (AgriEnvironmental indicators); GBIF; GBIC/GBIO; INSPIRE; EUROSTAT; ECLAC (CEPALSTAT); ECA Databank (StateBase); World Biodiversity Database (WBD); A Pan-European Species-directories Infrastructure (PESI); The Arctic Biodiversity Data Service; WB World Development Indicators (WDI) – mainly driving forces indicators; The Global Invasive Species Database (GISD); Ocean Biogeographic Information System Spatial Ecological Analysis of Megavertebrate Populations (OBIS-SEAMAP); MarineBio Species Database (MarineBio); EEA (Natura 2000, CLIM 026, SEBI, others); Eionet ETC DB; IUCN Red List DB; The Map of Life (MOL); USDA Plants DB; Freshwater Ecoregions of the World (FEOW); BirdLife International DB; The World Database on Protected Areas (WDPA); All Catfish Species Inventory; Arctos; AntWeb; ASEAN Biodiversity Information Sharing Service (BISS); FishBase; Freshwater Ecoregions of the World (FEOW); HerpNET; Reptile Database; iNaturalist; Integrated Botanical Information System (IBIS); Integrated Taxonomic Information System (ITIS); Natural History Information System; NatureServe; WikiSpecies; Naturdata; eBird; Netherlands Biodiversity Information Facility (NLBIF); A Database of Plant Biodiversity of West Bengal (WBPBDIVDB); ScaleNet; BioDat and IS BIODIV (NIS, RF); NatureServe (a public-private network and independent conservation data centres (CDCs) – USA, Canada, Latin America); The Russian Arctic Biogeographical Database (RABD, RF); BioModel (UKR); local-, thematic-, events-, ad hock-webresources on biodiversity; and at last mobile-application-based e-resources on species/habitats/tracks/localities (>240 titles). Pragmatic view on infrastructure is (simplification): GBIO {100 GBIO’s experts}; BioFresh {FP7, 9 countries, 19 institutions → GLM,GAM, MARS}, The Arctic Biodiversity Data Service {9 countries, 57 institutions}; WBD {21 separate projects → GRID}; MOL {55 datasets, 10 global partners, Google App Engine, Google Earth Engine, Google Maps Engine, CartoDB→ global location of species); NLBIF {258 institutions}; BioModel {8 EEBIO countries → GLM}. Everything is in a pulsating, breathing mode, which cannot be controlled from any one centre. At the same time, MEAs (and currently software vendors) serve as core initiatives. The well-known CBD infrastructure has changed since the 1992 convention which involved 194 Parties, 168 Signatures (CBD, 2014) each of which developed focal points, institutional network, basic legislation platforms, budgets, human resources, and data sets. CBD pushed the creation of numerous new institutions, MEAs, fundraisings, and data collection initiatives. Nevertheless, an understandable decrease of new contracting parties joining the convention has taken place, but, by no means does that Page 9 of 42
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
decrease the quality of CBD infrastructure. Thus, the indicator ‘number of MEA members’ is not so informative in the case study as it might seem at a first glance. GBIF arose from a recommendation in 1999 by the Biodiversity Informatics Subgroup of the Megascience Forum set up by OECD, and currently it provides access to more than 500 million records, shared freely by hundreds of institutions worldwide. Similar to CBD but contrary to an understandable decrease of new countries joining the GBIF in previous years, the total number of records published through GBIF over time has increased successfully and illustrates common tendencies of good capacity building related to biodiversity (GBIF, 2014; Figure 7.3).
Figure 7.3 GBIF: Expected advantageous infrastructural changes by 2020. Data source: GBIF (2014). Graph and interpretation: IPBES Deliverable 3 (c), Ch.7
Similar data-collection- and capacity building activity exhibits today many in other societies from large professional scientific organizations with databases on topical and training expertise, to taxon-specific databases and expert groups such as the Golden Jackal informal study Group in Europe (GOJAGE).
7.4.2 Key aspects of infrastructure requirements for BES modelling and scenario analysis 20
25
In the global context real gaps and problems of infrastructures tend to become invisible, as providers often do not mention them on the web portals. Table 7.1 provides a simple illustration of hierarchy inside the BM&SD infrastructure as well practically important rather than empirical steps required for modelling. A lighted interrelation of key sections of the matrix matched needs in knowledge, ICT, digital data, methodology, policy. Multi-criteria valuation and indicators can contribute more to assess how numerous factors affect the improvement of infrastructures. A portion of it has been transformed by us to a thematic map in the Section 1 of Chapter 7. Page 10 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
This matrix (Table 7.1) resonates well within the vision of IPBES Deliverable 4(c) on identification of seven families of policy support tools and methodologies: assembling data and knowledge, assessments and evaluation, participatory process, designing of policy instruments, implementation and enforcement, [new] capacity building, social learning and innovation and adaptive governance. 5 Table 7.1 Simplified matrix on biodiversity modelling practice and infrastructure to be involved
Tasks, activities, milestones
Steps/Needs BM&SD INFRASTRUCTURE ADAPTATION AND IMPROVEMENT BIODIVERSITY CONSERVATION LAWMAKING, APPROVAL AND IMPLEMENTATION
Period
↑↓
Years
↑
Years Months/ Years
↑
DECISSION-MAKING
Submission of results to decision makers and stakeholders, negotiations
Weeks/ Months
MODELLING OF SCENARIOS (2015, 2030, 2050…)
↑
Days/ Weeks
MODELLING OF PREDICTIONS
↑
Days/ Weeks
↑
Days/ Weeks
SDM (all types) [species-per-area, pressurebased-models, MSA, GLM etc.]
Weeks/ Months
FILTERING, ERROR DETECTION, ACCURACY IMPROVEMENT MODELLING OF CURRENT STATE OF BIODIVERSITY AND/OR SPECIES HABITAT CHANGES IT PROCESSING, BIODIVERSITY DATA INTERPRETATION, PRE-MODELLING Climate-changedata incorporation (Bioclim, IMAGE…)
Land-use-change data incorporation
RS-data incorporation (m, km)
Basic digital map (1:50.000, 1:200.000…)
Sea/Ocean related data incorporation
(PROFESSIONAL) IT ACTIVITY PLANNING PRE-PROJECT STAGE
DB
●IT processing and analyze ●Completing GIS ●Data receiving/acquisition, adaptation ●Cooperation with dataowners, decision-makers ●Development of Technical Requirements matched Data Base and future GIS-product (see CPT-step) ●Initiating, creation of modelling point/platform/project
●CPT: Capacity Planning Tool usage (to identify system design requirements and model performance, and scalability of proposed GIS operations) ●Receiving key questions
Months (0.5…1)
Infrastructure and HR to be involved Ministries, Committees, Academies, NGOs; ITC Countries, MEAs Countries, MEAs COPs Decision-makers, institutions (global, regional, national) Direct: modelling team, recipients, GOs&NGOs representatives, internal/external experts, scientists. Indirect: data owners, software providers
Direct: modelling team, beneficiary, recipients, representatives, internal/external experts and scientists. Indirect: data- and software providers
Weeks/ Months?
Direct: trained IT staff (RS, GIS, BM&SB) Indirect: CPT provider(s)
Weeks/
Beneficiary, future
Page 11 of 42
First Order Draft
IPBES Deliverable 3c
from decision-makers and stakeholders, and potential users
10
Months
recipients, decision- and policy makers Funds, beneficiaries, partners
FINDRAISING, BUDGETING
↑
Months/ Years
Post-Graduate Education (DB, GIS, indicating, modelling, mapping)
↑
Years (1…2)
Universities, trainings
Ecology and ITC (e-Ecology)
Years (18…20)
Countries’ educational net of all levels
Basic Education
5
DO NOT CITE
The matrix displays better potentially important but problematic infrastructural points in scope of our practical long-term experience on biodiversity modelling on the level of international and national projects. It suggests that the following parts of the infrastructure remain potentially important and, at the same time, problematic (highlighted in red): basic education needs (combined knowledge in ecology and ICT); post-graduate education needs (DB, GIS, indicating, modelling, mapping); fundraising and budgeting; trained IT specialists on a stage of modelling planning; basic digital maps of good scale for a future GIS; filtering, error detection, accuracy improvement – at the final stage; law-making and implementation of the modelling results; biodiversity conservation as a response to BM&SD; next adaptation and improvement of infrastructures.
7.5 Improving regional and national access to, and inter-operability of, quality standardised data‐sets that are appropriately prepared and served globally 15
Key message: Key aspects of the technical capacity required for scenario development and modelling include improving regional and national access to, and inter-operability of, quality standardised data‐sets. 20
25
30
7.5.1 What are data and why they are important for scenarios and modelling? During last 250 years, researchers and scholars have collected, produced and used a wealth of information from various sources, and guided the conservation and economic development of many nations, addressing subsistence livelihood of one third of the poverty ridden humanity and shaping the world economy through its goods and services. Such information gathered systematically and used for some purpose are generally terms as ‘data’. Therefore, data are a collective term for information gathered from any field of discipline (i.e. socio-economic data, floral data, faunal data, soil data, forest inventory data, precipitation data, temperature data etc). However, for the purpose of this section, biodiversity and ecosystem related data will be the focus for discussion. It is to be noted that the biologists have explored and documented 1.9 million of species records (Chapman 2009) from estimated 11 million species on our planet (Raven and Yeates 2007). Many of these data, the fundamental descriptors of living systems, are still locked away and inaccessible (GBIF 2013; Pimm et al. 2014). In
Page 12 of 42
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
order to preserve the variety of life on Earth, it is important to understand it better (Sala et al. 2000; GBIF 2013). The world’s governments missed their target to reduce the rate of biodiversity loss by 2010 (Secretariat of the CBD 2010) and the progress so far for 2020 Aichi targets seems to be far from achievable (Secretariat of the CBD 2014). These analyses were interpreted through existing datasets by utilizing biodiversity and ecosystem service modelling and scenario development processes (e.g. Sala et al. 2000; Leadley 2010; Pereira et al. 2010). One of the main reasons for this global failure was the shortage of comprehensive indicators and associated accessible data (Butchart et al. 2010; Secretariat of the CBD 2010). To create appropriate policies to protect biodiversity we must understand what they contain, how the species within interact, and how they might respond to changes and pressures, both natural and man-made (Mace et al. 2010). With the adoption of the Strategic Plan for Biodiversity 2011-2020, including the Aichi Biodiversity Targets, governments have re-affirmed the importance of preserving and restoring biodiversity and maintaining the planet’s ecosystem services. Thus the data does not merely contribute towards meeting these goals: it is fundamental to achieving them.
15
7.5.2 What are the relevant data types/sets available at global, regional and national levels?
20
25
30
35
Realizing the importance of data, many global, regional and national initiatives have progressed to archive different forms of data for the use and applications on various decision making processes (MA 2005a; Chettri et al. 2008; Yahara et al. 2014; Viciani et al. 2014). This is true even at the global level where Multilateral Environmental Agreements such as UNFCCC, CBD, RAMSAR, CITES and Millennium Development Goals are supported by a range of primary and secondary data both at national and global levels to reach common conservation and development goals. This was evident from the progressive and refining reporting from IPCC Report 4 (IPCC 2007) and IPCC Report 5 (IPCC 2014). Parties to such conventions are obliged to develop clearing housing mechanisms with established national level accessible datasets. These practices have significantly contributed to dataset development processes and accessibility. Thus, through such conservation and development processes, biodiversity and ecosystem data (among others) are becoming more relevant for shaping the future of this planet earth and humanity for generations to come. Some of the promising efforts on developing global biodiversity database could be referred to Encyclopedia of Life (Parr et al. 2014) and Global Biodiversity Information Facility (Robertson et al. 2014) among others. The GBIF data has made significant progress on publishing about 500 million species checklist and 50 million geo-referenced data worldwide (Figure 7.4) with the US being the highest contributor among the top ten countries (Figure 7.2). Efforts have also been made to develop thematic datasets on forests (Gilani et al. 2014; Pfeifer et al. 2014), wetlands (Lehner and Döll 2004; Chaudhary et al. 2014) and mountain ecosystems (Chettri et al. 2008; Guralnick and Neufeld 2005; Gurung et al. 2011). However, it is a paradox that biodiversity rich countries and regions are not the major contributors of global datasets.
Page 13 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 7.4 Density of geo-referenced species occurrence records published through GBIF till July 2014. Top ten contributing countries of geo-referenced data include the United States, Sweden, United Kingdom, Australia, Netherlands, Germany, France, Finland, Norway, and Spain) (Source: http://www.gbif.org/ocuurange)
5
7.5.3 What are the database management issues?
10
15
20
25
There are five broad groups of issues that stand out in any examination of research data access and sharing regimes (Arzberger et al. 2004). 1. Technological issues: Broad access to research data, and their optimum exploitation, requires appropriately designed technological infrastructure, broad international agreement on interoperability, and effective data quality controls; 2. Institutional and managerial issues: While the core open access principle applies to all science communities, the diversity of the scientific enterprise suggests that a variety of institutional models and tailored data management approaches are most effective in meeting the needs of researchers; 3. Financial and budgetary issues: Scientific data infrastructure requires continued, and dedicated, budgetary planning and appropriate financial support. The use of research data cannot be maximized if access, management, and preservation costs are an add-on or afterthought in research projects; 4. Legal and policy issues: National laws and international agreements directly affect data access and sharing practices, despite the fact that they are often adopted without due consideration of the impact on the sharing of publicly funded research data; 5. Cultural and behavioural issues: Appropriate reward structures are a necessary component for promoting data access and sharing practices. These apply to those who produce and those who manage research data. 6.
7.5.4 What are the standard requirements for quality data-sets? As new technologies and scientific approaches are evolving, modelling through new and old data could help us towards understanding our past and guide the future of humanity (Pimm et al. 2014). But, this Page 14 of 42
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
can happen only if we are able to enrich, maintain and use quality data effectively (GBIF 2013). This is possible when data, old and new, are archived in a structured and standardized form that would enable a vast range of uses, creating new opportunities for research and putting biodiversity-related policy making on a sounder footing. Therefore bundling of biodiversity and ecosystem data fed with modern tools can enable us to understand the trend and projections with scenarios. These could be important building blocks for future conservation and development goals.
7.5.4.1 Data quality 10
15
Quality refers to the proper description of uncertainties surrounding the production of the data (e.g., the techniques employed in their collection and archiving, and the measuring instruments and their calibration), the ability to ensure that the cited source and value are authentic, that the data retain integrity (complete and absent from introduced errors), and that they are secure against loss, destruction, modification, and unauthorized access. This is required to address inconsistent data with taxonomic anomaly limiting inter-operability across the temporal and geospatial scales (Natvi et al. 2009). Poor data limits the opportunity for comparison and inter-operability challenging possibility achieving the objective of data fed decision making process (Edward et al. 2000).
7.5.4.2 Data format 20
25
Data are the building blocks for any research, decision making process or even policy formulation. Hence, it is imperative to have a well-designed or globally accepted data format that allows peers and future generations to do repetitive analysis. Therefore, while choosing data format, special consideration should be given for its compatibility with multiple analytical, reporting and publishing options. The chosen format should be compatible for both spatial and temporal analysis as well as with available software. While archiving data, focus should also be made to have geographic representation within the stipulated grids considering representing global biomes, ecoregions, ecosystems as well as altitudinal and latitudinal representations. There are examples of recognized and widely used data standards such as Darwin core (Wieczorek et al. 2012).
7.5.4.3 Data inter-operability 30
Technical and software standards and protocols are required to ensure the access and usability of data. These should be clear to the user and adopted by as many data management organizations as possible.
7.5.4.4 Open access 35
Open access to data increases the efficiency of research by avoiding unnecessary duplication of data collection and permitting the creation of new data sets by combining data from multiple sources. Coupled with open access, comprehensive documentation of data sets and how to access them provides a more efficient use of resources.
Page 15 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
7.5.4.5 Flexibility In general, scientific communities will approach data management requirements more consistently within their discipline internationally, than they will across other disciplines on a national level. Data access regimes need to be sufficiently flexible to take account of this variation. 5
7.5.4.6 Data publishing platform and sustainability The data publishing infrastructures such as publishing platform and their sustainability needs to be ensured through state of art technological backups and financial sustainability for maintaining and enriching such data hubs periodically when new data are generated. 10
7.5.5 What are the gaps and challenges for developing quality data?
15
20
25
30
Biodiversity distribution across the globe is not consistent (Chapman 2009; Raven and Yeates 2007). Biodiversity rich regions or countries are often poorly equipped with technology, skills, infrastructure and financial resources. As a result, the data collection and management are among the least priority areas leading to limited representation or participation in global database development discourse. The vast information available amongst the traditional and indigenous people and their fading knowledge has not to properly documented and archived. Many of the existing global datasets such as Global Circular Model for temperature and forest datasets such as HYDE (Klein et al. 2011) are with coarse resolution. They do not capture the true picture of varied landscapes such as that of mountains or small sized wetlands and fragmented forests (Chettri et al. 2010; Pfeifer et al. 2014; Svob et al. 2014). Ironically, even the existing datasets maintained by Secretariats of multilateral agreements such as UNFCC, CBD, RAMSAR; global commons on bioinformatics such as GBIF and IUCN Redlist and other datasets maintained by developed countries does not show complementarity to each other and duplication of work is prominent. Geospatial datasets for the same location may use different geospatial projections, making datasets incompatible (e.g., the numerous geospatial projections available for the Antarctic region, and lack of consistency in usage for Antarctic datasets). In addition, taxonomic anomaly, provision for inter-operability among the existing datasets and quest for generating datasets and developing database infrastructure among the conservation communities are bringing more complexities in database management domain rather than contributing to it. There is a serious need for balanced and representative national and regional centers contributing the quality and open access database. Biodiversity by definition is not limited to species or taxonomic data. It is more complex to bring synergy. The compartmentalized data generation and management practices need innovative thinking to bring synergy.
35
7.5.6 What are the short term and long term strategies to address the existing gaps?
40
The existing data collection and management practices could be improved with emphasis on data quality, inter-operability, institutionalization process through short and long term strategies. A number of recommendations can result in increased capacity to use Geographic Information System (GIS)
Page 16 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
databases and analytical and visualization tools for rapid production and access of information products (Table 7.2).
5
Table 7.2 Short- and long-term strategies to address gaps in data collection and management strategies to support BES modelling.
Short term strategies (1-2 years) SWOT analysis Preparation of database development and management outlook at national and regional levels Development of database catalog and identification of gaps Thematic modules for capacity development (training resources- online available) Training for database developers and users Data sharing and users policies Develop guidelines for habitat assessment to allow approximation of ecosystem services based on land-cover/biotypes, and guide data collection to ground truth model functional relationships Develop tools for down-scaling of common databases (e.g. GBIF, climate models) Develop queries to enable dataset transformations of popular global indices to seasonal-monthly-daily scales; spatial queries to enable regional, national or local scale analysis of global datasets
Long term strategies (3-20 years) Establishment/improvement on data base infrastructures (portals) and accessibility Strengthen regional network and cooperation (particularly in nations without culture of data sharing) Establishment of global and regional advisory platforms to certify the quality of the datasets in conformity with the adopted standards Link results with policy development process Ensuring datasets are updated when new information is available Financial sustainability Develop priorities to enlarge data coverage of global datasets
7.6 Improving access to, and guidelines for, user-friendly and practicable software tools for scenario analysis, modelling and decision‐support 10
Key message: Key aspects of the technical capacity required for scenario development and modelling include improving access to, and guidelines for, user-friendly software tools for scenario analysis, modelling and decision‐support systems. 15
7.6.1 Models and software tools Models and the software tools that support model implementation are important components of biodiversity and ecosystem modeling at regional and global scales. Models are direct or indirect methods for approximating biodiversity and ecosystem services. Models are used within scenario Page 17 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
analyses to model relationships between indirect drivers, direct drivers, and biodiversity and ecosystem services resulting in predictions that relate to nature’s benefits to people. Widely used models include climate models such as EMICS, GCMs and RCMs, and biodiversity models such as Predicts, Globio, Madingley. 5
10
Software tools are program created to support model application, and may be free of cost or proprietary. Software used in BES models ranging from standard applications such ArcGIS and other geospatial software, to specialist tools developed specifically to model ecosystem services (e.g., InVEST), to applications for mobile phones, such as those created to support taxonomic identification and geospatial recording of biodiversity records (Table 7.3). Example of open source biodiversity software include Waterworld and Costing Nature (http://www.policysupport.org/costingnature). Costing Nature provides free web training for their user base, and include links to most global datasets in their TERRASIM server; this software also provides the option to upload other databases if there is better data available.
15 Table 7.3 Comparison of widely used software
Name
accessibility
ease of use
INVEST ARIES OCELET WATERWORLD/COSTING NATURE TERRSET
WEB/Easy Web/ difficult Not easy Web/easy
EASY Difficult Acceptable Easy
learning curve FAST Slow Acceptable Fast
Average
Average
Average
VENSIM ECOAIM ECOSERV ENVISION EPM ESVALUE INFOREST LUCI MIMES SOLVES ECOMETRIX
Difficult Difficult Difficult Average Difficult Difficult Average Average Average Average Average
Difficult
Slow Slow Slow Slow Slow Average Slow Easy Easy Average Slow
Type WEBAPPS Standalone Standalone webapps Standalone (or part of IDRISI) Standalone Webapps Webapps Webapps Webapps Webapps Webapps Webapps Webapps Webapps Webapps
7.6.2 Recommended characteristics of software tools to support BES scenarios analysis 20
The most important traits for successful biodiversity and ecosystem services models and software tools are accessibility, user-friendliness and robustness. While many tools are open source, and freely accessible, funds to support access to proprietary software can be attained through development Page 18 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
funding sources such as large international organization such as the United Nations, the World Bank, and CITES.
5
10
15
It is important that stakeholders and policy makers can interpret and believe in model results. As such, modelers and biodiversity and ecosystem specialists should work together to create model that are transparent, minimize complexity, and reflect a common understanding of the model objectives. Creating a network including the scientific community, stakeholders, decision makers, and policy makers can enable feedback at all stages of model development, including evaluating scenario and model outputs with empirical observations. While this process enhances the model through the feedback, it also creates capacity within stakeholders to familiarize with the model and how it works, which in turn enables easier integration and use of the model for planning and decision making processes. All tools (model, software and database) should be well documented, in an intelligible language so that the user base can understand. Metadata should be written following international standards, fully illustrated and idiot-proof. Creating network and user forums are useful for people to ask questions and interact with other users, exchanging knowledge; example include the Marxan forum (University of Queensland) which serves a network of users from over one hundred countries, and answers queries from software applications to dataset requirements related to this software.
7.6.3 Training and support for models and software tools 20
25
30
Training is an important component of software applications. Regular courses are run at global and at regional/national scales, provided training in use and application of different tools. Examples are numerous including Marxan workshops, GIS in Ecology training courses, Society for Conservation Biology training courses aligned with regional and global conferences. Another tactic for enhancing capacities to use tools is the reinforcement and support of regional centers for modeling ecosystem services. Such centers are already present in many places but they either lack funding for training or not well known enough for modelers to draw attention to. Through these places, it is possible to implement a “train the trainer” program that should exponentially enable capacities.
7.7 Developing flexible and effective methods for incorporating local data and knowledge into scenario analysis and modelling Key message: Key aspects of the technical capacity required for scenario development and modelling include developing methods for better incorporation of local data and knowledge
35
7.7.1 Recognition of the interdependence of knowledge systems, including traditional knowledge, to inform BES models and scenarios
40
“Traditional and local knowledge” refers to knowledge and ‘know-how’ accumulated by regional, indigenous or local communities over generations that guide human societies in their interactions with their environment (IPBES/2/17). The IPBES Conceptual Framework clearly recognizes the importance Page 19 of 42
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
and interdependence of the knowledge across multiple systems (local, scientific, technical, educational and traditional) (IPBES/2/17), and that understanding of these complex knowledge systems is necessary to determine system feedbacks within modelling and scenarios. Folke et al. (2002) highlight the importance of knowledge integration within environmental decision-making in order to create resilience (capacity to absorb impacts) in a world of uncertainty and transformations. This resilience requires understanding of traditional knowledge and spirituality, and access to this knowledge base (Tengö et al. 2012). Traditional and local knowledge offers us a vision of the world from a different epistemology, which provides a new perspective to define relationship between people and the environment and to construct "another possible world" (Leff 2011). Cultural understanding of every scenario or stage(?) will bear a commitment (a common vision established in the Plan 2011-2020 in the frame of the CBD and for Aichi Biodiversity Targets), and is a critical aspect of scenario analysis and modelling. For indigenous and local communities, environmental management decisions are intrinsically tied to culture and way of life, and their knowledge can enrich and inform scenarios and models (Feinsinger 2001). However, these systems are often quite complex due to multiple interactions between people and their environment. The main problem of such complex systems is limited skills to understand, to predict and to control socio-ecological systems (Pilkey and Pilkey-Jarvis 2007, Gnaws and Baker 2007). There is a need to include traditional values and knowledge along with the western knowledge, to develop an integrated system in the present times. To develop an integrated system of western and traditional knowledge, decision-making must avoid critical thresholds and incorporate uncertainty and risk in decisions, and these decisions should not be implemented exclusively from an expert analysis, but instead should include consultation and participation of the most relevant users (Cortner & Moote 1999, Bocking 2004, MA 2005a & b). Such integrated efforts could provide enriched information for BES scenarios and models. To develop effective BES scenarios and models for decision-making, diverse forms of local knowledge must come together by transcending spatial and temporal scales (Figure 7.5). Figure 7.5 demonstrates key aspects of integrating local knowledge, including feedbacks between different scales and knowledge systems. The dialogue of knowledge can form the platform for scenarios and modelling. This dialogue should integrate knowledge and cosmo-vision from local and indigenous perspectives, including civil society, scientific experts, private and economic sectors, and the government. In this process, knowledge is achieved through a combination of rights, obligations and responsibilities, resulting in integral, just and sustainable management of resources (Pacheco 2013).
Page 20 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Figure 7.5 Conceptual diagram on integration of local knowledge for developing BES scenarios and models for decision-making (Tëngo et al. 2014).
5
7.7.2 Mechanisms to include traditional knowledge in scenario analysis and modelling Key priorities of IPBES 3/3 (2014) for incorporating traditional knowledge systems into scenario analysis and modelling are efficient mechanisms to integrate knowledge, enhance participation and dialogue between actors at national and regional scales, as summarized below (Box 7.1).
10
15
Box 7.1. Priorities for the models and scenario in relation to capacity building (IPBES 3/3, 2014). Develop the capacity for enhancing collaboration among research institutions and policymakers at national and regional levels, in particular for encouraging multidisciplinary and cross-sectoral approaches Develop the capacity for the conversion of scientific and social assessments of biodiversity and ecosystem services into a format easily understood by policymakers Develop the capacity to understand how to combine modern science with local and indigenous knowledge, including facilitating the effective engagement of indigenous and local communities, scientists and policymakers
20
25
Examples of integration mechanisms include adaptive co-management, participation and ongoing collaboration that can minimize current knowledge gaps and enhance conservation of ecosystems (Folke et al 2002). Adaptive co-management incorporates traditional and modern knowledge, and encourages participation and collaboration amongst all the stakeholders. It is critical to include local/indigenous knowledge from the first stage of planning scenarios in order to allow co-definition of the problem, to increase trust and understanding between participatory stakeholders, and to reduce uncertainty in the scenarios (Peterson et al. 2003). The long-term success of a particular scenario will depend on
Page 21 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
cooperation among various stakeholders through scenario refinement, testing and iterations, to ensure acceptance for evaluating policies and informing decision-making.
5
10
15
20
25
30
35
40
We also need to consider local and national politics that can significantly influence the possible BES scenarios. For scenarios across local scales, we should recognize commonalities between regions that are critical for determining actions and/or for developing policies. Alternatively, there should be specific local scenarios and models to appropriately address the main issues in a region. The temporal scale is another significant aspect that needs to be addressed in order to incorporate temporality into models and decision-making, and to incorporate important drivers in the development of BES scenarios and models at local and regional scales. Incorporation of traditional knowledge is a process that goes hand in hand with empowerment and strengthening of local communities, and is directly related to goal 19 of the Aichi Targets. One method to incorporate traditional knowledge is to develop an integrated set of BES indicators that are based on scientific and traditional knowledge. Many examples exist including the SESELP Program of Biocultural Subantarctica Conservation (Rozi et al. 2010) conducted at a local scale, and National Program of Conservation and Sustainable Utilization (PNCASL) for the caiman (Caiman yacare) in Bolivia, as presented in Box 7.2. Box 7.2. Incorporation of local knowledge in the management and conservation of Caiman yacare (a crocodile species) in Bolivia Bolivia developed a National Program of Conservation and Sustainable Utilization (PNCASL) for the caiman (Caiman yacare). Initially, the annual assignment of local harvest quota was estimated across the "Scientific Authority" based on random counts of relative abundance. With the increasing interest among local communities to be a part in the PNCASL, a need to strengthen the system was identified for incorporating new indicators. These included both biological indicators (based upon models of the species) as well as socioeconomic and cultural indicators of species health. One of the first trials under this new approach was done in the TIPNIS (Indigenous Territory and National Park Isiboro Sécure) where local knowledge was initially the most reliable source on the status of Caiman yacare. The quantitative indicators moved beyond quantitative measures for including qualitative indicators such as perceptions of increase in abundance of charismatic species, e.g. “there are a lot more caiman than before". This helped to develop robust indicators on estimation of population size. This was achieved as the traditional resource users participated in workshops where they informed resource quotas using traditional knowledge of caiman, defining concepts, harmonizing criteria, and defining the space and territory conceptualized across maps and knowledge of habitats where caiman live. From scientific perspective, a population count was measured by the researchers involving indigenous techniques suggested by the communities. The process was repeated by the communities after an integration of knowledge systems and harvest estimates were developed based on local knowledge fortified with scientific concepts and criteria that were internalized (e.g., sizes of hunt allowed). These estimates were used to determine and manage abundance of caiman within the whole protected area. This participatory experience informed a national scale predictive model of abundance of Caiman yacare.
Page 22 of 42
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
Another way to integrate indigenous knowledge for BES scenarios and models is to understand and evaluate the role of BES in people’s well-being. For this, we need to develop and apply a holistic model of well-being by incorporating ES. Sangha et al. (2011) evaluated the role of ES from tropical rainforests in indigenous well-being in north Queensland, applying the MA approach (Figure 7.6). Each ES-wellbeing link highlighted the importance of an ES in terms of well-being of indigenous people that could be used in developing scenarios and models.
Ecosystem services
Human well-being
Provisioning Bush food and medicine Fishing and fish traps Hunting for food and recreation Teaching places Camping ground
Basic material for living Air, water, food and shelter (provision of timber and fibre)
Fire places Timber, fuel wood, bark, tool materials Public recreation Public tracks Regulating and Supporting Biodiversity Soil stability (soil erosion, nutrient levels) Reef protection Hydrological balance Carbon sequestration Cultural Sacred/traditional sites – initiation, burial, remembrance and ceremonial sites Story places Healing places Spiritual sites Identity sites (art or other features) to keep the culture alive Social gathering with family Knowledge transfer to young generation Linkage key: High strength Medium strength Low strength
Good health Provision of good air, water and land resources for good health Security Availability of natural resources for the present and future generations, and opportunity to have recreational/cultural sites for the present and future use. Social relations Cultural celebrations linked to land and other natural features of the landscape, hunting and gathering food, learning techniques and listening stories from elders. Freedom and choice Freedom to access the natural resources Cultural importance Sites of cultural significance, art and artefacts. ABS socio-economic indicators Economic resources Work Education and training Housing Family and community Culture and leisure Crime and justice
Indirect link
10
Figure 7.6 Relationships between ES and the constituents of well-being identified by the Mullunburra-Yidinji community, north Queensland (Sangha et. al. 2011).
Moreover, indigenous knowledge of BES further links with people’s capabilities. For example, knowledge of bush food and medicine of local plants benefits people for providing good health and a capability to Page 23 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
pass on this knowledge to the next generation. According to Sen (1999), enhancing people’s capabilities (e.g. health, education etc.) will enhance their well-being. There is a need to consider the role of BES in terms of indigenous and local capabilities and well-being to communicate information to policy decision making. 5
10
15
20
Based upon literature and personal experience of working with indigenous communities, Sangha et al. (2014) proposed an integrated well-being framework (Fig 7.7) that could be used as a tool to develop possible scenarios and models to suggest and analyze the role of BES in the economic, social and cultural worlds of indigenous and local communities. This tool can highlight each connection that indigenous and local communities may have with their ecosystems, and it can directly be relevant to the policy decision makers as well to the local communities for suggesting feasible scenarios applying current use and value of BES. By incorporating traditional knowledge, the BES scenarios and models can actually strengthen and consolidate traditional knowledge.
Figure 7.7 The social, cultural and economic links between ecosystems and indigenous well-being, a proposed framework (Sangha et al. 2014)
To support integration of traditional knowledge, people’s capacities need to be identified through mapping of key actors, their interests and powers, and feasibility of key actors to participate in scenario analysis and modelling (Table 7.4) (CONDESAN/UMBROL, 2014). This requires knowledge of local information, monitoring and collection of new data to fill information gaps; this information should be quantifiable and verifiable. Often, information gaps in traditional knowledge cannot be fulfilled through conventional scientific evaluations.
Page 24 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
7.7.3 Strategies to communicate information with the local knowledge stakeholders
5
10
15
A main barrier between effective participation and real involvement of the communities and local actors is a lack of provision of information in real time, which in turn influences the positive feedback between communities and decision-makers (Primack, 2001). The spread of knowledge is greatly important for the local actors to take suitable decisions regarding the management as part of the process of empowerment. The knowledge to inform scenarios and models must be freely accessible, and translated into products that are compatible with both local language and modern knowledge systems. A number of communication sources are available such as graphical pamphlets, television and print media, educational systems, and internet and social media. Choice of communication media will depend on the community of interest and their technical capacity. The communication materials must have key messages and presentation format that is of relevance to local communities, avoiding excessive technical information. For example, if one seeks to register data for species or changes in the ecosystems from the local perspective, the graphical material must include the needs of the local actors and its community such as linking data collection with agricultural calendars or cultural events. Highlighting the importance of a particular species in people’s living based upon their current values and usages can also help to engage and communicate with the locals for future scenarios. An important part of information dissemination is that it should be reach to the different sectors including children, women, and aging populations.
20
25
Learning must occur in both directions, with enhanced understanding of local stakeholders in regional, national and international policy and management goals, in addition to the incorporation of local knowledge into local, national and regional collaborative processes that support sustainable development alternatives and biocultural conservation. To prevent the imposition of biocultural standardization (i.e. meaningful or accidental imposition of structures including biotas, languages, cultural models, administrative, educational and economic global structures), those involved in local initiatives must establish collaborative dialogues that include local communities and decision makers (Rozi et al. 2010). Educational initiatives are valuable outlets for enhancing partnerships between the scientific community and local communities through universities and school centers.
30
35
Long-term support for collaborative partnerships are important to give flexibility for the long-term survival of the systems of common property and integration of traditional knowledge into management decisions (Merino and Robson 2006). Global partnerships include organizations such as the Group on Earth Observations Biodiversity Observation Network (GEO BON) which coordinates activities relating to the Societal Benefit Area (SBA) on Biodiversity of the Global Earth Observation System of Systems (GEOSS). GEO BON draws on data-sharing principles to promote full and open exchange of data, and on the GEOSS Common Infrastructure to enable interoperability through adoption of consistent standards.
7.7.4 Integration of traditional knowledge with western science 40
Many of the typical scientific indicators have little relevance to local communities, and limit collaboration in data collection and scenario analysis and modelling. Indicators can vary depending on Page 25 of 42
First Order Draft
20
Table 7.4 Development of capacities and levels of involvement to integrate knowledge for scenario analysis and modelling
Level
Local stakeholders and organizations
15
Integration of traditional knowledge must also incorporate datasets that are often qualitative or metaphysical. For example, Māori, the indigenous people of New Zealand are increasingly expressing values, goals and objectives based on their local knowledge systems within biocultural conservation. Monitoring of locally based indicators is typically used to articulate indigenous values as well as assess, measure and monitor changes to biocultural ecosystems from an indigenous Māori perspective. Robb et al. (2014) found that implementation of locally based indicators into biocultural conservation can be used to build the capacity and capability of Māori communities in both local Māori knowledge and western science. Furthermore, utilsing locally based indicators along with Māori collaboration in biocultural conservation provides an opportunity for cross-cultural learning in scenario development and modelling.
Involved Local stakeholders, national and regional organizations
Public Regional associations Private Institutions
10
DO NOT CITE
local and regional information gaps, advances, uncertainties, conservation programs, priorities and implemented policies, and these indicators can be used in the model validation and scenarios. (BIP, 2011). At a larger scale, the Essential Biodiversity Variables proposed by Pereira et al (2013) are advantageous in allowing for global comparison. At a local scale, indicators that include or have links to local and regional traditional knowledge systems will better contribute to collaborative involvement and to enhance socio-ecological scenarios and models (IUCN, 2006).
Practice
5
IPBES Deliverable 3c
Inter-scientific community Associations that work with local stakeholders
Capacities Representation Leadership Inclusion of biodiversity use/value into policy decision-making Adaptability to ecosystem and functional change Knowledge register Information from people’ systems of life Feedback on indicators of direct drivers Lessons learnt Basics TICs (?) integration Procedures and legal instruments for biodiversity value and conservation Transparency and credibility Measurement of indicators of indirect and direct drivers Interaction with local communities Organizational support for biodiversity and eco system services Transverse incorporation of Biodiversity knowledge in the educational system Generation of TICs in the form of Ecosystem Data Groups Exchange of information among Data Groups at regional level Active Participation Measurement across qualitative and quantitative indicators in relation to the direct drivers Technical integration and multilevel..? Recapitulation of lessons learned on managing and conservation Page 26 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Inter-scientific dialogue Establish and support networks on BES
5
7.8 Developing effective strategies and methods for mainstreaming scenarios and models into participatory assessment and decisionmaking processes across scales (local, regional, global), and across different policy, planning and management contexts. 7.8.1 Strategies to mainstream scenarios into the Science-Policy interface
10
15
20
25
30
35
Scenario development is a way to explore possibilities for the future that cannot be predicted by the extrapolation of past and current trends. The future could be far better or worse than any of the scenarios, depending on choices made by key decision makers and other people in society who bring about change. In this section we see capacity building as a continuous process that should be aimed at strengthening or developing long-term relevant human resources, institutions, and organizational structures (Ash et al 2010) to carry out scenarios exercises and develop models within IPBES assessments. These have to necessarily be of relevance to decision makers with the final goal of them acting on the findings. The purpose of using scenarios and developing stories is to encourage decision makers to consider certain positive and negative implications of different development trajectories (MA, 2005). For example, the failure to meet the 2010 biodiversity targets (CDB, 2010) stimulated a set of new future targets for 2020 (the Aichi targets). At the national scale, most governments recognize the social role of ecosystems and their biodiversity due to their influence on human health and quality of life, apart from their contribution to social and economic development through the supply of essential ecosystem services. It emphasizes the social and economic value of ecosystem services and the importance of their inclusion in policies. However, significant efforts in both science and policy domains need to be made in the next several years if the Aichi targets are to be met (Cardinale 2012). As highlighted by Perrings et al. (2011), the first strategic goal to meet the 2020 targets is to “address underlying causes of biodiversity loss by mainstreaming biodiversity across government and society”. The Millennium Ecosystem Assessment has shown that there is no clear institutional response to address these underlying causes (indirect drivers of change), and new sets of responses are necessary to fill this political gap and to meet the 2020 targets. To meet these targets, further structural changes are required that recognize biodiversity as a global public service as well as integrate biodiversity conservation into policies and decision frameworks for resource production and consumption by focusing on wider institutional and societal changes to enable more effective implementation of policies (Rands et al. 2010). Implementation of policies and development of effective strategies for mainstreaming scenarios and models into decision-making processes across scales (local, regional, global), and across different policy, planning and management contexts within the framework of IPBES relies on a set of capacity building objectives summarized later in this chapter (Table 7.6). Page 27 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
At this stage, there is little information on scenarios that focus on BES and suit the policy decision makers. Costanza et al. 2014 reviewed various scenarios at a global and national scale (Australia) and most of the scenarios were related to businesses or economy, not to BES. Some global scenarios that include some element of ES (except for the MA scenarios (MA 2005)) are mentioned in Table 7.5. 5 Table 7.5 Global scenarios that include ES (Source: Costanza et al. 2014).
GLOBAL SCENARIOS Visions of Alternative (Unpredictable) Futures and Their Use in Policy Analysis (Costanza 2000) Future Vision (Watson and Freeman 2012) World Business Council for Sustainable Development scenarios (World Business Council for Sustainable Development (WB, 2000)) Great Transitions scenarios (Raskin et al. 2002) Millennium Ecosystem Assessment Scenarios (MA (Millennium Ecosystem Assessment) 2005; Cork et al. 2006) UK Foresight Futures (Office for Science and Technology 2002) IPCC (Intergovernmental Panel on Climate Change 2000)
10
15
Technological Optimism vs. Scepticism Real state of the world Collectivism vs. Individualism Pessimism vs. Optimism Market-driven growth, economic globalisation Top-down vs. bottom-up approach to sustainability Alliances, innovation Essential continuity Fundamental but undesirable social change Fundamental and favourable social transformations World development: Globalisation vs. Regionalisation Environmental management: Proactive vs. reactive Social values: Individualistic vs. community-oriented Governance: Interdependent vs. autonomous Relative orientation towards: economic or environmental concerns global and regional focus
Indeed, a mix of approaches such as valuation, scientific (ecological functioning) and modeling may be applicable to develop suitable ‘prototype’ BES scenarios for IPBES assessments. These prototypes can be further developed to provide an integrated information base for the policy decision makers as well as for the public for a specific region/area. There are two different ways in which scenarios and models could be useful for BES assessments: ‘Scenarios based on models’ could be developed to project possible futures where there is a greater degree of certainty in data. For example, population models could be used to develop scenarios on use of ES in a particular region. ‘Models based on scenarios’ could be used to project possible future options. A model can use different scenarios to suggest different options that may occur in the future. For example, a model can suggest the difference in values of ES over time based upon current use of ES.
20 Either of the two methods mentioned above can be applied to project long-term impacts for future decision-making. However, the second approach seems more appropriate in relation to BES assessment
Page 28 of 42
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
given the intangible nature of many ES as well as uncertainty in BES data. The experts, locals and other stakeholders can apply their common judgment to predict for future alternatives.
5
To mainstream scenarios and models into policy sector, it is vital to include people’s wellbeing, economy and status and trends in BES as important domains in any BES scenario, to appropriately inform the policy makers. A preliminary proposal of a ‘BES prototype’ is presented below (Fig. 7.8): 1. Sustainable livelihoods:
3. Reversing the current trend (short-term): Increase in ES as people compromise their utilitarian needs using less resources over a short-term; compromising with the Economy
15
BES
10
Increase in ES and wellbeing of people. It may lead to a decline in GDP, but improvements in (Green) Economy over a long-term Economy
Economy 2. Business as usual: GDP focused economy
25
Figure
Decrease in ES and wellbeing of people, with decline in Economy over a long-term due to exploitation of BES
BES
4. Unsustainable livelihoods: 20
Decrease in ES and social capital (wellbeing of people except for income), but increase in GDP at the expense of BES (exploitation and over-consumption of resources) over long-term
7.8 A prototype of BES scenarios on linking BES with economy (focused on Gross Domestic Product (GDP)) and wellbeing of people (Sangha 2014).
30
35
40
Each type of scenario mentioned in Fig. 7.8 can further include studying the impacts of change in BES over a long-term on: 1. Government (development and policy sector) 2. Natural resources (capital) 3. Social values (capital) These scenarios can further be modeled to project how the future may look like under each scenario. A combined approach of scenario planning and modeling can be useful for the policy decision makers to comprehend various values and changes that may occur in a system over the long-term. However, it is important when working with local or indigenous communities to develop scenarios that suit people’s values. This is one major difference from the modeling approach where pre-developed models are applied without including local values. Scenarios can help us exploring options from local perspectives, and can accommodate local knowledge, thus may prove very useful for IPBES assessments to demonstrate the role of ES in people’s wellbeing, beyond the tangible measures. Moreover, scenarios Page 29 of 42
First Order Draft
5
10
15
IPBES Deliverable 3c
DO NOT CITE
can prove as appropriate tools for many Indigenous and local communities across the globe as these can encompass indigenous/local perspectives of natural systems from a much broader perspective rather than just the biophysical or the economic perspectives that is used in many models. However, combining scenarios with modeling can be an effective tool for the policy decision makers for providing a long-term vision to support decisions. For example, each of the scenarios as mentioned in Fig. 1 could be further processed using MIMES (Multi-scale Integrated Model of ES) or any other such model, to project the outcomes over the long-term in the future. Currently, there is need to develop a range of BES scenarios for IPBES assessments that suit the local perspectives and could be applicable at a local, regional and national scale. For a start, BES prototypes could be developed in advance, that are inclusive and flexible to incorporate values/changes that may occur at any given scale. A final set of BES scenarios could be further developed in consultation with the stakeholders, i.e. local communities, experts, policy decision makers etc. For this, there is a need to focus on building capacity at local, regional and national scales. Within the framework of IPBES, there are a set of capacity building objectives on scenario planning as summarized in Table 7.6.
Page 30 of 42
Deliverable 3c, Chapter 7. Draft version 19-Jan-15 Table 7.6 Capacity building objectives, strategies and actions for developing IPBES scenarios across scales. Entry points
Capacity Building Objective to enhance national networks, individuals and team capacities to carry out and use scenarios exercises within ecosystem assessments
Strategies Establishing or strengthening regional networks of experts Update and complement knowledge and skills in scenarios Improving research capacities of universities and other research and training institutions Implementation of structured training
to enhance institutional expertise, particularly on the science–policy interface, for effective adoption and use of the scenario findings.
Engaging stakeholders Enhance the science policy interface in support of implementation of scenarios Improving the shared knowledge base Improving an understanding of the decision making process on the part of the scientific community Improving capacity for transdisciplinary and transsectorial communication
to strengthen institutional and organizational structures at all levels
Assess, revise and develop scenarios capacities Enhance the capacity to participate effectively in IPBES assessments Develop capacity to locate and mobilize financial and technical resources; Establishing exchange program and technical assistance
Methods Map current expertise/capacities (local and regional) Identify needs Regular training workshops to share methodologies Seminars for specific technical aspects Assistance in conducting scenarios within assessments on the ground. Train new and emerging actors (in an applied setting) Developing curricula relating to ecosystem services and development of scenarios (classroom setting) Involving students and young researchers (fellowships) Establishment of inclusive assessment governance structure (stake holders, scientists, policy makers, local organizations or individuals) Networking and Multiple stakeholders face to face meetings Dialogue and development of visioning exercises with multiple actors (scientists, government officials, policymakers and other stakeholders) Dialogues on methods for scenarios to improve the shared knowledge base (including qualitative and participatory approaches) Training on communication skills Develop plans of actions Establishing institutional partnerships at all scales IPBES matchmaking … bring together specific capacity building need with expert practitioners, guidance and financial resources increase cooperation between centers of excellence/institutions
Page 31 of 42
17/10/2014 The current accelerated changes in economic, social and environmental aspects require flexible policies. Policy is subject not only to a political process but also to urgent or sudden calls for decisions, before any scientific result is available (Scheraga et al. 2003); so there is a growing need of scientific knowledge that is understandable by the different stakeholders. Besides, the complexity of ecosystems and their services demands reliable data and analysis (UNEP 2012) (Figure 7.7).
Figure 7.7 Connection between policy makers and scientific community and scenario capacity needs objectives inter-linkages (red circle indicates Capacity Building objective as referenced in Table 7.6) (Source: Adapted from Swanson, D. and Bhadwal, S. (2009). Creating adaptive policies. IISD-TERI: Ottawa. IUCN (2010). Information Paper on IPBES).
7.8.2 Levels of action where to target identified needs Specific set of instruments to stop the degradation of ecosystems and loss of biodiversity can be classified into the following major groups, each of which exhibits advantages and limitations: I. environmental legislation applicable to the conservation of biodiversity and ecosystems, II. economic incentives and market-based strategies, and III. adaptive governance strategies based on community and education strategies for sustainability. IV. a mix of all the above approaches The first level of action is aimed at developing the main principles of a new paradigm of sustainability to correct management actions of the past that have been proven to be inappropriate and offer new insights into the development of a legal framework that respects the role that ecosystems play in human wellbeing. Ecosystem Assessments serves as a step forward in responding to this policy demand and can be used to reach agreements on BES at national, regional and global levels. For instance, the EU Page 32 of 42
17/10/2014 Biodiversity strategy calls on member States to map and assess the state of ecosystems and their services in the national territory (EU 2011). Maps are useful for spatially explicit prioritization and identification of problems, especially in relation to the synergies and trade-offs among different ecosystem services and between ecosystem services and biodiversity. Furthermore, maps can be used as a communication tool to initiate discussions with stakeholders, visualizing the locations where valuable ecosystem services are produced or enjoyed and explaining the relevance of ecosystem services to the public in their territory. Maps can, and to some extent already contribute to the planning and management of biodiversity protection areas and, implicitly, of their ecosystem services at a subnational level. The second level refers to fundamental market-based instruments for building a framework or governance institutional architecture that is suitable for sustainability. The interactions between human society and ecosystems must be modulated by large-scale rules consistent with values, social attitudes and the role of the economy in a socially fair development model that is ecologically sustainable. It is assumed that ecosystem management is extremely complex and therefore requires institutional complexity. In general, institutions have approached the conservation of biodiversity and ecosystems either indirectly or directly using market strategies and planned conservation and restoration strategies. However, this process is much more complex, and to complete the institutionalization of the conservation of ecosystems and its biodiversity, multilevel coordinated institutional organization will be required. The third level refers to the importance of non-formal institutions, which have been and are currently a key aspect in the conservation of ecosystems and its biodiversity. Maintaining local, social capital generates and contributes, to empowering traditional ecological knowledge and maintaining a workforce of rural people. The political-institutional model in many countries is far from the proposed multi-level governance model by the MA. However, pointing necessary and realistic goals in the sense of strengthening the coordination mechanisms between different levels of government and involvement of different actors in the management and implementation of public policies could be a way forward from the present situation. A multi-level governance model, defined by a number of features related to the relationships between formal and non-formal institutions and the public and private actors, is required. The fundamental objective is to achieve good environmental policy coordination between both global and regional administrations through advisory bodies to obtain good articulation of the different policies for ecosystem management and biodiversity conservation. In any case, it should be noted that these set of tools only make sense when applied together at different spatial and temporal scales, and never when considering each individual tool as an end in itself. Ultimately, greening sectorial policies with a territorial impact are required to build an adaptive governance model based upon the coupling between human and ecological systems. Designing future scenarios encourages reflection on current management strategies and on the conservation of biodiversity and ecosystems as well as the need to move towards strategies designed based on an inclusive (including both formal and non-formal institutions) and polycentric (incorporating Page 33 of 42
17/10/2014 different organizational levels) governance model. The prototype BES scenarios can help to develop a set of scenario proposals for future IPBES assessments. These set of proposals should organize as different management tools with short-, medium- and long-term effects applied at different scales. Finally, the proposal incorporates an interconnected component for communication, education and participation for achieving sustainability, which is considered a cornerstone of the three levels of performance.
7.9 Consolidation, strategy, recommendations Based on the analysis and assessment in the previous chapters, the following broad recommendations are proposed that would improve the use and application of BES projection of data:
7.9.1 Improving and expanding projection of data and capacity building efforts
Improved historical data is needed. Historical data are paramount to the examination and understanding of present variability and the implications of future scenarios. Reliable, quality controlled historical data is often not available. Data rescue activities that facilitate the digitization of archives currently on paper that are in many cases in the hands of a range of stakeholders, including farmers, would improve this. Often these data consist of long time series dating back as much as 40-50 years. Data rescue efforts to convert these to workable/digital formats could be a very valuable contribution to overcoming the current data constraints, improving the quality of BES modeling efforts. More capacity building efforts are needed to train scientists, especially in developing countries, in the development and application of scenarios techniques as well as in the use of multi-model approaches by the international bodies that mentioned in section 7.4 as they have the right and proper resources to building the capacity of BES scenarios and modelling. The ability to draw correct inference from BES data from different sources and to apply these data in decision making could be standardized in the format of a training module. The module could be made available to government officials, decision makers and practitioners as a means of strengthening their capacity to draw appropriately on available data. The training module could also, of course, raise awareness of the available data as it evolves. Standard modules could be developed that would need country specific input to make them relevant to local stakeholders.
7.9.2 Bridging the gap between data producers and data users More opportunities that join decision makers and biodiversity scientists, and provide an environment in which the skills necessary for closer collaboration between these two groups could be created. These interactions should aim to: Improve data users' understanding of the possibilities and limitations of BES projection data and, as such build their capacity in interpreting and applying the data in a correct and appropriate way.
Page 34 of 42
17/10/2014
Improve data producers' understanding about the information needs of different user groups and, as such, enable them to tailor and package their data in ways that are more useful for decision makers. Support translators who understand the challenges on both sides and can act as information conduits. The challenge with this task is that it requires skills that many of the people simply have not developed. Specifically it requires the ability to translate science concepts into those that users understand and can use, without distorting the concepts. It also requires in-depth understanding of users’ needs and the potential opportunities for using BES projection data.
7.9.3 Improve the capacity building regarding the data modelling infrastructure
Interested countries have to strengthen national ecological (and involved) institutions and infrastructure on permanent adaptation to biodiversity modelling and scenario usage practices (through policy, legislation, resources management, statistics, education, communication, others), support respective multidisciplinary research, activities, planning and budgeting. Invite existing MEAs focal points, matched projects/programs/events to involve possible resources for the purposes of modelling. To demonstrate together with communities new cost effectiveness of forecast-based-policy-making primary in developing sectors such as agriculture and biotechnology, protected areas management, forestry, nature conservation, coastal zone management, others. Permanent improvement and adaptation to modelling species occurrence data. Mobilizing existing national and international institutional network as well as respective human resources important for practically sound modelling. Improve digital segments of national reporting system matched biodiversity. To take into account grooving regular/irregular income of new users (Android, iOS, WindowsPhone, others) since they map annually many ecosystems/habitats/pressures/changes. Involve NGOs data sets to GOs net. To cultivate corporative web-based digital products on biodiversity, modelling, scenarios building, accuracy improvement, implementation to practice. Developing actively synergies platforms since a common trend is that growing number of MEAs increasing synchronically to ecologically sound questions on which countries must answer annually through the mechanism of national reporting. Hence, there is a direct benefit from the synergies of the ability to use biological data of MEAs for various goals and objectives, including modelling and scenarios building. CBD Secretariat has named seven Biodiversity-related Conventions (CBD, CITES, CMS, ‘Treaty’, Ramsar, WHC, IPPC) which are important for synergy. ICT-interlinkages inside of BM&SB infrastructure would be a new advantage for worldwide modelling not only for the named seven conventions but also for neighbouring agreements like European Landscape Convention, The Alpine Convention, The Bat Agreement (EUROBATS), econet etc. To incorporate related outcomes of ‘corporative’ and/or ‘civil’ modelling and scenario building to reports of international and national levels (SoE, CBD NR, GBO, EPR) other global assessments (plants, fish, amphibians, mammals, birds). Actively use a unique new ingrowth of ICT-based-civilnetwork, which have demonstrated its effective role in learning-and- studying of the subject, benefit sharing, best practices and mapping. To intensify georeferencing of archived/historical and ‘civil data’ and to lead improving the new e-geosociety. Page 35 of 42
17/10/2014
Development of indicators-and-indices on BM&SB activity for IPBES is increasingly important due to absence of respective statistics. Indirect statistics of WB and OECD as well as previous CBD experience on biodiversity indicators/indices and DPSIR approaches possibly would serve as a starting guideline. IPBES could initiate development of a set of BM&SD indicators-and-indices (activity/quality/veracity/cost-results/etc.) and discuss it with NGOs and GOs.
7.9.4 “Set up a concept” rather than raising the awareness There is a need for developing so-called “set up a concept” – examples of decision makers that have successfully drawn on BES projection data to make decisions that have improved human well-being. “Set up a concept” could also provide an opportunity to quantify and assess the potential value and benefits of using BES projection data as opposed to not using BES projection data at all.
7.9.5 Sustainable donor funding Many donor agencies are currently undergoing a similar process of gap analysis and exploration of opportunities for coordination in order to direct future funding strategies. Since most of these processes are still ongoing it is difficult at this moment to draw final conclusions. Nevertheless some issues seem to be emerging. Firstly, it is important to support processes not projects. Although projects can facilitate the development of technical capacity to interpret and use biodiversity information, experience has shown that building capacity for the appropriate use of biodiversity information is a tedious and resource intensive process that takes time. Therefore it is important that existing networks are strengthened and funding goes beyond project specific skills creation to develop institutional capacity and stability, which requires a lot of inputs from different stakeholders. Even though there is an institutional deficit in developing countries, there are local and international institutions starting to explore the field. Building on these existing institutions and networks has many advantages as opposed to creating new ones. In general, existing institutions can benefit from a longstanding presence in the different regions and they have typically built up credibility and trust among stakeholders. Often they are already grounded in the local reality with an understanding of the local needs and some knowledge about how to deal with institutional constraints that characterize a certain country. Finally, there is an increasing number of private sector donors such Google Foundation and NGOs targeting research and implementation opportunities. To ensure an effective response and avoid overlap, discussions will need to take place between the existing donors and these new players in order to identify roles and responsibilities.
7.9.6 Incorporating traditional knowledge To achieve an effective integration of traditional knowledge and socio-ecological feedbacks in models and scenarios for BES we suggest: Page 36 of 42
17/10/2014 1. To identify universities, research institutions and NGOs with experience and/or are related in the formulations and building of scenarios or models into the scale and the time (long term or short term) 2. To identify the public actors and local representatives who have local networks and/or mechanisms for a distribution of information (indigenous technical personnel, organizations), and local governments in a region so that they are the mechanism of IPBES's connection with the local environment. 3. To identify the coverage of mass media from rural zones up to regional and national scales, according to the goals of monitoring indicator. 4. Coordination and establishment of agreements of cooperation with authorities, technical persons (nets) and organizations for transfer of knowledge as well as for coordination with educational entities for incorporation of information in educational curricular. 5. To select appropriate mechanisms for transfer of knowledge for selecting specific indicators and feedback of the system, and for the models and the scenarios. The generation of knowledge must differ among the standard and/or adapted models and scenarios (in case of models or scenarios applicable to a more local context).
7.9.7 General recommendations There is, firstly, a need to develop scenarios that focus on BES and are appropriate for policy sector, and secondly, to mainstream these scenarios into policy sector. Scenario planning can provide a number of advantages over other techniques such as valuation or modelling particularly when the latter are used in isolation of scenario planning, for simple reasons such as scope to include diversity in ES values from a local/indigenous/scientific perspective, and/or for linking BES with people’s wellbeing. A proposed stepwise approach that may help to mainstream scenarios and models on BES into policies, could involve: 1. Develop trans-disciplinary scenarios and models that include socio-economic and ecological values: For BES scenarios, there is need to consider various social, economic and ecological values of an ecosystem(s) at a given local, regional or national scale, to appropriately inform the future policies. This may incorporate an explicit link to wellbeing of people, demonstrating status of people’s wellbeing under possible scenarios for the future. 2. Develop scenarios and models in consultation with all the stakeholders at a local or regional scale: This includes particularly engaging with the public, government organisations and local communities at a local and regional scale, and this information can be further extrapolated at a national scale. 3. Develop policy relevant scenarios backed by rigorous scientific data and local knowledge: This involves ‘good’ integration of scientific, social, economic and local information, for telling a good storyline. 4. Apply a balanced approach (just enough data to appropriately inform the stakeholders) to develop scenarios: There should be sufficient scientific data that help policy makers comprehend the impacts or changes under a given scenario.
Page 37 of 42
17/10/2014 5. Apply simple language (avoid scientific jargon) to engage the public and the policy decision makers: Simple but realistic presentation of information is necessary to engage the public and the policy decision makers. 6. Incorporate the impact of each scenario on changes in natural, social and built capital for the Government (including economy/policies), Public and Nature.
Page 38 of 42
17/10/2014
References Arzberger, P., Schroeder, P., Beaulieu, A., Bowker, G., Casey, K., Laaksonen, L. et al. (2004). Promoting access to public research data for scientific, economic, and social development. Data Science Journal 3, 135-152. Ash,N.,Blanco,H.,Garcia,K.,Tomich,T.,Vira,B.,Brown,C.,Zurek, M., 2010.Ecosystems and Human Well-Being: A Manual for Assessment Practitioners. Island Press,Washington, DC. Bousquet, F., Barreteau O., Aquino P., Etienne M., Boissau S., Aubert S., Le Page C., Babin D., Castella J. C., 2002. Multi-Agent systems and role games: collective learning processes for ecosystem management. In: Jansen, M. (Ed.), Complexity and Ecosystem management. Edward Edgar Publishers. Butchart, S.H.M., Walpole, M., Collen, B., van Strien, A., Scharlemann, J.P.W., Almond, R.E.A., et al. (2010). Global Biodiversity: Indicators of Recent Declines. Science 328:1164-1168. Cardinale B, Duffy J, Gonzalez A (2012) Biodiversity loss and its impact on humanity. Nature 486: 59–67. Carpenter, S., E. Bennett, and G. Peterson. 2006. Scenarios for ecosystem services: An overview. Ecology and Society 11(1): 29. Castella, J. C., Trung, T. N., Boissau, S., 2005. Participatory simulation of landuse changes in the northern mountains of Vietnam: the combined use of an agent-based model, a role-playing game and a geographic information system. Ecology and Society 10. CBD (2010) Global Biodiversity Outlook 3. Montreal. Chapman A. D. (2009). Numbers of Living Species in Australia and the World. Biodiversity Information Services, Toowoomba, Australia. Pp 80. Chaudhary, S., Kandel, P., Banerjee, S., Uddin, K., Chettri, N. (2014). An integrated assessment of the effects of natural and human disturbances on a wetland ecosystem: A retrospective from Koshi Tappu Wildlife Reserve, Nepal. Kathmandu: ICIMOD. Chettri, N., Shakya, B., Thapa, R. & Sharma, E. (2008). Status of protected area system in the Hindu Kush Himalaya: an analysis of PA coverage. International Journal of Biodiversity Science and Management. 4(3): 164– 178 Cork S. J., Peterson G. D., Bennett E. M., Petschel-Held G. & Zurek M. (2006) Synthesis of the storylines. Ecology and Society 11, [online] Costanza R., Kubiszewski I., Cork S., Atkins P, Bean A, Diamond A, Grigg A, Korb E, Logg-Scarvell J, Navis R, Patrick K(2014). Scenarios for Australia in 2050: A synthesis and proposed survey. Submitted for the Journal of Future Studies (under revision Oct, 2014). Costanza, R. (2000). Visions of alternative (unpredictable) futures and their use in policy analysis. Conservation Ecology 4(1):5. [online] URL: http://www.consecol.org/vol4/iss1/art5 Edwards, J. L., Lane, M. A. & Nielsen, E. S. (2000). Interoperability of biodiversity databases: biodiversity information on every desktop. Science, 289(5488), 2312-2314. EU (European Commission) (2011) Our life insurance, our natural capital: an EU biodiversity strategy to 2020. 244. Brussels GBIF (2013). Global Biodiversity Informatics Outlook: Delivering biodiversity knowledge in the information age. Global Biodiversity Information Facility, Copenhagen, Denmark. GBIF (2014). Data sharing with country of origin/Global data trends. Available at [http://cdn.gbif.org/sites/default/files/gbif_analytics/global/figure/occ_repatriation.png]. GBIO (2012). Contribution of the GBIO to Aichi Targets/Delivering Biodiversity Knowledge in the Information Age//Global Biodiversity Informatics Outlook. Available at: [http://imsgbif.gbif.org/CMS_ORC/?doc_id=5353&download=1]
Page 39 of 42
17/10/2014 Gilani, H., Shrestha, H.L., Murthy, M.S.R., Phuntso, P., Pradhan, S., Bajracharya, B. et al. 2014). Decadal land cover change dynamics in Bhutan. Journal of Environmental Management, http://dx.doi.org/10.1016/j.jenvman.2014.02.014 Guralnick, R. P. & Neufeld, D. (2005). Challenges Building Online GIS Services to Support Global Biodiversity Mapping and Analysis: Lessons from the Mountain and Plains Database and Informatics project. Biodiversity Informatics. 2, 56-69. Gurung, D.R., Amarnath, G., Khun, S.A., Shrestha, B., Kulkarni, A.V. (eds) (2011) Snow-cover mapping and monitoring in the Hindu Kush-Himalayas. Kathmandu: ICIMOD Horigue, V., Aliño, P.M., White, A.T., Pressey, R.L. (2012) Marine protected area networks in the Philippines: Trends and challenges for establishment and governance. Ocean & Coastal Management, 64(0): 15-26. http://dx.doi.org/10.1016/j.ocecoaman.2012.04.012 IPCC (2007). IPCC Summary for Policymakers: Climate Change 2007: Climate Change Impacts, Adaptation and Vulnerability. IPCC WGII: Fourth Assessment Report. IPCC (2014). IPCC Fifth Assessment Report-Summary for Policymakers. Intergovernmental Pannel on Climate Change, Geneva, Switzerland IUCN (2010). Information Paper on IPBES Klein G.K., Beusen, A., Van Drecht, G., & De Vos, M. (2011). The HYDE 3.1 spatially explicit database of human‐induced global land‐use change over the past 12,000 years. Global Ecology and Biogeography, 20(1), 73-86. Kowalski, K., Stagl S., Madlener R., Omann I. 2013. Sustainable energy futures: methodological challenges in combining scenarios and participatory multi-criteria analysis. European Journal of Operational Research. Leadley, P. (2010). Biodiversity Scenarios: Projections of 21st Century Change in Biodiversity, and Associated Ecosystem Services: a Technical Report for the Global Biodiversity Outlook 3 (No. 50). UNEP/Earthprint. Lehner, B., & Döll, P. (2004). Development and validation of a global database of lakes, reservoirs and wetlands. Journal of Hydrology, 296(1), 1-22. MA (Millennium Ecosystem Assessment) (2005). Ecosystems and Human Well-being: Synthesis. Washington, DC. Island Press, Washington, DC. 155 pp MA (Millennium Ecosystem Assessment). 2005a. Ecosystems and human well-being: Scenarios. Washington DC: Island Press. MA (Millennium Ecosystem Assessment). 2005b. Ecosystems and human well-being: Multiscale assessments. Washington DC: Island Press. Mace, G.M., Cramer, W., Dıaz, S., Faith, D. P., Larigauderie, A., Prestre, P. L., et al. (2010). Biodiversity targets after 2010. Current Opinion in Environmental Sustainability 2:1–6 Madlener, R., Kowalski K., Stagl S. 2007. New ways for the integrated appraisal of national energy scenarios: the case of renewable energy use in Austria. Energy Policy 35, 6060 – 6074. Maes J., Hauck J., Paracchini M. L., Ratamäki O., Termansen M., Perez-Soba M., Kopperoinen L., Rankinen K., Jan Philipp Schägner, Peter Henrys, Iwona Cisowska, Marianne Zandersen, Kurt Jax, Alessandra La Notte, Niko Leikola, Eija Pouta, Simon Smart, Berit Hasler, Tuija Lankia, Hans Estrup Andersen, Carlo Lavalle, Tommer Vermaas, Mohammed Hussen Alemu, Paul Scholefield, Filipe Batista, Richard Pywell, Mike Hutchins, Morten Blemmer, Anders Fonnesbech-Wulff, Adam J. Vanbergen, Bernd Münier, Claudia Baranzelli, David Roy, Vincent Thieu, Grazia Zulian, Mikko Kuussaari, Hans Thodsen, Eeva-Liisa Alanen, Benis Egoh, Peter Borgen Sørensen, Leon Braat, Giovanni Bidoglio 2012. A spatial assessment of ecosystem services in Europe: methods, case studies and policy analysis - phase 2. Synthesis report. PEER Report No 4. Ispra: Partnership for European Environmental Research Magnuszewski, P., Sendzimir J., Kronenberg K. 2005. Conceptual modelling for adaptive environmental assessment and management in the Barylz valley, lower Silesia, Poland. International Journal of Environmental Research and Public Health 2, 194 – 203. Page 40 of 42
17/10/2014 Mills, M., Weeks, R., Pressey, R.L., Gleason, M.G., Eisma-Osorio, R.-L., Lombard, A.T., Harris, J.M., Killmer, A.B., White, A., Morrison, T.H. (2015) Real-world progress in overcoming the challenges of adaptive spatial planning in marine protected areas. Biological Conservation, 181(0): 54-63. http://dx.doi.org/10.1016/j.biocon.2014.10.028 Morrison, T.H., McAlpine, C., Rhodes, J.R., Peterson, A., Schmidt, P. (2010) Back to the Future? Planning for environmental outcomes and the new Caring for our Country program. Australian Geographer, 41(4): 521-538. 10.1080/00049182.2010.519763 Nativi, S., Mazzetti, P., Saarenmaa, H., Kerr, J., & Tuama, É. Ó. (2009). Biodiversity and climate change use scenarios framework for the GEOSS interoperability pilot process. Ecological Informatics, 4(1), 23-33. Office for Science and Technology. (2002) Foresight Futures 2020. Department of Trade and Industry, London. Parr, C. S., Wilson, M. N., Leary, M. P., Schulz, K. S., Lans, M. K., Walley, et al. (2014). The Encyclopedia of Life v2: Providing Global Access to Knowledge About Life on Earth. Biodivers Data J. 2), e1079. Pereira, H. M., Leadley, P. W., Proença, V., Alkemade, R., Scharlemann, J. P., Fernandez-Manjarrés, et al. (2010). Scenarios for global biodiversity in the 21st century. Science, 330(6010), 1496-1501. Perrings C, Duraiappah A, Larigauderie A, Mooney H (2011). The Biodiversity and Ecosystem Services Science-Policy Interface. Science 331: 17–19. Pfeifer, M., Lefebvre, V., Gardner, T. A., Arroyo‐Rodriguez, V., Baeten, L., Banks‐Leite, C. et al. (2014). BIOFRAG–a new database for analyzing BIOdiversity responses to forest FRAGmentation. Ecology and Evolution, 4(9), 1524-1537. Pimm, S. L., Jenkins, C. N., Abell, R., Brooks, T. M., Gittleman, J. L., Joppa, L. N., Raven, P. H. et al. (2014). The biodiversity of species and their rates of extinction, distribution, and protection. Science 344 (6187), 1246752 Rands MRW, Adams WM, Bennun L, Butchart SHM, Clements A, et al. (2010) Biodiversity Conservation: Challenges Beyond 2010. Science 329: 1298–1303. doi:10.1126/science.1189138. Raskin P., Banuri T., Gallopin G., Gutman P., Hammond A., Kates R. & Swart R. (2002) Great Transition: The Promise and Lure of the Times Ahead. Stockholm Environment Institute, Boston. Raven, P. H. & Yeates, D. K. (2007) Australian biodiversity: threats for the present, opportunities for the future. Aust. J. Entomol. 46, 177–187. Reed M. S., Kenter J., Bonn A., Brad A., Burt T. P., Fazey I. R., Fraser E. D. G., Hubacek K., Nainggolan D., Quinn C. H., Stringer L. C., Ravera F. 2013. Participatory scenario development for environmental management: A methodological framework illustrated with experience from the UK uplands. Renn, O. 2006. Participatory processes for designing environmental policies. Land Use Policy 23, 34 – 43. Robb, M., Harmsworth, G., Awatere, S. 2014. Māori Values and Perspectives to Inform Collaborative Processes and Planning for Freshwater Management. Technical report. Landcare Research, Hamilton, New Zealand. Robertson, T., Döring, M., Guralnick, R., Bloom, D., Wieczorek, J., Braak, K. et al. (2014). The GBIF integrated publishing toolkit: facilitating the efficient publishing of biodiversity data on the internet. PloS one, 9(8), e102623. Sala, O.E., Chapin, F. S., Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R. et al. (2000). Global biodiversity scenarios for the year 2100. Science 287(5459), 770-1774. Sangha, K., Butler, J., Delisle, A., Stanley, O., (2011). Identifying links between ecosystem services and Aboriginal well-being and livelihoods in north Australia: applying the Millennium Ecosystem Assessment framework. Journal of Environmental Science and Engineering, 381-387. Sangha, K., LeBrocque, A., Costanza, R., Cadet-James, Y. (2014). Ecosystems and Indigenous well-being: An integrated framework. Submitted to the journal Ecosystem Services (Dec 2014). Scheraga, J. D., Ebi, K. L., Furlow, J., & Moreno, A. R. (2003). From science to policy: developing responses to climate change. Climate Change and Human Health: Risks and Responses (McMichael AJ, Campbell-Lendrum D, Page 41 of 42
17/10/2014 Corvalan CF, Ebi KL, Githeko A, Scheraga JD, et al., eds). Copenhagen: World Health Organization/World Meteorological Organization/United Nations Environment Programme, 237, 266. Secretariat of the Convention on Biological Diversity (2010) Global Biodiversity Outlook 3. Montréal, 94 pages. Secretariat of the Convention on Biological Diversity (2014) Global Biodiversity Outlook 4. Montréal, 155 pages. Sen, A., (1999). Commodities and Capabilities. Oxford University Press. Svob, S., Arroyo-Mora, J. P., & Kalacska, M. (2014). The development of a forestry geodatabase for natural forest management plans in Costa Rica. Forest Ecology and Management, 327, 240-250. Swanson, D y Bhadwal, S. (2009). Creating Adaptive Polices. Ottawa: IISD. Tompkins, E. L., Few R., Brown K. 2008. Scenario-based stakeholder engagements: incorporating stakeholders preferences into coastal planning for climate change. Journal of Environmental Management. 88, 1580 – 1592. UNDP (2011). Practitioner’s Guide: Capacity Development for Environmental Sustainability UNEP (2012). GEO 5. Nairobi: UNEP van den Belt M. 2004. Mediated Modelling: a System Dynamics Approach to Environmental Consensus Building. Island Press Washington D. C. Viciani, D., Lastrucci, L., Dell’Olmo, L., Ferretti, G., & Foggi, B. (2014). Natura 2000 habitats in Tuscany (central Italy): synthesis of main conservation features based on a comprehensive database. Biodiversity and Conservation, 23(6), 1551-1576. Watson R. & Freeman O. (2012) Future Vision. Scenarios for the World in 2040. Scribe Publications, Brunswick, Victoria. WB (World Business Council for Sustainable Development) (2000) The Wizard of Us. World Business Council for Sustainable Development. WB (World Business Council for Sustainable Development) (2012). Internet users, total number/World Development Indicators. Available at: [http://data.worldbank.org/indicator/IT.NET.USER]. Last accessed 24 May 2012. Wieczorek, J., Bloom, D., Guralnick, R., Blum, S., Döring, M., Giovanni, R., et al. (2012). Darwin Core: An evolving community-developed biodiversity data standard. PLoS One, 7(1), e29715. Yahara, T., Ma, K., Darnaedi, D., Miyashita, T., Takenaka, A., Tachida, H., et al. (2014). Developing a Regional Network of Biodiversity Observation in the Asia-Pacific Region: Achievements and Challenges of AP BON. In Integrative Observations and Assessments (pp. 3-28). Springer Japan.
Page 42 of 42
First Order Draft
8
5
IPBES Deliverable 3c
DO NOT CITE
Improving the rigor and usefulness of scenarios and models through ongoing evaluation and refinement Coordinating Lead Authors: Resit Akçakaya, Henrique M. Pereira Lead Authors: Graciela Canzani, Cheikh Mbow, Akira Mori, Maria Gabriela Palomo, Jorge Soberon, Wilfried Thuiller, Shigeo Yachi Contributing Authors: Miguel Fernandez, Niklaus E. Zimmerman KEY MESSAGES
10
15
20
25
30
35
40
We recommend that IPBES engages with existing processes on identifying essential variables for monitoring, modelling and reporting of biodiversity and ecosystem services. A concerted effort is required for improving the spatial, temporal and taxonomic coverage and resolution of the monitoring of those variables. This effort requires cost-effective approaches that are geared towards the needs of scenario analysis and modelling at global, regional and local scales. More efforts are required towards easier access to well-documented data and models. IPBES should adopt existing data and model documentation standards and expand those as needed, make use of existing central repositories and liaise with organizations to develop new ones, and participate on on-going efforts to assure proper credit to data and model providers. The task force on Data and Knowledge could have a key role in these efforts. We recommend that IPBES develops mechanisms―such as a task force on research prioritisation―to encourage basic research that advances scenario analysis and modelling, including research on (i) linkages between different aspects of biodiversity and ecosystem services; (ii) ecological processes at temporal and spatial scales relevant to IPBES assessments; (iii) early warning systems to anticipate ecological breakpoints and regime shifts; and (iv) coupling of, and feedbacks between, social and ecological components of ecosystems. We recommend that IPBES develop guidelines and standards for verification and validation of models, and uncertainty propagation in scenario analysis and modelling; and ensure continued evolution of these standards through regular updates based on scientific developments. We recommend that IPBES develop mechanisms―such as a task force that regularly updates Deliverable 3c to make it an evolving guide―to ensure that the review of available policy support tools and methodologies for scenario analysis and modelling of biodiversity and ecosystem services continues to reflect best available science. IPBES can pave the way to instil the use of models and scenarios to improve the transparency of policy-making, by making the assumptions explicit and facilitating the comparison of multiple options.
Page 1 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
IPBES assessments should identify all the stakeholders relevant at the scale of the problem, including scientists, decision-makers and people with different types of knowledge, and engage them early in the modelling and scenario analysis process. 5
10
15
20
25
We recommend IPBES supports activities to give planners and policy makers a better understanding of models and scenarios, including limitations and uncertainties, and that IPBES should develop activities to assist modellers in engaging further with policy and planning processes. Previous chapters have demonstrated the variety of approaches to scenario analysis and modelling that can be used to inform decisions and evaluate policy options. They have also identified the problems or challenges, and reviewed existing solutions. The goal of this chapter is to chart the way forward in terms of the additional research and development that is required to develop more effective solutions and to shift scenario analysis and modelling of biodiversity and ecosystem services to a whole new level of rigour and utility. The chapter is organised into three main sections. We first discuss the highest priority approaches to improving the data used to calibrate and validate models, emphasizing linkages to various existing initiatives for biodiversity monitoring at national, regional and global scales. We then discuss basic and applied science research needed to improve models, both by promoting development of new models and by encouraging and facilitating functional linkages among existing models and modelling platforms. Finally, we discuss institutional changes, educational developments and improvements in communication needed to make models more useful and relevant to decision making at local, national and global levels.
8.1 Improving data 8.1.1 Identification of common metrics
30
35
40
Biodiversity has multiple dimensions, including genetic diversity, species diversity and ecosystem diversity, and can be measured in a multitude of ways (Noss, 1990; Pereira et al. 2012). Similarly, there are many ecosystem services and each ecosystem service can be quantified using different approaches, including biophysical, cultural and economic measurements (Daily et al. 2009). Researchers often face the challenge of accessing adequate data for calibration and validation of models as different initiatives monitor different biodiversity metrics and there is a lack of harmonization and integration across observation communities and countries (Pereira et al. 2013). On the other hand, the needs of the modelling community have not been clearly articulated to the observation community. A key challenge is the identification of common metrics that could be used by the modelling and observation communities. A common set of variables and parameters 1 for observation and
Variables are dynamic and are monitored and modelled over time, while parameters are static and are observed once and set up once in the models. Sometimes the separation is hard to make, as a variable in a model can be a parameter in a different model, particularly if they look at different time scales.
1
Page 2 of 33
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
modelling, of biodiversity and ecosystem services would allow easier integration of data from different sources, and would facilitates calibration, validation of models and inter-model comparison. Currently two approaches, at different levels of data abstraction (Figure 8.1), show promise: the Essential Biodiversity Variables being promoted by the Group on Earth Observations Biodiversity Observation Network (Pereira et al. 2013), and the biodiversity indicators being adopted by the Convention on Biological Diversity to assess progress towards the Aichi targets (Tittensor et al. 2014; Nicholson et at. 2012).
Figure 8.1. The Essential Biodiversity Variables framework (reprinted from Pereira et al. 2013).
In recent years, scientific communities of different physical and biological phenomena have started to identify essential variables that are critical for monitoring and modelling. The first such effort was the identification of the Climate Essential Variables by the Global Climate Observing System. Similarly, the Group on Earth Observations Biodiversity Observation Network (GEO BON) has developed a process to identify Essential Biodiversity Variables. The idea behind this concept is to identify, using a systems approach, the key variables that we need to monitor in order to understand biodiversity change. The Essential Biodiversity Variables are an intermediate layer of abstraction between the raw data, from in situ and remote sensing observations, and the derived high-level indicators used to communicate the state and trends of biodiversity (Figure 8.1). These variables can be used in models of the whole biosphere or parts of it as the main system variables. They can then be used to compare model simulations with data. For example, the population abundance variable is defined as a three dimensional matrix of population abundances per species, per location, per time. A gridded dataset of population abundances for a group of species requires the integration of population estimates from different methods and observers, and the interpolation of gap areas with modelled and remote sensing data. Such gridded dataset can be used to calibrate and validate species distribution models, or models of the impacts of drivers on biodiversity such as GLOBIO (Alkemade et al. 2009).
Page 3 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
A list of 23 Essential Biodiversity Variables candidates has been identified, organized into six major classes (Pereira et al. 2013): genetic composition, species populations, species traits, community composition, ecosystem structure, and ecosystem function. An effort is on-going to identify appropriate monitoring schemes, propose data standards and develop global or regional datasets for each variable. A similar process to identify essential variables for Ecosystem Services is under discussion in GEO BON. It is important for IPBES to engage in these processes in order to guarantee that the final set of variables and associate monitoring methods and data standards serves the user needs of the assessments. We envision a particular role for the Task Force on Data and Knowledge on this process. An alternative approach is to work at a higher level of data abstraction by using biodiversity indicators and indices (van Strien et al. 2012). Over the last decade several biodiversity indicators have been used to report on biodiversity change at the national and global level (Titensor et al. 2014). Indicators condense a wealth of data into a few values. For instance the Living Planet Index condenses information on population counts of several thousands of vertebrate populations into a single global value per year, which inform on global vertebrate population reductions relative to a base year. The Red List Index condenses assessments of endangered species status of a few thousands species into a single value for a time point which can be compared with values from previous time points to assess if there has been an acceleration or deceleration of biodiversity loss. Similarly, a few indicators have been developed for supply, demand and benefits of ecosystem services, including fish catch, timber production, and annual carbon sequestration (Tallis et al. 2012; MAES 2014). Recently, some models have been developed to project the evolution of these indicators globally or regionally (Nicholson et al. 2012; Visconti et al. in press). The power of this approach is to provide large-scale calibration and validation of scenarios models. However, it seems that not all indicators are equally amenable to simulation (Visconti et al. in press). IPBES assessments should report results of models and scenarios using a small set of indicators. For example, models assessing the impacts on species populations could report geometric or arithmetic mean abundance changes or aggregated changes in extinction risk. Models assessing the impacts on ecosystem services should use a group of indicators for a standard classification of ecosystem services such as the Common International Classification of Ecosystem Services or CICES (MAES 2014). The set of indicators to adopt should be further explored by a Task Force on Modelling and Scenarios, which would also update regularly the guidelines presented in the current report.
8.1.2 Increasing data availability for model calibration and validation Despite recent increases in the variety and amount of biodiversity-related data, there are significant gaps with respect to quantity and quality, and significant biases in the availability of biodiversity and ecosystem services data: • Regional biases in coverage: Historically, ecologists studied pristine but relatively humanaccessible areas, in wealthy countries, resulting in a very uneven global distribution of study areas (Figure 8.2). The disparity among terrestrial, freshwater and marine realms is also noteworthy (Loh et al 2005). • Taxonomic biases in coverage: Ecological studies have focused disproportionately on conspicuous species. Vertebrates, particularly birds and mammals, are much more often the
Page 4 of 33
First Order Draft
• 5
10
15
•
IPBES Deliverable 3c
DO NOT CITE
focus of ecological studies than invertebrates and plants (Pereira et al 2012). One of the most popular indices to measure global biodiversity change, the Living Planet Index (LPI), is based on vertebrate populations only (Loh et al. 2005). Spatial and temporal resolution: Most ecological studies either have high spatial resolution and small spatial extent, focusing in detail on small areas, or have low spatial resolution and focus on larger regions. For some scenario analysis and modelling approaches, high resolution data with global coverage are needed (Pereira et al. 2010). Such data exist for some biodiversity-related variables (such as forest cover data available at http://earthenginepartners.appspot.com/), but this is rare. Thematic gaps: There is a lack of regional and global consensus on what to monitor. Monitored variables are often selected with the goal of determining the current status of biodiversity, and may not always fill the needs for models designed to projecting the future status. In terms of biodiversity at the species level, data on species occurrence locations, geographical ranges, time series of abundance indices, and threat levels are available for more species than demographic variables (such as fecundity) and life history variables (such as generation length.)
IPBES will need to use a range of approaches to fill these gaps including: improving monitoring programs, mobilizing data, and modelling. 20
25
Figure 8.2: Maps of (a) the global distribution of ecological field sites (kernel densities), (b) study site position (crosses) overlaid on the distribution of potential vegetation biomes (Ramankutty and Foley 1999), and (c) study site position (crosses) overlaid on the distribution of anthromes (Ellis et al. 2010). All maps are expressed in Eckert IV Equal Area projection. Reprinted from Martin et al. 2012 (need to obtain permission).
Page 5 of 33
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
In many cases, existing data bases can be improved with concerted and coordinated efforts to increase spatial (regional) coverage, spatial resolution (e.g., smaller grid size or denser sampling points), temporal resolution (regular and frequent observations), and temporal coverage (long-term, sustainable monitoring for the future; historical reconstruction for the past). Monitoring programs should have a data strategy that leads to making optimal choices about what and how to measure; and be cost-efficient, sustainable through space and time, and effective, avoiding duplication (see BOX 8.1). For instance, in terms of taxonomic coverage, adding large numbers of species in poorly studied taxonomic groups may not be cost effective. However, a taxonomically sampled approach, as used in the Sampled Red List Index (Baillie et al. 2008), can provide taxonomic coverage in a costeffective way. Although much of the biodiversity data available today is suitable for scenario analysis and modelling, future assessments by IPBES would benefit if monitoring programs are designed with modelling in mind, and in collaboration with modellers. BOX 8.1: Data strategy (modified from Scholes et al. 2012 and other sources)
15
20
25
1. Relevant to the goals and data needs of scenario analysis and modelling at global, regional and local scales. 2. Global in coverage, but with sufficient resolution and accuracy at subnational scales to be useful to the main decision-makers at this scale. 3. Statistically sound basis for repeated measurements of biodiversity. 4. Following best practices for metadata specification. 5. Provisions for coordinating and managing data that are collected by disparate institutions and individuals for different purposes. 6. Sufficiently comprehensive in terms of taxonomic coverage. 7. Quality controlled, with well-defined standards for formats, codes, measurement units and metadata; traceability of the observation (incl. place and time of origin, the techniques used to make the observation, and methods used to modify the data); enforced data-typing 8. Cost efficient. Avoiding duplicative work in recording or analyzing the same observations for the same time period. 9. Sustained. Ensuring data continuity and comparability over time, including provisions for long-term storage and data management. 10. Adaptive. Responsive to new technical possibilities and emerging societal needs. 11. Interoperable. Data available to (and discoverable by) other parts of the system, with tools to enable analyze data from different parts together. Requires metadata (see above) and New and promising approaches to building and curating datasets include citizen science and crowdsourcing (Silvertown 2009, Wiggins and Crowston 2011), and new technological tools such as automated data collectors and sensor networks that are embedded in the environment (Collins, et al. 2006; Porter et al. 2009; Rundel et al. 2009; Benson et al. 2010). The new field of ecoinformatics envisions building ecological data sets in the context of a "data life cycle" that encompasses all facets of data generation to knowledge creation, including planning, collection and organization of data, quality assurance and quality control, metadata creation, preservation, discovery, integration, and analysis and visualization (Michener and Jones 2012). Ecoinformatics tools that support and assist various steps of the data life cycle include data management planning tools (e.g., http://dmp.cdlib.org/); metadata standards and tools; relational data bases which allow specification of constraints on the types of data that can be entered (i.e., data typing) assuring data
Page 6 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
integrity; scientific workflow systems, such as Kepler, Taverna, VisTrails and Pegasus (see section 8.2.1.2); and cloud-computing resources.
5
10
15
20
25
30
35
40
In some cases, gaps in datasets can be filled using quantitative approaches such as statistical and modelling methods. One approach is imputation, which is a particularly useful approach when analysing large data sets of demographic traits (e.g., Di Marco et al. 2012). Imputation is a valuable alternative to removing missing observations in databases, because it produces low errors while retaining statistical relationships among the variables (Penone et al. 2014). Another option for filling data gaps is making inferences based on allometric relationships between biological variables such as body size, metabolic rates, population density, generation time and maximum population growth rate (e.g., Damuth 1987). Although allometric relationships have rarely been used in a practical way to estimate model parameters because of the large uncertainty in the predicted values, they may be useful if limited to groups of functionally related species (such as herbivorous mammals). A third approach involves sampling demographic parameters of population models using a "generic life history modelling" approach. Although linking ecological niche and population models gives more realistic predictions of the effects of changing environmental conditions on species (Keith et al. 2008), widespread application of such coupled niche-population models is hampered by the availability of species-specific demographic data. The generic life history modelling (Pearson et al. 2014; Stanton et al. 2014) gets around this problem by using ensembles of population models designed to encompass the full set of life history parameters characteristic of a particular group of species. This approach avoids the need to obtain species-specific demographic parameters, which are rarely known, and enables generalising results beyond the well-studied species, at the cost of not being able to make species-specific predictions of population dynamic s (Pearson et al. 2014). Remote sensing and in situ data are a vital for modelling and monitoring environmental parameters relevant for biodiversity conservation (Buchanan et al. 2009, Koghan et al. 2010). Satellite remote sensing is useful to collect data across different spatial and temporal scales. However, there is a lack of capabilities of users to deal with these data; access to training and education in using satellitebased observations are essential in the future (Turner et al. 2015). Some initiatives for increasing access to remote sensing data globally are the international Group on Earth Observations (GEO) and ESA Climate Change Initiative (Bontemps et al.,2011). There is a need to improve the existing statistical and modelling approaches discussed above, as well as developing new quantitative approaches for filling data gaps. In recent years, ecosystem services modelling has improved with governmental demand for standardized practices to measure, value and map ecosystem services (Waage & Kester 2014). The best known, generalisable model is InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) that combines data on the economy, human well-being, and the environment, in an integrated way for 16 services and models the impacts of alternative resource management choices (Daily et. al 2009). However, there is a need for understanding the relationships between ecological mechanisms and ecosystem services to create the realistic end products for managers (Wong et al. 2015). Approaches to include institutional realities (e.g. regulations, policies) into ecosystem service values should be improved to satisfy the needs of scenario analysis and modelling at global, regional and local scales.
45
Page 7 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
8.1.3 Facilitating data access for model calibration and validation Data availability is drastically increasing, especially for biodiversity information (Pimm et al. 2014); however major barriers remain associated with the limited usability of and accessibility to data. 5
10
15
20
25
30
35
40
45
8.1.3.2 Data sharing across stakeholders There is currently a major movement to “open data”, reflecting an increasing interest and demand for data being made publicly available (Molloy 2011; Reichman et al. 2011). CBD Aichi Target 19 also emphasizes that biodiversity information need to be “widely shared and transferred, and applied.” In coming years, data release is expected to be more often required by funding sources and by research journals, and it will become a common norm of conduct of scientific societies. Note that this is not just a response to increasing calls for transparency from stakeholders; archiving data in public domains can potentially yield multiple benefits to the scientific community and the data providers. The opening-up of data does not only help reduce the duplication of work needed for data collection but also facilitates scientific exploration (Ruegg et al. 2014). Archiving more data can possibly reduce uncertainty in the urgent situations in environmental management. For example, it is claimed that for many taxa sufficient information does not exist when they become threatened, so that monitoring programs implemented by local agents are often too late and ineffective (Lindenmayer et al. 2013). Considering the fact that combining past inventory data with present data can serve as a surrogate of long-term monitoring (e.g., estimating a temporal change in species distribution in response to climate change (Moritz et al. 2008)), archiving data is crucial for better coping with complexity and uncertainty inherent in ecological systems. By enriching data in public domains, there is a high possibility of faster scientific discovery as well as savings in organizations and governments, both of which would help develop and apply models for biodiversity and ecosystem services management. Note that data are undoubtedly precious, so that some stakeholders may feel reluctant to provide their data to public domains. Field data, which are the crucial part for the majority of models, need enormous effort to be collected. Creating large datasets spanning several temporal, geographical and biological scales, which are essential for global assessments, require numerous inputs from a large number of contributors. The trend to make data open could make people (especially early career researchers) feel averse collecting data on their own, which might have negative consequences in the long-term. Therefore, incentives, including career rewards, are of importance to ensure further development of data archives (Borgman 2012; Costello et al. 2013). While potential benefits of open data have been extensively discussed recently, no clear emphasis has been place on crediting and rewarding aspects of providing data. Advocates for opening up data tend to stand on the side of the “data user”, and do not necessarily view the issue from the side of the “data collector”. According to a survey, the most dominant answer from data collectors as a condition for the use of data is formal citation (Michener et al. 2012). However, given a strict, short page limit with a limited number of citations allowed in many journals, it is often difficult to “formally” cite all data sources, especially for papers that rely on spatially and temporally large multiple datasets. Metadata could be more acceptable for those who are willing to collect and provide data. However, conflicts exist as raw data are often required by other stakeholders. Archiving data as a metadata, which require users for multiple procedures to access raw data, could be one of the causes of delayed delivery of information between different scientific disciplines. Given the “top-down pressure”(Molloy 2011) for open data, development of additional incentives and initiatives will be necessary for shortening the time between the development of models and
Page 8 of 33
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
the delivery of the resulting outcomes to policy development. In this regard, inviting data collectors to be involved in data analysis can potentially help the study of focus, because data collectors have first-hand knowledge about the strengths and weaknesses of the data. This co-development and collaboration between data collectors and users may benefit both, leading to “win-win solutions”. Although this is one possible way to overcome the issue and would not provide an ultimate solution, data collectors should be more encouraged and acknowledged.
8.1.3.2 Accessing and using data by different stakeholders 10
15
20
25
Both biodiversity and ecosystem services data, which are increasingly available to public (e.g., Boxes 8.2 and 8.3), have been evaluated with different spatial/temporal coverage and resolution, and metrics themselves are very diverse. In using such data, there are several issues to be considered. An important issue is data standardization. Models and scenarios often require multiple data types, sourced from different databases. Combining data from multiple sources may be difficult; for example, biodiversity information such as taxonomic names are often stored in different ways or following different published taxonomies. Although a number of tools are available to unify data from different sources, such as uBio (http://www.ubio.org/), which can help match and integrate names of species from different sources, the lack of common language in data management could be a large barrier that prevents information of biodiversity and ecosystem services from being widely usable to different stakeholders. Also, operability of data is different between databases and between data types, largely limiting the existing data to be applied for model calibration and validation. Considering the increasing visibility of data, platforms that facilitate user access such as Data Observation Network for Earth (DataONE; https://www.dataone.org/) will play a crucial part in the coming years. Specifically, while biodiversity information such as those archived in the Global Biodiversity Information Facility (GBIF; http://www.gbif.org/) and in the Ocean Biogeographic Information System (http://www.iobis.org/) are widely recognized and relatively-well organized, data for ecosystem services tend to be collected individually and more diversely.
Page 9 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
BOX 8.2: Examples of existing biodiversity databases at the species level A. Databases of occurrences, trends, and threats • GBIF -- http://www.gbif.org • IUCN Red List -- http://www.iucnredlist.org/ • Living Planet Index -- http://livingplanetindex.org/ • Global Population Dynamics Database -- http://www3.imperial.ac.uk/cpb/databases/gpdd • North American Breeding Bird Survey -- https://www.pwrc.usgs.gov/bbs/ • Global Invasive Species Database -- http://www.issg.org/database/welcome/ • WORMS World register of marine species (http://www.marinespecies.org) • EOL Encyclopedia of Life Global access to knowledge about life on Earth (http://eol.org) • AlgaeBASE: a list of the world’s algae. (http://www.algaebase.org) B. Databases of demography and life history characteristics • COMPADRE Plant Matrix Database & COMADRE Animal Matrix Database (http://www.compadre-db.org/) • MAPS: Monitoring Avian Productivity and Survivorship (http://www.birdpop.org/nbii2006/NBIIHome.asp) • BROT: plant trait database for Mediterranean Basin species (http://www.uv.es/jgpausas/brot.htm) • AnAge: Database of Animal Ageing and Longevity (http://genomics.senescence.info/species/) • PanTHERIA: life history, ecology, and geography of extant and recently extinct mammals (Jones et al. 2009 http://esapubs.org/archive/ecol/E090/184/) • FishBase: a global information system on fishes (http://www.fishbase.org/home.htm) • The Primate Life History Database (Strier et al. 2010 Methods in Ecol. Evol. https://plhdb.org/) • OBIS Ocean biogeographic information system (http://www.iobis.org) BOX 8.3: Examples of existing biodiversity databases at the ecosystem level • BEDIC numerical information on biodiversity and ecosystems. (https://www.naturalsciences.be) • Biodiversity information system for European biodiversity and habitat types (http://biodiversity.europa.eu) • Ecosystem service indicators database (http://www.esindicators.org) • ESP The Ecosystem services partnership (http://www.fsd.nl) • EcoDB Ecosystem database (http://ecomdb.niaes.affrc.go.jp) • NOAA National Oceanic and Atmospheric Administration (http://www.ngdc.noaa.gov) • MESP Marine Ecosystem Services Partnership: ecosystem service database (http://www.marineecosystemservices.org) • United Nations Framework Convention on Climate Change Database on ecosystem-based approaches to adaptation (http://unfccc.int) • EBM The Ecosystem-Based Management information about coastal and marine planning and management tools (http://www.ebmtools.org) 5
Generally, ecosystem services data are often produced by combining datasets sourced from multiple databases into a focal type of data (Tallis et al. 2012; MAES 2014). These datasets are diverse and can be physical, biological and social, such as satellite images, digital elevation models, LIDAR (Light Detection and Ranging) data, land/ocean use information, crowd-sourced data (e.g., for taxa
Page 10 of 33
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
distribution and phenology), meteorological data, human health statistics, and economic/financial statistics. Another reason why these diverse datasets are required is that, in the real-world decisionmaking, it important to identify trade-offs and synergies between multiple services (e.g., Bateman et al. 2013; Brandt et al. 2014). Handling such different datasets needs special skills and knowledge, a task generally not easy for the majority of stakeholders. At the local scale, the shortage of human resources can be as serious as data incompleteness. Currently, some synthesized information, which potentially facilitate non-experts to use ecosystem services information, are available online. For example, the Ecosystem Service Valuation Database of the Economics of Ecosystems and Biodiversity (TEEB) (http://www.fsd.nl/esp/80763/5/0/50), gives the global overview of the estimates of monetary values of ecosystem services, potentially benefiting local stakeholders who are unfamiliar with environmental economics. Another example is the Global Forest Change (http://earthenginepartners.appspot.com/), which makes it possible for groups without a remotesensing expertise to visualize and assess the changing status of forest coverage in a specific region of interest (Hansen et al. 2013). In addition to open data, open tools are drastically increasing; however, it is crucial to assist stakeholders use diverse datasets. In this regard, it is desirable to expand opportunities for learning how to handle different types of data, including online-learning modules and webinars that can be accessible worldwide. Many organizations, universities and research institutes now provide various databases; in addition to the information regarding to what types of available data they have, they must also be encouraged to provide the knowledge on how to use these data. Online learning, web database and crowd-sourcing can certainly bridge people working at different regions/nations, having different backgrounds, and lacking expertise for using scenarios and models for policyformation and decision-making.
8.2 Improving models 8.2.1 Basic research to fill thematic gaps and build functional linkages
30
35
As previous chapters have demonstrated, there is a wide variety of approaches to scenario analysis and modelling that can now be used to inform assessment of status and trends, to assess future risks, and to evaluate policy options. Despite recent advances in these approaches, there are significant gaps, both in types of models for analyzing and forecasting different ecological processes, and in linkages between different types of models. This section focuses on basic science needs, i.e., research directed towards further development of theoretical and conceptual underpinnings of ecological systems. This is the type of basic science research carried out by most academic ecologists. This section gives examples of research that would advance scenario analysis and modelling, in contexts and scales that are of interest to IPBES.
8.2.1.1 Thematic gaps 40
There is a need for research that leads to development of new types of models to analyse and forecast ecological processes that have so far not been the focus of much research. In this section, we give a few examples of these "thematic gaps".
Page 11 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Species interactions and community dynamics Models for performing scenario analyses and projecting regional biodiversity dynamics under IPBES will need to incorporate species interactions and community dynamics (including, for example, trophic interactions and disease dynamics). Although there is much theoretical and empirical research on trophic interactions and disease dynamics, these developments have not been translated into predictive tools at large temporal and spatial scales (Thuiller et al. 2013). For instance, while it is generally acknowledged that much of the impact of climate change will be through disruption of existing species interactions and emergence of new ones (Van der Putten et al. 2010), most large-scale models that project impacts of climate change on biodiversity either exclude such interactions or incorporate them only implicitly or under strong, simplifying assumptions (Albouy et al. 2014). When species interactions are explicitly included in predictive models of biodiversity, they are often limited to only two or few species, such as one-predator-one-prey (Fordham et al. 2013) and predator-prey-pathogen (Shoemaker et al. 2013); or they are limited to specific types of well-studied interactions such as pollination (Bascompte et al. 2006). Part of the reason for this thematic gap is that, in the context of projecting the effects of particular policy or management actions on specific systems, the challenges in community ecology are even greater than in population ecology of single species. Therefore, basic science investments that lead to incorporation of species interactions and community dynamics into scenario analysis and modelling at large spatial and temporal scales would benefit global and regional IPBES assessments. Research needs include large scale experiments (e.g., experimental translocations), long-term and large spatial scale monitoring of the effects of conservation or policy actions (e.g., monitoring following the establishment of protected areas and invasive species control measures), and studies designed to translate measurable properties (such as comparison of ecological niche models of potentially interacting species) to parameters commonly used in theoretical models of species interactions (such as interaction coefficients or partial derivatives of population growth equations; Tang et al. 2014). Early warning of regime shifts Another research need is developing practical early warning systems to anticipate ecological breakpoints, tipping points, and regime shifts (Leadley et al 2014). At the species level, warning systems based on current status and recent trends of populations have been in use for decades (Mace et al. 2008), and have been recently tested under scenarios of climate change (Stanton et al. 2014). At community or ecosystem levels, warning systems based on statistical properties of time series―such as increasing temporal variance and autocorrelation, and slowdown of system recovery from small perturbations―have been proposed (Scheffer et al. 2009) and empirically tested (Carpenter et al. 2011). Although much research has been done on regime shifts in ecosystems, there are significant gaps, with the result that no practical early warning system for regime shifts (i.e., a set of generally agreed-upon measurable indicators), is currently available for adoption by IPBES. Practical limitations include dependence on long-term time series data (which are not as practical as static measures such as spatial patterns often used at the species level), difficulty of determining critical thresholds for a specific ecosystem, difficulty of predicting the timing of the transition and the nature of the altered state. A promising research direction is linking theoretical research on network robustness and empirical research on indicators of resilience, which have been largely unconnected so far (Scheffer et al. 2012). A related, and also promising, research direction is using time series data of ecological variables to infer causal drivers of ecological change. Regime
Page 12 of 33
First Order Draft
5
10
15
20
25
IPBES Deliverable 3c
DO NOT CITE
shifts may be more predictable if we understand what drives the underlying ecological processes. Methods such as maximum likelihood (Wolf & Mangel 2008), convergent cross-mapping (Sugihara et al. 2012), and Bayesian model selection (Shoemaker and Akçakaya 2014) have been used to infer causes of species decline and to separate causality from correlation. Further development and refinement of these approaches will help advance the use of mechanistic models for building early warning systems as well as for evaluating the effect of policy options on biodiversity and ecosystem services. Response to variability and extreme events One critical research need related to regime shifts, at both species and ecosystem levels, involves the effects of changes in environmental variability and environmental regimes, and biodiversity responses to extreme events (Zimmermann et al. 2009). In particular, global climate change is expected to result in increased frequency and intensity of extreme weather events. Predicting the effects of this on the stability of biological systems requires multidisciplinary collaboration among climatologists and ecologists, as well as integration information from demographic models, physiological models, and predictions of climatic variability. Linking indicators to models There is need for research that links indicators and modelling, directed towards developing indicators that can be used to not only measure the current status but also to forecast the future state of biodiversity and ecosystem services, based on scenario analysis and modelling. Many of the currently used or proposed indicators (see section 8.1.1) are useful for assessing current status and recent trends of components of biodiversity and ecosystem services, but few can be projected into the future. Research that links indicators and modelling can fill this gap. The key is to develop indicators that are firmly based on scenario analysis and modelling so that future values of the index can be calculated for alternative policy options.
30
IPBES-relevant scales Most basic ecological research involves short time periods and small spatial scales, which may not be relevant to the global and regional assessments to be undertaken by IPBES. There is a need for investments in research on ecological processes at the spatial and temporal scales relevant to IPBES assessments. In addition, there is bias in taxonomic and regional coverage of basic research, with a disproportionate amount of research involving populations of few groups (such as birds and mammals), and focusing certain regions (such as northern temperate regions).
35
8.2.1.2 Functional linkage gaps
40
45
A second type of research need concerns the development of linkages between existing approaches to modelling different aspects of biodiversity and ecosystem services. One type of linkage that is needed is between human socio-economic systems and natural systems. Improving the coupling of the social and ecological components of models and scenarios requires well developed, specific feedbacks from the ecological to the social systems and vice versa (Carpenter et al. 2009). Linkages between human and natural systems may have complex structures, and may form cascades. For example, the effect of human activities on the world's climate is fairly well studied. There are also studies on the second link, the effects climate change on
Page 13 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
human activities, such as shifts in agriculture and urbanization. And the third link is the effect of these changes in human activities on biodiversity and ecosystem services, compounding the direct effects of climate change on natural systems. Other examples include the linkages among human population growth, land-cover change, and ecosystem services (Pereira et al. 2010). Such cascades of causal connections are often difficult to predict (Chapman et al. 2014; Watson et al. 2014). Research on the coupling of human and ecological systems that focuses on these causal chains will help modellers to make more realistic projections of future changes in biodiversity and ecosystem services. In addition, approaches to use time series data to infer causality mentioned above (under Early warning of regime shifts) will help untangle these causal chains. From the point of view of connecting scenario analysis to policy options, a critical research need involves the functional linkages between biodiversity and ecosystem services (Mace, 2012; Díaz et. al. 2007). As the previous chapters have emphasized (e.g., see Chapters 4, 6), models that predict the impact of ecological changes on human well-being are not well developed. One of the few well developed connections is between pollinators and human well-being (see IPBES Thematic assessment of pollinators, pollination and food production). There is a need for developing new models that link components of biodiversity to nature's benefits to people, including ecological services. Developing such models, tools, and methods will require basic research involving multidisciplinary teams of scientists as well as policy makers and other stake holders (see section 8.3). Development of the types functional linkages between different types of models of biodiversity and ecosystem services discussed above can be facilitated by research into mechanistic as well as statistical (e.g., correlative) relationships. For example, analysis of statistical relationships between environmental drivers (climate, land-cover) and biodiversity components (e.g., species occurrence) allows some predictive ability. Such an approach has been successfully implemented as ecological niche models and used to project the future potential distribution of species in response to environmental change (e.g. Guisan and Thuiller 2005). However, in order to predict beyond the current conditions, and to evaluate the impact of management and conservation options, a deeper understanding of ecological processes is needed. This need has led to the development of more mechanistic models that incorporate ecological processes such as dispersal and demography (e.g., Keith et al. 2008), and coupling of correlative and mechanistic approaches (Boulangeat et al. 2014). Similarly, the development of linkages discussed in this section will likely benefit from coupling of correlative or statistical methods with mechanistic models of ecological and socio-economic processes, such as some of the models incorporated in the InVest package (Daily et al. 2009). On the technological side of developing these linkages, there is a need to encourage development of models that can communicate with (or embedded in) software platforms that are designed for linking different models. Two main types of such platforms are "scientific workflow managers" and "integrated environmental modelling frameworks". Both of these approaches allow users to assemble and run a system composed of existing simulation models that can exchange data at run time. Examples of scientific workflow managers include Kepler (kepler-project.org), with applications in areas such as ecological niche modelling (Pennington et al. 2007) and environmental sensor data analysis (Barseghian et al. 2010); VisTrails (vistrails.org), recently applied to habitat modelling (Morisette et al. 2013); and Taverna (http://www.taverna.org.uk), recently applied to mapping potential distribution patterns (Leidenberger et al. 2014). The integrated modelling frameworks include OpenMI (openmi.org), Object Modeling System
Page 14 of 33
First Order Draft
5
10
IPBES Deliverable 3c
DO NOT CITE
(www.javaforge.com/project/oms), and Metamodel Manager (www.vortex10.org/MeMoMa.aspx), which have been applied to models of hydrology (Butts et al. 2014), sediment transport (Shrestha et al. 2013), trophic interactions (Prowse et al. 2013), and solar radiation (Formetta et al. 2013). An important difference between these systems is that the workflow managers are mainly designed for infrequent, unidirectional transfer of data among component models whereas the integrated modelling frameworks are designed for among-component interactions (i.e., feedbacks) and for frequent exchange of data among modules (e.g., passing key information at every time step). Other technological improvements include compatible spatial and temporal scales (coverage and resolution; see Chapter 6); data-based and region- or system-specific functional relationships; and interacting drivers (see Chapter 2).
8.2.1.3 Evolving guidelines, evaluations, prioritization 15
20
25
Research on many aspects of scenario analysis and modelling of biodiversity and ecosystem services is progressing at a rapid rate. Many of the approaches reviewed in this report will be further developed in the near future; others may become obsolete. Therefore, there is a need to ensure, through ongoing updates and new evaluations, that the review of available policy support tools and methodologies for scenario analysis and modelling of biodiversity and ecosystem services continues to reflect best available science. Similarly, there is a need for ongoing prioritization of research needs. Some of the research and development directions and needs identified in this chapter will have already matured in the next few years, and others will not be pursued, or will be proven to be not beneficial. Therefore, it is critical that IPBES develops mechanisms for research prioritization, in order to encourage basic research that advances scenario analysis and modelling in contexts and scales that are relevant to IPBES. This could be in the form of a standing Task Force on research prioritization that makes recommendations to research funding agencies about the significant gaps that remain in our understanding of the fundamental processes that are the subject of scenario analysis and modelling used in IPBES assessments.
8.2.2 Anticipating, verifying and validating models 30
35
40
45
8.2.2.1 More complex models need more verification Ecological systems are complex by nature since they typically consist of various components of differing dimension (nutrients, genes, individual organisms, populations, communities, ecosystems) and of different types (biotic and abiotic) that interact in many different ways (fluxes of matter or energy) at various spatial and temporal scales. Additionally, the biotic compartments of ecosystems can evolve, and interactions between biotic and abiotic components may create feedback loops that result in nonlinear, hence unexpected, ecosystem responses. Intuitive reasoning may suggest that considering an increasing number of details in models would improve the accuracy of ecological predictions, but this is not necessarily the case. There exists a clear gap in our current understanding of the link between model complexity and predictive accuracy, although statistical criteria exist to compare the predictive accuracy of sets of models. Filling this gap is critical to better understand under what circumstances additional model complexity is advised, and what kind of complexity is needed (see section 8.2.1). Complex models can cope with more details on processes or interactive effects. However, increasing complexity is also faced with challenges. Understanding the behavior of such models becomes difficult, since they commonly lead to emergent effects that cannot be
Page 15 of 33
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
predicted from the knowledge of their building blocks alone. Calibrating such models, assessing their sensitivity relative to assumptions, and measuring data uncertainty are additionally challenging. The analysis of complex models is the focus of intense research, not only in biology, but also in climatology, industry and statistics, with an increasing number of software packages to disseminate state of the art techniques (Pujol et al. 2013). Ecological research should capitalize from current advances in other fields that deal with complex systems.
8.2.2.2 Uncertainty propagation 10
15
20
25
30
35
40
With the rise of statistical and mechanistic predictive models of biodiversity and ecosystem services, quantifying, incorporating, and propagating uncertainty has become a key issue. Although uncertainty in complex models can be dealt with probabilistic techniques, it remains important to decipher epistemic uncertainty as a consequence of model predictions, and to promote a better communication between disciplines and stakeholders (Li et al 2013, see also section 8.3.4). The input data for biodiversity and ecosystem services models are often uncertain and are specified as a range of values or as statistical distributions. Using these types of quantification of uncertainty in the input data, Uncertainty Analysis aims to quantify the overall uncertainty of model results, in order to estimate the range of values the output could take (Regan et al. 2002). In recent years there has been an increasing interest in uncertainty analyses, mainly motivated by keeping imperfect data in data poor model environments instead of discarding them. Uncertainty and dependence modeling, model inferences, efficient sampling, screening and sensitivity analysis, and probabilistic inversion are among the most active research areas (Kurowicka & Cooke 2006). To date, despite few examples, and the awareness that different algorithms likely result in different projections, biodiversity and ecosystem services models are still much too often used blindly, without clear reporting of the underlying uncertainty in parameter estimation or the uncertainty resulting from the input data. Better integration of statistical analyses into mechanistically based fitting should foster appropriate characterization and reporting of uncertainty. Promising approaches for doing this include methods such as inverse modeling or approximate Bayesian computation (ABC), which produce a probability distribution for the estimated parameters (the posterior distribution) that are relevant for the reporting of uncertainty (Hartig et al. 2012). So far, however, a full treatment of uncertainty has been considered too time-consuming and complex to be achieved in biodiversity and ecosystem services models, and a full integration and partitioning of the uncertainty coming from different sources (such as climate or land-use models) is difficult to achieve. To meet this challenge, there is a need for mathematical, statistical and computational skills that extend beyond the range of standard ecological expertise, and include novel techniques mixing deterministic and random concepts that are usually considered as independent of each other. Despite these caveats, pragmatic approaches are encouraged, for instance by sub-sampling alternative climate projections for the same scenario to still give a basic representation of the uncertainty and by considering that parameters in mechanistic models should not be fixed to one value but rather to a probability density function based on prior knowledge. Progress will also come from integrating methods from other fields that already consider uncertainties (Fisher et al. 2011).
Page 16 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
8.2.2.3 Validation of models
5
10
15
20
25
30
35
40
45
Validation is at the basis of model judgment but also essential for the credibility of the scientific community towards stakeholders. It thus should be given top priority. Validation is the process of testing the behaviour of a model using a dataset not used for calibration or parameterization (Rykiel 1996). There are inherent limitations to the validation process: (i) the perception of the predictive performance of the model depends on the techniques used, and (ii) by definition validation also strongly depends on the availability and quality of datasets. Overcoming these limitations requires building clear predictions, using robust statistical methods and generating enough data (either experimental or observational) so as to reach the level of quality needed for validation. Biodiversity and ecosystem services models are subject to these limitations as they usually face a strong difference between the scale of the prediction and the scale of measurement. For instance, most dynamic vegetation models use growth curves that are calibrated over dozen of individuals measured in situ with precise climate measurements, while these curves are then extrapolated at large spatial scales and with a resolution over 20x20 km for which climate is highly smoothed. To overcome this limitation cross-scale validation has been proposed (using data generated at a lower scale to validate models built for a larger scale), but even here the question of the interchangeability of processes between scales has not been truly addressed (Morozov & Poggiale 2012). Another approach is to calibrate models on a dataset and validate them on a spatially or temporally independent dataset. In practice, however, two independent datasets might not be available, and therefore the calibration and validation datasets are often defined as random subsets of the original one (Araújo et al. 2005). The calibration and validation datasets are thus often partially dependent because abiotic variables show strong spatial or temporal autocorrelation. The non-independence of these datasets thus may yield overly optimistic estimates of the accuracy of model projections. The use of simulated data could be considered an alternative and a priority to test the ability of the model to retrieve known patterns (Zurell et al. 2010). Moreover, while spatially distinct samples are frequent, temporally distinct samples are scarce because of the lack of long-term datasets at the temporal resolution needed in anticipatory predictions. Spatially and temporally dynamic models of biodiversity or ecosystem services must be validated against dynamic data. Although fruitful initiatives were launched in the past decades (e.g., the ILTER network: http://www.ilternet.edu, GEO-BON: http://geobon.org), long-term datasets are still the exception more than the rule. Promising new directions become available from sediment archives (Willerslev et al. 2014) or from remote sensing/air-borne data that now provide systematic information about change in diverse communities through time, or about spatial and temporal change in biophysical variables such as NDVI (Petorelli et al 2014). Resolving these issues will require special efforts in long-term data collection, coordination and sharing (see section 8.1.3). However, more than their quantity, a particular effort will be necessary to improve the quality of the data used and their spatial and/or temporal independence, which condition the quality of the validation process (section 8.1.2).
8.3 Improving Utility In this section we tackle the challenge of improving the utility of scenarios and models for policy and decision-making. We first examine how to best engage stakeholders. Next we discuss how to improve the links between modeling parameters and policy options. We examine how uncertainty
Page 17 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
can be better communicated to scenarios and model users. Finally, we propose avenues for improving the relevance of scenarios and models outputs to policy makers.
8.3.1 Stakeholders 5
10
15
20
25
30
35
40
8.3.1.1 Why adequate stakeholder involvement is indispensable? Assessment of biodiversity and ecosystem services is a multi-scale issue in its essence (Cash et al. 2002; Reid et al. 2006). What is regarded as ecosystem services depends on scale, i.e., it changes when the stakeholder characteristic of its focusing scale changes. Different stakeholders dispersed over the assessment region have different interests according to their lives, which are tightly connected to specific scales. This plurality of recognition is the cause of potential conflict of interest and at the same time the source of resilience in the assessment process (Sabatier et al. 2005; Reid et al. 2006; Cash et al. 2006; Berks 2007). Among the reasons of past failures in assessments is the exclusion of stakeholders from the assessment process, because of ignorance, prioritizing efficiency to equity, and intuitive exclusion (Cash et al. 2002; Sabatier et al. 2005; Berks 2007; Lucas et al. 2010). Exclusion of stakeholders sometimes decreases manpower for the assessment. Furthermore, it makes the excluded people distrust assessment process and perceive it as unfair, hindering communication that is essential to resolve conflicts, and leading to the failure of the project (Cash et al. 2002; Sabatier et al. 2005; Lucas et al. 2010). Sound assessments have three fundamental qualities: relevance, credibility and legitimacy. Relevance requires that the assessment information is significant in relation to stakeholders’ priorities or decision-making issues. Credibility requires that the assessment meets standards of scientific rigor and technical adequacy. Finally, legitimacy requires that stakeholders perceive the assessment process as unbiased and meeting standards of political and procedural fairness (Cash et al. 2002; Lucas et al. 2010) (section 8.3.4). Models and scenarios, however, can help overcome some of the challenges in assessments with multiple stakeholder involvement. Relevance, credibility and legitimacy of assessment can be adequately balanced and enhanced through the participatory process of making models and scenarios (section 8.3.2), and the risk communication process about the inherent uncertainties (section 8.3.3).
8.3.1.2 Multiple uses of models and scenario for diverse stakeholders With the involvement of diverse stakeholders, models and scenarios for assessment can exert functions other than supplying credible information for decision-making, including: enhancing communication between stakeholders and scientists, fostering trust, and empowering participants toward the assessment goal. In order for this multiple functions to occur, clear identification and early involvement of stakeholders in the assessment process are indispensable (Lucas et al. 2010). Models and scenarios can be used to improve the transparency and relevance of policy making, by incorporating necessary demands and information of each stakeholder. Models and scenarios allow comparing multiple options, and making assumptions, trade-offs and potential conflicts of interests between stakeholders explicit. The information incorporated in models and scenarios with the help of stakeholders with multiple backgrounds can strengthen credibility and usefulness of these tools (Cash et al. 2002; Sabatier et al. 2005; Lucas et al. 2010). This can be achieved by developing participatory models and scenarios.
Page 18 of 33
First Order Draft
5
IPBES Deliverable 3c
DO NOT CITE
Improved visualization of model and scenario outputs is another desirable direction. GIS is a common platform for models and scenarios to enhance communication between multiple stakeholders. GIS visualization promotes transparency: it can project model and scenario outputs, display the appropriate spatial level for each stakeholder, and spatially identify conflict of interests. Visualization of the outcome of stakeholders’ efforts by adopting adequate indicators is also effective for the enhancement of the feedback process by empowering the concerned stakeholders.
8.3.2. Linking scenarios and policy options to model parameters 10
15
20
25
8.3.2.1 Identifying the problem Modelling and scenario development are among the primary tools used to support a back-casting of past evolutions and projecting future trends based on historical and current dynamics (Leeuw 2004). Using modelling and scenario tools helps explore a broad range of policy options and socio-economic developments and their impacts, not only on the environment, but also on well-being, and to identify different pathways towards sustainability (Abdelgalil, E.A. et al. 2001). An overarching question is how to include advanced social systems in modeling to respond to policy relevant information? Changes in biodiversity and ecosystem services are the results of individual, local and regional decisions guided by processes and actors operating across different spatial and temporal scales and organizations (Lenton, M. et al. 2008, Leadley, P. et al. 2014). Representing these cross-scale interactions and feedbacks will constitute a major step in advancing our capability of modelling biodiversity and will better connect the different communities. Many studies showed the need for better methods to convert socioeconomic scenarios and general policy options to specific changes in model’s input parameters (e.g., how a particular law, regulation or policy changes the demand for particular types of biodiversity distribution and ecosystem services) (Elmqvist, T. et al. 2003, Lontzek and Narita 2010).
30
Social structures have particular patterns for the processes of diffusing knowledge and technologies (knowledge systems), and for institutional and cultural preferences (socio-political contexts). Those elements are very place dependent and have a huge influence on the evolution of biodiversity and drivers of ecosystem dynamics. Therefore, there is a need to represent them as collective macroprocesses in models inputs and parameters (if they cannot be captured as individual parameters) and in socioeconomic scenarios to foster a new understanding of coupled human-environment system (Costanza, R. et al. 2007, Butzer and W. 2012).
35
8.3.2.2 Linking policy needs with model and scenarios
40
45
The current and future development of biodiversity models and scenarios requires a full consideration of policy options and actions. A starting point to address current and future policy needs is a greater interdisciplinary understanding of the connections between natural and human risks on biodiversity in past, present and future timescales (Costanza, R. et al. 2007). The modeling communities in the natural and social sciences are relatively isolated from each other, and a substantive collaboration effort is needed. Model co-design will promote intellectual fusion between communities, helping them to formalize and integrate different discourses into a consistent framework (Rindfuss, R.R. et al. 2004). Such an effort will necessitate overcoming linguistic, epistemological, technical and other hurdles between the modelling communities.
Page 19 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
The participations of stakeholders on the co-design of inputs, parameters and outcome of models is also essential (Future-Earth 2013). Feedbacks through the ecological and physical earth system both constrain and moderate the future options of decision makers. From the outset, policy makers can help identify topics, questions and approaches that are relevant to them. The solutions brought by science will thus be driven by user needs. This approach will constitute a mutual learning experience whereby the modeling community will learn about the more imperious challenges and operative options on biodiversity from stakeholders, while stakeholders will gain from scientists a better understanding of the environmental challenges faced and the solutions that science can provide. A key issue is how to manage tradeoffs and also opportunities for synergies between biodiversity conservation, food security and livelihoods across contrasting social-ecological regions. In particular the community needs to: i) identify the nature of these tradeoffs and synergies across socialecological systems and regions of the world; ii) identify the key ecosystem services that are at stake in these tradeoffs; iii) identify the biophysical and societal drivers that contribute to exacerbating the tradeoffs and those that contribute to reducing them; iv) identify opportunities for synergies between biodiversity conservation, food security and livelihoods that are most suitable for particular social-ecological contexts (McCarthy, Lipper et al. 2012, Smith, Haberl et al. 2013, Klapwijk, J. et al. 2014). Most current biodiversity models emphasize the predictive power of environmental processes, but are not driven by socioeconomic storylines. One exception is Integrated Assessment Models (IAM), which were designed for evaluating policy pathways (Vuuren, Isaac et al. 2011, Vuuren, P. et al. 2012). Additionally, most models are not able to reproduce state transitions in the coupled humanenvironment system, and cannot structurally produce qualitative shifts or collapse (tipping-points) (Leadley, P. et al. 2014). However, research shows that even simple systems with smooth behavior, when coupled as a co-evolving system, can display strong non-linearity and abrupt transitions (Scheffer, M. et al. 2003, Folke, C. et al. 2004, Leadley, P. et al. 2014). Simple complex systems models address feedbacks and qualitative shifts, but cannot speak to the fuller processess that includes regime shifts and tipping points.
8.3.2.3 Exploratory versus normative scenarios Scenarios can be developed using a variety of approaches (Kok et al 2011; Alcamo et al. 2001). In exploratory scenarios, the analysis starts in the present and different plausible future trajectories are explored by stakeholders, often across major axes of uncertainty on social-ecological dynamics, and using associated narratives for the unfold of events from present to the future. In normative scenarios, stakeholders agree on a set of desirable futures, and then a backcasting analysis is performed of the socio-ecological pathways that may lead to those desirable futures. There are advantages and disadvantages to each approach, and some exercises have tried to combine elements of both approaches (Kok et al. 2011). Exploratory scenarios foster creative thinking and exchange of viewpoints between different stakeholders, but do not always provide clear actions that decision makers should implement to reach desirable outcomes. Normative scenarios are more likely to provide clear policy pathways but have been criticized for being value laden.
Page 20 of 33
First Order Draft
5
10
15
20
25
30
IPBES Deliverable 3c
DO NOT CITE
Recent large scale environmental scenarios have been following an approach that is perhaps more akin to normative scenarios, and that should be considered by IPBES activities. The new scenarios of IPCC, define plausible relative concentration pathways (RCP) of greenhouse gases to achieve different levels of radiative forcing for the end of the century (Moss 2010). Then, emission pathways and associated socio-economic scenarios were developed in order to produce those RCP scenarios. In the end, these are integrated with climate models and analysis of impacts. Another approach was followed by the Rethinking Global Biodiversity Strategies scenarios (ten Brink et al. 2010). These scenarios consider a set of policy options aimed at reducing biodiversity loss, such as increase in protected areas, changes in diet and improving forest management. The effects on biodiversity of the implementation of those options are then assessed over time. More recently, the Roads from Rio+20 Scenarios (PBL 2012) defined a vision for biodiversity in 2050, and then examined three pathways, each one with its one set of policy options, that can lead to that vision.
8.3.2.4 Building a framework for collaboration IPBES activities should facilitate and create conditions, frameworks and infrastructure for collaborative design of biodiversity and ecosystem function monitoring schemes, data harmonization and integration, prediction, reporting and information sharing. IPBES activities should: • Identify key global biodiversity questions to which we can develop robust answers; • Overcome disciplinary barriers in modeling, data collection, selection and management; • Identify co‐design and co-development of best practices that respond to policy needs; • Define modelling methodologies appropriate for the different social contexts and policy needs • Identify robust model integration techniques that respond to current and future development requirements; • Establish a permanent dialogue between modelers and decision-makers to address issues such as common understanding of concepts, transdisciplinarity, and Infrastructure for resource and knowledge sharing;
8.3.3 Communicating uncertainties 8.3.3.1 The importance of proper communication
35
40
45
Natural systems are complex systems and uncertainty is just another unavoidable component. Uncertainties on how the various natural components interact, or on whether all the relevant components have been identified usually appear. Models are considered useful tools for organizing knowledge in a systematic way, testing its consistency and identifying gaps. However, uncertainties in various aspects of the modeling and analysis process, from input data to model output, will remain. Complexity increases many fold when the human components, both social and cultural, that intervene in the management of natural resources need to be incorporated into the models. A critical challenge in communicating the results of scientific research arises when those results contain uncertainties. It is highly important that the various types of uncertainties that will necessarily appear in the modeling process, as well as in the scenario analysis, be clearly communicated to all stakeholders and decision-makers so that there is full understanding of the
Page 21 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
relative weight of the output, their implications and the risks involved. However, decisions can be made even when gaps in information appear, or data are not totally reliable, or ample variability is observed and risks are identified (see Section 8.2.2). 5
10
15
Recent experience, mostly related to the communication of uncertainties related to climate change or to potential pandemics, has opened the way to a more systematic analysis of how people perceive the uncertainty inherent to scientific research. These problems have captured the attention of both climate and social scientists (Janssen et al., 2005; Handmer & Proudley, 2007; Kloprogge et al., 2007; Pidgeon & Fischhoff, 2011). Research communities have emerged in which people from different fields, such as climate and environmental scientists, historians, social scientists, philosophers, examine issues of uncertainty with respect to global environmental problems with the purpose of improving the capacity to discuss and weigh related policy recommendations. It is particularly important that the process of building a dialogue between scientists/modelers and stakeholders/decision-makers explicitly involves communicating the weaknesses that inevitably appear regarding present knowledge and the way it can be used. Being clear about what the shortcomings are should permit to increase the confidence between interlocutors.
8.3.3.2 Understanding uncertainty in model construction 20
25
30
35
40
Models are a representation of the real world but what they represent depends strongly on the perception that the modeler has about the real world and on the components of the perceived system that are taken into account in building the conceptual model. A model may be general (can be useful in many different situations), realistic (parameters and variables are measurable), and precise (accurate quantitative output), but it is impossible to have a perfect model that can be all three simultaneously. Models are often built for gaining a deeper understanding of the interactions between systems’ components and for responding to questions about the systems’ functioning. Depending on how well the model answers the questions that are being posed, it will be labeled more or less adequate. Hence, the limitations of a model need to be assessed and adequately informed to the stakeholders that will be using the outputs. Weakness and gaps in data Data are essential for drawing conceptual models that will later translate into quantitative or clean qualitative models but also indispensable for running those models and for performing scenario analysis. When the information is incomplete, not reliable, imprecise, fragmented, contradictory, or in any way deficient, it is also fundamental that stakeholders can understand that even a simple model based on very general data can still be useful at providing insight on the possible effects of different alternatives. There are diverse mathematical or statistical techniques that allow dealing with information deficiencies, such as fuzzy inference systems and uncertainty-based information theory (Klir & Yuan, 1995; Cao 2010). One advantage of fuzzy inference systems is that they allow incorporating qualitative information that “local experts” and stakeholders may volunteer into the more rigorous framework of model construction. Qualitative reasoning is an area of artificial intelligence that is concerned with the construction of knowledge models that capture insights domain experts have of systems’ structure and their functioning (Recknagel, 2006).
Page 22 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
One valuable source of information may be found in indigenous peoples who own a wealth of traditional knowledge. However, they may not trust persons outside their community enough as to share that knowledge straightforwardly. This requires the development of participation channels through the experience of anthropological scientists. Efforts should be made to systematically gather and organize this type of information. Weakness in assumptions A model is as good as the assumptions that were made in its construction, that is, what it has been accepted as true or as certain to happen. When stakeholders and policymakers are involved in the modeling process, these assumptions can be discussed and analyzed in the light of the data that will be used. Stakeholders can provide, and modelers should ask for, additional information to enrich the conceptual model as the discussions generate feedback and the resulting iterative process allows improving the model. Also, the knowledge of “local experts”, that is people who may not be scientist or have any particular training but know the ecosystem functioning well and/or understand its response to diverse types of pressure can and should be taken into consideration. In the process, modelers have to be open and ready to listen to what the other actors have to say, without presuppositions or prejudice. It is crucial that this be organized in such away that all stakeholders may be satisfied with the manner in which the discussions are conducted as well as be aware of the gaps and weaknesses that may remain in the conceptual model. Missing and overlooked components It may happen that in the process of building the conceptual model, some component is left aside because of technical reasons (serious gaps in data series, difficulties in parameterization, lack of knowledge on particular aspects, etc.) or because its importance has not been detected or assessed until later. However, it might be possible to overcome these. When data series are incomplete, it is possible to generate artificial data and then perform a statistical analysis of the output of the model. When some parameters are not known, it is possible to utilize a range of reasonable values inspired on better known similar species and check the model output for consistency. When there is insufficient knowledge on some particular aspect, it is possible to state various hypotheses and test whether the model outputs are reasonable. This is not a sensitivity analysis and requires contrasting the output with other data sets. In any case, these shortcomings should be acknowledged and discussed as part of the participative process of model improvement.
8.3.3.3 Understanding outputs and the limitations in their scope, translation and scope definition Once the models have gone through a process of iterative improvement, have been properly scaled, calibrated, and verified, the models results need to be understood within the context of the data and the assumptions. Here, the users may be those stakeholders involved in the modeling process, but often they could be decision-makers at other levels who may want to use the available information and results, or even the public in general. If the model users participated in the modeling process, it might be easier to give the adequate weight to the output and translate the meaning of the results because of the previous involvement and the understanding of the modeling process. Nevertheless, the results should be presented in a clear, consistent, and precise way, giving preference to graphic forms or to tables that synthesize the
Page 23 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
main points. It should be made clear what the uncertainties in the output are, what are the implications, and all that which is not implied (Janssen et al., 2005).
5
10
15
If the intended audience has not participated in the model-construction process, much more attention needs to be given to communicating the outputs in a way that minimizes misinterpretation and does not generate confusion or mistrust. Kloprogge et al. (2007) recommend Progressive Disclosure of Information (PDI), which “entails implementation of several layers of information to be progressively disclosed from non-technical information through more specialized information, according to the needs of the user”. The audience needs to make sense of the information that is being provided, so it may be important to give it in the context of the assessment and to include some idea of how it was obtained. Uncertainties need to be set in the context of the key messages that are being conveyed, and the implications of the uncertainties need to be explained. It may also be important to offer information on how the uncertainties can be or may be treated or dealt with. Over all, it is essential that the audience perceives the information as being useful for making policy decisions, or for use in political debates, or for forming personal opinions relative to policy advice.
8.3.3.4 The need to improve the communication of probabilistic results 20
25
30
35
40
45
All biological dynamical systems evolve under stochastic forces. In a stochastic or random process there is some indeterminacy. Even if the initial condition or starting point is known, there are several directions in which the process may evolve. Models of biological systems that need to be realistic may have to include random influences, since they are linked to subsystems of the real world that cannot be sufficiently isolated from effects external to the model. In some situations it will be necessary to use stochastic models, for instance, in order to understand how random events may affect certain outcomes. In nonlinear dynamical systems with system noise, the noise will often change the corresponding deterministic dynamics. Stochastic effects influence the dynamics, and may enhance, diminish or even completely change the dynamic behavior of the system. Translating the meaning of output from stochastic models to persons without professional or specialized knowledge in the subject often generates confusion because there is a whole set of possible outcomes and the results are given in terms of probabilities. Normally information concerning probabilities is susceptible to biases and misinterpretations, as the average citizen does not understand the concept easily (Handmer & Proudley, 2007). Research on cognitive biases and prospect theory (behavioural economic theory that describes the way people choose between probabilistic alternatives that involve risk) indicates that people have difficulty in correctly interpreting probabilities because they are more likely to act to avoid a loss than to achieve a gain (Kahneman & Tversky 1979; Kahneman et al. 1982; Kahneman 2011).
8.3.3.5 Dealing with uncertainties in decision-making The process of constructing models, proposing scenarios and analyzing them as means of learning in advance about the effects and implications of policies on ecosystems and ecosystem services is not only a technical matter. The whole process is embedded in the cultural setting of the societies that are part of those ecosystems and use their resources. Communicating effectively with those stakeholders requires the participation of interdisciplinary professionals with diverse skills and
Page 24 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
broader intellectual capabilities, in particular social scientists who understand the institutions and the social structure in the region, helping modelers to notice relevant issues, but who can also contribute in helping society better understand and solve environmental problems. 5
10
15
20
25
30
35
40
45
8.3.4 Linking Outputs to Policy 8.3.4.1 The role of Boundary Institutions The process whereby stakeholders engage in a modeling assessment includes the definition of the relevant variables, assumptions, methods and parameterization, all the way to communicating results, uncertainties and caveats, in the appropriate language, to different audiences (Cash et al. 2003; Folke et al. 2005). There is a variety of science-to-policy interfaces (van der Hoven and Chabason 2009), but most have in common the existence of some institutional way of facilitating or enabling the above functions, over the usually long periods of time that are necessary for some knowledge to be converted into the policies of a government, or a group of governments, or local decision-makers. Such institutions have been called boundary or bridging institutions (Cash et al. 2003; Folke et al. 2005; Cash et al. 2006). The role of boundary institutions in facilitating the science to policy process is crucial given the multiscale features of most realistic biodiversity-governance problems (see 8.3.1), the variety of stakeholders (8.3.2), and the serious problem of communicating (8.3.3) the assumptions and the results of “boundary objects” (BO) like maps, models, scenarios and assessments (Cash et al. 2003).
8.3.4.2 Features of effective science to policy processes. Analysis of a variety of case-studies allowed Cash and co-workers (Cash et al. 2003) to propose that there are several features of successful transfer of the results of knowledge to policy. Legitimacy means that the relevant stakeholders are included in the process and accept it as consistent with their values and perceptions of legality and fairness. For instance, early discussion of the features of the Access and Benefit Sharing protocol of the CBD were marred by exclusion of indigenous people’s representatives from the negotiation. Saliency means that the features of the BO must be relevant for the decision makers. It is useless to develop a very sophisticated BO if its features or outputs are legally, economically or culturally irrelevant to the policy makers. Credibility means that the stakeholders are willing to accept the features of the BO as believable, performed with at least acceptable standards of rigor. If stakeholders belonging to a traditional or indigineous culture are part of the assessment, credibility often becomes a very difficult issue due to non-overlapping perceptions about local vs. global contexts, non-shared values and differences in ontological assumptions like the object/subject dichotomy (Agrawal 1995). Besides the above, problems with the multi-scalar features of many governance processes often arise in biodiversity policy-making (Cash et al. 2006). From a governance perspective, important factors changing with scale are political and jurisdictional units, the characteristic time-frames for decision making, institutional arrangements and the scopes of different types of knowledge systems (formal scientific claiming to be general and abstract, whereas local and indigenous knowledge tends to be specific and experiential). The challenges of producing BOs that are relevant across scales is a very difficult one, since it is easy to fail to recognize the heterogeneities in the way that scales are perceived by different stakeholders (Cash et al. 2006). A successful science to policy process then should be aware of the scale problems and be able to tackle those challenges.
Page 25 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
Finally, the BOs resulting from a science to policy process should be communicated actively using the right translation of terms, and concepts, and, if needed, a mediation between stakeholders with different languages, usages and histories (Cash et al. 2003). 5
10
15
20
25
Such demanding and complicated tasks are better performed institutionally. Boundary organizations, mandated to act as intermediaries between the worlds of knowledge and policymaking can specialize in organizing legitimate, salient and credible science to policy processes, can be accountable to different groups and can translate and communicate in appropriate ways the BOs resulting from the processes. There are many examples of such organizations, working at very different levels of the biodiversity governance scales. For example, the CGIAR and other international centers (Cash et al. 2003), research bodies of the MEAs (van der Hoven and Chabason 2009), international or large ational NGOs, and national research centers and universities (Sarukhán et al. 2014). Many of these have been proved to communicate rather complex BOs, for instance, the result of sophisticated modeling (Guisan et al. 2013).
8.3.4.3 What should IPBES do to become a successful boundary institution? One major question for IPBES is whether it can attempt to become a boundary organization. This question very much depends on the scale (8.3.1.1) at which it decides to concentrate its operations. At a global scale, working for stakeholders accustomed to the procedures and rules of multilateral organizations, and used to accept scientific results as credible in general, it is very likely that IPBES can mandate, support or perform assessments fulfilling the features recommended in the literature. However, at scales closer to the local it is much more difficult that a multinational organization, with procedures and rules borrowed from the culture of international diplomacy can be legitimate, credible and relevant to the large variety of stakeholders that biodiversity governance demands at such more local scales. Perhaps in these cases IPBES should partner with existing boundary institutions, which can be NGOs, international research centers, or universities, searching for organizations with the experience, contacts and even the language capacities to become true boundary organizations at scales closer to the local.
30
Page 26 of 33
First Order Draft
IPBES Deliverable 3c
DO NOT CITE
References
5
10
15
20
25
30
35
40
45
Abdelgalil, E.A., Cohen and S.I. (2001). "Policy modelling of the trade-off between agricultural development and land degradation—the Sudan case." Journal of Policy Modeling(23): 847–874. Agrawal A (1995) Dismantling the divide between indigenous and scientific knowledge. Development and Change 26: 413-439. Albouy, C., Velez, L., Coll, M., Colloca, F., Le Loc'h, F., Mouillot, D. & Gravel, D. (2014) From projected species distribution to food-web structure under climate change. Global Change Biology, 20, 730-741. Alcamo, J. (2001), Scenarios as tools for international environmental assessments, Copenhague, European Environment Agency. Alkemade, R., van Oorschot, M., Miles, L., Nellemann, C., Bakkenes, M. & ten Brink, B. (2009) GLOBIO3: A Framework to Investigate Options for Reducing Global Terrestrial Biodiversity Loss. Ecosystems 12: 374–390. Araújo M.B., Pearson R.G., Thuiller W. & Erhard M. (2005). Validation of species-climate impact models under climate change. Glob. Change Biol., 11, 1504-1513. Baillie J.E.M., Collen B., Amin R., Akcakaya H.R., Butchart S.H.M., Brummitt N., Meagher T.R., Ram M., Hilton-Taylor C. & Mace G.M. (2008). Toward monitoring global biodiversity. Conservation Letters, 1, 18-26. Barnosky A.D., Matzke N., Tomiya S., Wogan G.O.U., Swartz B., et al. (2011). Has the Earth’s sixth mass extinction already arrived? Nature 471(7336):51–57 Barseghian, D., Altintas, I., Jones, M. B., Crawl, D., Potter, N., Gallagher, J., ... & Hosseini, P. R. (2010). Workflows and extensions to the Kepler scientific workflow system to support environmental sensor data access and analysis. Ecological Informatics, 5(1), 42-50. Bascompte J., Jordano P. & Olesen J.M. (2006). Asymmetric coevolutionary networks facilitate biodiversity maintenance. Science, 312, 431-433. Bateman I.J., Harwood A.R., Mace G.M., Watson R.T., Abson D.J., Andrews B., Binner A., Crowe A., Day B.H., Dugdale S., Fezzi C., Foden J., Hadley D., Haines-Young R., Hulme M., Kontoleon A., Lovett A.A., Munday P., Pascual U., Paterson J., Perino G., Sen A., Siriwardena G., van Soest D. & Termansen M. (2013). Bringing Ecosystem Services into Economic Decision-Making: Land Use in the United Kingdom. Science, 341, 45-50. Benson, B.J. et al. (2010) Perspectives on next-generation technology for environmental sensor networks. Front. Ecol. Environ. 8, 193–200 Berkes, F. 2007. Community-based conservation in a globalized world. PNAS 104, 15188-15193. Bontemps, S. et al. (2011). Revisiting land cover observations to address the needs of the climate modelling community. Biogeosciences 8, 7713–7740 Borer et al. 2014…..Methods in Ecology and Evolution Borgman C.L. (2012). The conundrum of sharing research data. Journal of the American Society for Information Science and Technology, 63, 1059-1078. Boulangeat, I., Georges, D., Dentant, C., Bonet, R., Van Es, J., Abdulhak, S., Zimmermann, N.E. & Thuiller, W. (2014) Anticipating the spatio-temporal response of plant diversity and vegetation structure to climate and land use change in a protected area. Ecography, doi: 10.1111/ecog.00694 Brandt P., Abson D.J., DellaSala D.A., Feller R. & von Wehrden H. (2014). Multifunctionality and biodiversity: Ecosystem services in temperate rainforests of the Pacific Northwest, USA. Biol Conserv, 169, 362371. Buchanan, G.M., Nelson, A., Mayaux, P., Hartley, A., Donald, P.F. (2009). Delivering a global, terrestrial biodiversity observation system through remote sensing. Conservation Biology 23, 499–502. Butts, M. et al., 2014. Embedding complex hydrology in the regional climate system – Dynamic coupling across different modelling domains. Advances in Water Resources, 74, pp.166–184 Butzer and K. W. (2012). "Collapse, environment, and society." Proceedings of the National Academy of Sciences 109(10): 3632-3639. Cao, Bing-Yuan (2010) “Optimal Models and Methods with Fuzzy Quantities”. Springer. Carpenter, S.R., Mooney, H.A., Agard, J., Capistrano, D., Defries, R.S., Díaz, S., Dietz, T., Duraiappah, A.K., Oteng-Yeboah, A., Pereira, H.M., Perrings, C., Reid, W.V., Sarukhan, J., Scholes, R.J. & Whyte, A. (2009) Science
Page 27 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
for managing ecosystem services: Beyond the Millennium Ecosystem Assessment. Proceedings of the National Academy of Sciences of the United States of America 106: 1305–1312. Carpenter S.R., Cole J.J., Pace M.L., Batt R., Brock W.A., Cline T., Coloso J., Hodgson J.R., Kitchell J.F., Seekell D.A., Smith L. & Weidel B. (2011). Early warnings of regime shifts: a whole-ecosystem experiment. Science, 332, 1079-1082. Cash DW, Adger N, Berkes F, Garden P, Lebel L et al. (2006) Scale and cross-scale dynamics: governance and information in a multilevel world. Ecology and Society 11(2): 8-19. Cash DW, Clark WC, Alcock F, Dickson NM, Eckley N et al. (2003) Science and technology for sustainable development special feature: knowledge systems for sustainable development. Proc Natl Acad Sci USA 100(14): 8086-8091. Cash et al. 2006. Scale and cross-scale dynamics: governance and information in a multilevel world. Ecology and Society 11(2):8. Chapin, E. S., et al. 2009. Ecosystem stewardship: sustainability strategies for a rapidly changing planet. TREE 25, 241-249. Chapman S., Mustin K., Renwick A.R., Segan D.B., Hole D.G., Pearson R.G. & Watson J.E.M. (2014). Publishing trends on climate change vulnerability in the conservation literature reveal a predominant focus on direct impacts and long time-scales. Diversity and Distributions, 20, 1221-1228. Collins, S.L. et al. (2006) New opportunities in ecological sensing using wireless sensor networks. Front. Ecol. Environ. 4, 402–407 Costanza, R., Graumlich, L., Steffen, W., Crumley, C., Dearing, J., Hibbard, K., Leemans, R., Redman, C., Schimel and D. (2007). "Sustainability or to collapse: What can we learn from integrating the history of humans and the rest of nature?" Ambio 36(7): 522-527. Costello M.J., Michener W.K., Gahegan M., Zhang Z.Q. & Bourne P.E. (2013). Biodiversity data should be published, cited, and peer reviewed. Trends Ecol Evol, 28, 454-461. Daily, G.C., Polasky, S., Goldstein, J., Kareiva, P.M., Mooney, H.A., Pejchar, L., Ricketts, T.H., Salzman, J. & Shallenberger, R. (2009) Ecosystem services in decision making: time to deliver. Frontiers in Ecology and the Environment 7: 21–28. Damuth, J. 1987. Interspecific allometry of population density in mammals and other animals: The independence of body mass and population energy-use. Biological Journal of the Linnean Society 31, 193-246. Di Marco M, M. Cardillo, H.P Possingham, K.A. Wilson, S.P. Blomberg, L. Boitani and C. Rondinini. 2012. A novel approach for global mammal extinction risk reduction. Conservation Letters 5, 134-141. Diaz, S., Lavorel, S., de Bello, F., Quétier, F., Grigulis, K. & Robson, T.M. (2007) Incorporating plant functional diversity effects in ecosystem service assessments. Proceedings of the National Academy of Sciences 104: 20684. Elmqvist, T., Folke, C., Nystrom, M., Peterson, G., Bengtsson, J., Walker, B., Norberg and J. (2003). "Response diversity, ecosystem change, and resilience." Frontiers in Ecology and the Environment 1(9): 488494. Fisher J., Harel D. & Henzinger T.A. (2011). Biology as Reactivity. Communications of the ACM, 54, 7282. Folke C, Hahn T, Olsson P, Norberg J (2005) Adaptive governance of social-ecological systems. Annual Review of Environmental Resources 30: 441-473. Folke, C., Carpenter, S., Walker, B., Scheffer, M., Elmqvist, T., Gunderson, L., Holling and C. S. (2004). "Regime shifts, resilience, and biodiversity in ecosystem management." Annual Review of Ecology Evolution and Systematics 35: 557-581. Fordham D.A., Akçakaya H.R., Brook B.W., Rodriguez A., Alves P.C., Civantos E., Trivino M., Watts M.J. & Araujo M.B. (2013). Adapted conservation measures are required to save the Iberian lynx in a changing climate. Nature Climate Change, 3, 899-903. Formetta, G., Rigon, R., Chávez, J. L., & David, O. (2013). Modeling shortwave solar radiation using the JGrass-NewAge system. Geoscientific Model Development, 6(4), 915-928. Future-Earth (2013). Future Earth Initial Design, Interim Secretariat-ICSU.
Page 28 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Grose M.J. (2014). Thinking backwards can inform concerns about 'incomplete' data. Trends Ecol Evol, 29, 546-7. Guisan A, Tingley R, Baumgartner JB, Naujokaitis-Lewis I, Sutcliffe PR et al. (2013) Predicting species distributions for conservation decisions. Ecol Lett 16(12): 1424-1435. Guisan A. & Thuiller W. (2005). Predicting species distribution: offering more than simple habitat models. Ecology Letters, 8, 993-1009. Handmer. J.; Proudley, B. (2007). “Communicating uncertainty via probabilities: The case of weather forecasts”. Environmental Hazards 7:79–87. Hansen M.C., Potapov P.V., Moore R., Hancher M., Turubanova S.A., Tyukavina A., Thau D., Stehman S.V., Goetz S.J., Loveland T.R., Kommareddy A., Egorov A., Chini L., Justice C.O. & Townshend J.R. (2013). Highresolution global maps of 21st-century forest cover change. Science, 342, 850-3. Hartig F., Dyke J., Hickler T., Higgins S.I., O'Hara R.B., Scheiter S. & Huth A. (2012). Connecting dynamic vegetation models to data - an inverse perspective. Journal of Biogeography, 39, 2240-2252. Janssen, P.H.M.; Petersen, A.C.; van der Sluijs, J.P.; Risbey, J.S.; Ravetz, J.R. (2005). “A guidance for assessing and communicating uncertainties”. Water Science & Technology 52 (6):125–131. Kahneman, D. (2011). “Thinking, Fast and Slow”. Macmillan. Kahneman, D., Slovic, P., & Tversky, A. (Eds.) (1982). “Judgment under Uncertainty: Heuristics and Biases”. Cambridge University Press. New York. Kahneman, D.; Tversky, A. (1979). “Prospect Theory: An Analysis of Decision under Risk”. Econometrica 47 (2): 263. Keith D.A., Akçakaya H.R., Thuiller W., Midgley G.F., Pearson R.G., Phillips S.J., Regan H.M., Araújo M.B. & Rebelo T.G. (2008). Predicting extinction risks under climate change: coupling stochastic population models with dynamic bioclimatic habitat models. Biology Letters, 4, 560–563. Keith D.A., Akçakaya H.R., Thuiller W., Midgley G.F., Pearson R.G., Phillips S.J., Regan H.M., Araújo M.B. & Rebelo T.G. (2008). Predicting extinction risks under climate change: coupling stochastic population models with dynamic bioclimatic habitat models. Biology Letters, 4, 560–563. Kerr, J. 2007. Watershed management: lessons from common property theory. International Journal of the Commons 1, 89-109. Kok, K., van Vliet, M., Dubel, A., Sendzimir, J. & Barlund, I. (2011) Combining participative backcasting and exploratory scenario development: Experiences from the SCENES project. Technological Forecasting and Social Change 78: 835–851. Klapwijk, C. J., v. Wijk, M. T., Rosenstock, T. S., v. Asten, P. J. A., Thornton, P. K., Giller and K. E. (2014). "Analysis of trade-offs in agricultural systems: current status and way forward." Current Opinion in Environmental Sustainability 6(0): 110-115. Klir, George J.; Yuan Bo (1995). “Fuzzy Sets and Fuzzy Logic. Theory and Applications”. Prentice Hall. Kloprogge, P.; van der Sluijs, J.; Wardekker, A. (2007). “Uncertainty Communication: Issues and good practice”. Copernicus Institute for Sustainable Development and Innovation. Utrecht University. The Netherlands. Kogan, F., Powell, A., Fedorov, O. (2010). Use of Satellite and In-Situ Data to Improve Sustainability. Springer. Kurowicka D. & Cooke R. (2006). Uncertainty Analysis with High Dimensional Dependence Modelling, John Wiley & Sons. Leadley, P., Proenca, V., Fernandez-Manjarres, J., Pereira, H. M., Alkemade, R., Biggs, R., Bruley, E., Cheung, W., Cooper, D., Figueiredo, J., Gilman, E., Guenette, S., Hurtt, G., Mbow, C., Oberdorff, T., Revenga, C., Scharlemann, J. P. W., Scholes, R., Smith, M. S., Sumaila, U. R., Walpole and M. (2014). "Interacting RegionalScale Regime Shifts for Biodiversity and Ecosystem Services." BioScience 64(8): 665-679. Leeuw, S. E. v. d. (2004). "Why Model?" Cybernetics and Systems 35: 117-128. Leidenberger, S., De Giovanni, R., Kulawik, R., Williams, A. R., & Bourlat, S. J. (2014). Mapping present and future potential distribution patterns for a meso‐grazer guild in the Baltic Sea. Journal of Biogeography.
Page 29 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
50
IPBES Deliverable 3c
DO NOT CITE
Lenton, T. M., Held, Hermann, Kriegler, Elmar, Hall, J. W., Lucht, Wolfgang, Rahmstorf, Stefan, Schellnhuber and H. Joachim (2008). "Tipping elements in the Earth's climate system." Proceedings of the National Academy of Sciences 105(6): 1786-1793. Levy, P.E., Cannell, M.G.R., Friend and A.D. (2004). "Modelling the impact of future changes in climate, CO 2 concentration and land use on natural ecosystems and the terrestrial carbon sink." Global Environmental Change (14): 21-30. Li, Y., Chen, J. & Feng, L. (2013) Dealing with Uncertainty: A Survey of Theories and Practices. IEEE Transactions on Knowledge and Data Engineering 25: 2463–2482. Lindenmayer D.B., Piggott M.P. & Wintle B.A. (2013). Counting the books while the library burns: why conservation monitoring programs need a plan for action. Front Ecol Envion, 11, 549-555. Loh et al 2005. Phil. Trans. R. Soc. B.360: 289-295. Lontzek, T. S. and D. Narita (2010). Climate change mitigation and ecosystem services : A stochastic analysis. Kiel Working Paper No. 1593, Kiel Institute for the World Economy, Hindenburgufer 66, 24105 Kiel, Germany: 22 pp. Mace G.M., Collar N.J., Gaston K.J., Hilton-Taylor C., Akçakaya H.R., Leader-Williams N., Milner-Gulland E.J. & Stuart S.N. (2008). Quantification of extinction risk: IUCN's system for classifying threatened species. Conservation Biology, 22, 1424-1442. Mace, G.M., Norris, K. & Fitter, A.H. (2012) Biodiversity and ecosystem services: a multilayered relationship. Trends in Ecology & Evolution 27: 19–26. MAES. (2014) Mapping and Assessment of Ecosystems and their Services: Indicators for ecosystem assessments under Action 5 of the EU Biodiversity Strategy 2020. European Commission. Martin et al 2012 Frontiers in Ecology….. McCarthy, N., L. Lipper, W. Mann, G. Branca and J. Capaldo (2012). Evaluating synergies and trade-offs among food security, development and climate change. Climate Change Mitigation and Agriculture. E. Wollenberg, A. Nihart, M.-L. Tapio-Boström and M. Grieg-Gran. London-New York, ICRAF-CIAT: 39-49. Metzger et al 2013. Ecol. Ind. 33: 26-35 Michener and Jones 2012. Ecoinformatics: supporting ecology as a data-intensive science. Trends in Ecology and Evolution, Vol. 27, No. 2 Michener W.K., Allard S., Budden A., Cook R.B., Douglass K., Frame M., Kelling S., Koskela R., Tenopir C. & Vieglais D.A. (2012). Participatory design of DataONE—Enabling cyberinfrastructure for the biological and environmental sciences. Ecological Informatics, 11, 5-15. Molloy J.C. (2011). The Open Knowledge Foundation: open data means better science. PLoS Biol, 9, e1001195. Morisette, J. T., Jarnevich, C. S., Holcombe, T. R., Talbert, C. B., Ignizio, D., Talbert, M. K., ... & Young, N. E. (2013). VisTrails SAHM: visualization and workflow management for species habitat modeling. Ecography, 36(2), 129-135. Moritz C., Patton J.L., Conroy C.J., Parra J.L., White G.C. & Beissinger S.R. (2008). Impact of a century of climate change on small-mammal communities in Yosemite National Park, USA. Science, 322, 261-4. Morozov A. & Poggiale J.C. (2012). From spatially explicit ecological models to mean-field dynamics: The state of the art and perspectives. Ecological Complexity 10, 1-11. Nicholson, E., Collen, B., Barausse, A., Blanchard, J.L., Costelloe, B.T., Sullivan, K.M.E., Underwood, F.M., Burn, R.W., Fritz, S., Jones, J.P.G., McRae, L., Possingham, H.P. & Milner-Gulland, E.J. (2012) Making Robust Policy Decisions Using Global Biodiversity Indicators. Plos One 7. Noss, R.F. (1990) Indicators for monitoring biodiversity: a hierarchical approach. Conservation biology 4: 355–364. Ostrom, E., 2009. A general framework for analyzing sustainability of social-ecological systems. Science 325, 419-422. Pearson R.G., Stanton J.C., Shoemaker K.T., Aiello-Lammens M.E., Ersts P.J., Horning N., Fordham D.A., Raxworthy C.J., Ryu H.Y., McNees J. & Akcakaya H.R. (2014). Life history and spatial traits predict extinction risk due to climate change. Nature Climate Change, 4, 217-221.
Page 30 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Pennington, D. D., Higgins, D., Peterson, A. T., Jones, M. B., Ludäscher, B., & Bowers, S. (2007). Ecological niche modeling using the Kepler workflow system. In Workflows for e-Science (pp. 91-108). Springer London. Penone C., Davidson A.D., Shoemaker K.T., Di Marco M., Rondinini C., Brooks T.M., Young B.E., Graham C.H. & Costa G.C. (2014). Imputation of missing data in life-history trait datasets: which approach performs the best? Methods in Ecology and Evolution, 5, 961-970. Pereira, H.M., Leadley, P.W., Proenca, V., Alkemade, R., Scharlemann, J.P.W., Fernandez-Manjarres, J.F., Araujo, M.B., Balvanera, P., Biggs, R., Cheung, W.W.L., Chini, L., Cooper, H.D., Gilman, E.L., Guenette, S., Hurtt, G.C., Huntington, H.P., Mace, G.M., Oberdorff, T., Revenga, C., Rodrigues, P., Scholes, R.J., Sumaila, U.R. & Walpole, M. (2010) Scenarios for Global Biodiversity in the 21st Century. Science 330: 1496–1502. Pereira, H. M., Navarro, L. M., & Martins, I. S. (2012). Global biodiversity change: the bad, the good, and the unknown. Ann Rev Environ Resour, 37: 25-50. Pereira, H.M., Ferrier, S., Walters, M., Geller, G.N., Jongman, R.H.G., Scholes, R.J., Bruford, M.W., Brummitt, N., Butchart, S.H.M., Cardoso, A.C., Coops, N.C., Dulloo, E., Faith, D.P., Freyhof, J., Gregory, R.D., Heip, C., Höft, R., Hurtt, G., Jetz, W., Karp, D., McGeoch, M.A., Obura, D., Onoda, Y., Pettorelli, N., Reyers, B., Sayre, R., Scharlemann, J.P.W., Stuart, S.N., Turak, E., Walpole, M. & Wegmann, M. (2013) Essential Biodiversity Variables. Science 339: 277–278. Pettorelli, N., Laurance, W.F., O’Brien, T.G., Wegmann, M., Nagendra, H. & Turner, W. (2014) Satellite remote sensing for applied ecologists: opportunities and challenges. Journal of Applied Ecology 51: 839–848. Pidgeon, N. ; Fischhoff, B. (2011). “The role of social and decision sciences in communicating uncertain climate risks”. Nature Climate Change 1:37-41. DOI:10.1038/nclimate1080. Pimm S.L., Jenkins C.N., Abell R., Brooks T.M., Gittleman J.L., Joppa L.N., Raven P.H., Roberts C.M. & Sexton J.O. (2014). The biodiversity of species and their rates of extinction, distribution, and protection. Science, 344, 987-+. Porter, J.H. et al. (2009) New eyes on the world: advanced sensors for ecology. Bioscience 59, 385–397 Prowse, T.A.A., C.N. Johnson, R.C. Lacy, C.J.A. Bradshaw, J.P. Pollak, M.J. Watts, and B.W. Brook. 2013. No need for disease: testing extinction hypotheses for the thylacine using multi-species metamodels. Journal of Animal Ecology 82:355-364. Pujol G., Looss B. & Janon A. (2013). sensitivity: sensitivity Analysis. R package version 1.7, http://CRAN.R-project.org/package=sensitivity. Recknagel, F. (2006). “Ecological Informatics”. Springer-Verlag. Berlin Heidelberg. Regan H.M., Colyvan M. & Burgman M.A. (2002). A taxonomy and treatment of uncertainty for ecology and conservation biology. Ecological Applications, 12, 618-628. Reichman O.J., Jones M.B. & Schildhauer M.P. (2011). Challenges and opportunities of open data in ecology. Science, 331, 703-5. Reid, W. V., Berkes, F. , Wilbanks,T. and Capistrano, D. eds. 2006. Bridging scales and knowledge systems: Concepts and applications in ecosystem assessment. Washington, DC: Island Press, for World Resources Institute. Rindfuss, R.R., Walsh, S.J., Turner, B.L., Fox, J., Mishra and V. (2004). "Developing a science of land change: Challenges and methodological issues " PNAS 101(39): 13976-13981. Ruegg J., Gries C., Bond-Lamberty B., Bowen G.J., Felzer B.S., McIntyre N.E., Soranno P.A., Vanderbilt K.L. & Weathers K.C. (2014). Completing the data life cycle: using information management in macrosystems ecology research. Front Ecol Envion, 12, 24-30. Rundel, P.W. et al. (2009) Environmental sensor networks in ecological research. New Phytol. 182, 589– 607 Rykiel E.J. (1996). Testing ecological models: The meaning of validation. Ecological Modelling, 90, 229244. Sabatier, P. A., Focht, W., Lubell, M., Trachtenberg, Z., Vedlitz, A. and Matlock M. 2005. Swimming upstream: collaborative approaches to watershed management.
Page 31 of 33
First Order Draft
5
10
15
20
25
30
35
40
45
IPBES Deliverable 3c
DO NOT CITE
Sarukhán J, Urquiza-Haas T, Koleff P, Carabias J, Dirzo R et al. (2014) Strategic actions to value, conserve, and restore the natural capital of megadiversity countries: the case of Mexico. BioScience. Scheffer M., Bascompte J., Brock W.A., Brovkin V., Carpenter S.R., Dakos V., Held H., van Nes E.H., Rietkerk M. & Sugihara G. (2009). Early-warning signals for critical transitions. Nature, 461, 53-59. Scheffer M., Carpenter S.R., Lenton T.M., Bascompte J., Brock W., Dakos V., van de Koppel J., van de Leemput I.A., Levin S.A., van Nes E.H., Pascual M. & Vandermeer J. (2012). Anticipating critical transitions. Science, 338, 344-348. Scheffer, M., Carpenter and S. R. (2003). "Catastrophic regime shifts in ecosystems: linking theory to observation." Trends in Ecology & Evolution 18(12): 648-656. Scheffer, Marten, Carpenter, Steve, Foley, J. A., Folke, Carl, Walker and Brian (2001). "Catastrophic shifts in ecosystems." Nature 413(6856): 591-596. Scholes et al. 2012. Current Opinion in Environmental Sustainability, 4:139–146 Shoemaker K.T. & Akçakaya H.R. (2014). Inferring the nature of anthropogenic threats from long-term abundance records. Conservation Biology. DOI: 10.1111/cobi.12353 Shoemaker K.T., Lacy R.C., Verant M.L., Brook B.W., Livieri T.M., Miller P.S., Fordham D.A. & Akçakaya H.R. (2014). Effects of prey metapopulation structure on the viability of black-footed ferrets in plagueimpacted landscapes: a metamodelling approach. Journal of Applied Ecology, 51, 735-745. Shrestha N.K.; Leta O.T.; De Fraine B.; van Griensven A.; Bauwens W. (2013): OpenMI based integrated sediment transport modelling of the river Zenne, Belgium. Environmental Modelling & Software 47(0), pp. 193206 Silvertown, J. (2009). A new dawn for citizen science. Trends in Ecology & Evolution, 24(9), 467-471. Smith, P., H. Haberl, A. Popp, K. Erb, C. Lauk, R. Harper, F. Tubiello, A. d. S. Pinto, M. Jafari, S. Sohi, O. Masera, H. Böttcher, G. Berndes, M. Bustamante, H. Ahammad, H. Clark, H. Dong, E. A. Elsiddig, C. Mbow, N. H. Ravindranath, C. W. Rice, C. Robledo-Abad, A. Romanovskaya, F. Sperling, M. Herrero, J. I. House and S. Rose (2013). "How much land based greenhouse gas mitigation can be achieved without compromising food security and environmental goals?" Global Change Biology Review(DOI.10.1111/gcb.12160). Stanton J.C., Pearson R.G., Horning N., Ersts P. & Akcakaya H.R. (2012). Combining static and dynamic variables in species distribution models under climate change. Methods in Ecology and Evolution, 3, 349-357 Stanton J.C., Shoemaker K.T., Pearson R.G. & Akçakaya H.R. (2014). Warning times for species extinctions due to climate change. Global Change Biology, DOI: 10.1111/gcb.12721. Sugihara G., May R., Ye H., Hsieh C.-h., Deyle E., Fogarty M. & Munch S. (2012). Detecting Causality in Complex Ecosystems. Science, 338, 496-500. Tallis, H., Mooney, H., Andelman, S., Balvanera, P., Cramer, W., Karp, D., Polasky, S., Reyers, B., Ricketts, T., Running, S., Thonicke, K., Tietjen, B. & Walz, A. (2012) A Global System for Monitoring Ecosystem Service Change. BioScience 62: 977–986. Tang S., Pawar S. & Allesina S. (2014). Correlation between interaction strengths drives stability in large ecological networks. Ecology Letters, 17, 1094-1100. Ten Brink, B. (2010) Rethinking Global Biodiversity Strategies. Netherlands Environmental Agency (PBL), The Hague. Thuiller, W., Münkemüller, T., Lavergne, S., Mouillot, D., Mouquet, N., Schiffers, K. & Gravel, D. (2013) A road map for integrating eco-evolutionary processes into biodiversity models. Ecology Letters, 16, 94-105. Tittensor, D.P., Walpole, M., Hill, S.L.L., Boyce, D.G., Britten, G.L., Burgess, N.D., Butchart, S.H.M., Leadley, P.W., Regan, E.C., Alkemade, R., Baumung, R., Bellard, C., Bouwman, L., Bowles-Newark, N.J., Chenery, A.M., Cheung, W.W.L., Christensen, V., Cooper, H.D., Crowther, A.R., Dixon, M.J.R., Galli, A., Gaveau, V., Gregory, R.D., Gutierrez, N.L., Hirsch, T.L., Höft, R., Januchowski-Hartley, S.R., Karmann, M., Krug, C.B., Leverington, F.J., Loh, J., Lojenga, R.K., Malsch, K., Marques, A., Morgan, D.H.W., Mumby, P.J., Newbold, T., Noonan-Mooney, K., Pagad, S.N., Parks, B.C., Pereira, H.M., Robertson, T., Rondinini, C., Santini, L., Scharlemann, J.P.W., Schindler, S., Sumaila, U.R., Teh, L.S.L., Kolck, J. van, Visconti, P. & Ye, Y. (2014) A midterm analysis of progress toward international biodiversity targets. Science 346: 241–244.
Page 32 of 33
First Order Draft
5
10
15
20
25
30
35
IPBES Deliverable 3c
DO NOT CITE
Turner, W. et al. (2015). Free and open-access satellite data are key to biodiversity conservation, Biological Conservation 182: 173-176. van der Hoven S, Chabason L (2009) The debate on an Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES). 1-25 Van der Putten, W.H., Macel, M. & Visser, M.E. (2010) Predicting species distribution and abundance responses to climate change: why it is essential to include biotic interactions across trophic levels. Philosophical Transactions of the Royal Society B-Biological Sciences, 365, 2025-2034. Van Strien, A.J., Soldaat, L.L. & Gregory, R.D. (2012) Desirable mathematical properties of indicators for biodiversity change. Ecological Indicators 14: 202–208. Vuuren, D. P. v., M. Isaac, Z. W. Kundzewicz, N. Arnell, T. Barker, P. Criqui, F. Berkhout, H. Hilderink, J. Hinkel, A. Hof, A. Kitous, T. Kram, R. Mechler and S. Scrieciu (2011). "The use of scenarios as the basis for combined assessment of climate change mitigation and adaptation." Global Environmental Change 21: 575– 591. Vuuren, v., D. P., Bayer, L. B., Chuwah, C., Ganzeveld, L., Hazeleger, W., v. d. Hurk, B., v. Noije, T., O'Neill, B., Strengers and B. J. (2012). "A comprehensive view on climate change: coupling of earth system and integrated assessment models." Environmental Research Letters 7(2). Waage, S. & Kester, C. (2014). Global Public Sector Trends in Ecosystem Services: 2009–2013. Business for Social Responsibility,San Francisco, CA, USA. Watson J.E.M. (2014). Human responses to climate change will seriously impact biodiversity conservation: it's time we start planning for them. Conservation Letters, 7, 1-2. Willerslev E., Davison J., Moora M., Zobel M., Coissac E., Edwards M.E., Lorenzen E.D., Vestergard M., Gussarova G., Haile J., Craine J., Gielly L., Boessenkool S., Epp L.S., Pearman P.B., Cheddadi R., Murray D., Brathen K.A., Yoccoz N., Binney H., Cruaud C., Wincker P., Goslar T., Alsos I.G., Bellemain E., Brysting A.K., Elven R., Sonstebo J.H., Murton J., Sher A., Rasmussen M., Ronn R., Mourier T., Cooper A., Austin J., Moller P., Froese D., Zazula G., Pompanon F., Rioux D., Niderkorn V., Tikhonov A., Savvinov G., Roberts R.G., MacPhee R.D.E., Gilbert M.T.P., Kjaer K.H., Orlando L., Brochmann C. & Taberlet P. (2014). Fifty thousand years of Arctic vegetation and megafaunal diet. Nature, 506, 47-51. Wolf, N. & Mangel, M. 2008. Multiple hypothesis testing and the declining-population paradigm in Steller sea lions. Ecological Applications, 18, 1932–1955. Wong, C., Jiang, B. Kinzig,A. P., Lee K. N. & Ouyang Z. (2015). Linking ecosystem characteristics to final ecosystem services for public policy. Ecology Letters,18: 108-118 Zimmermann, N.E., Yoccoz, N.G., Edwards, J., T.C., Meier, E.C., Thuiller, W., Guisan, A., Schmatz, D.R. & Pearman, P.B. (2009) Climatic extremes improve predictions of spatial patterns of tree species. Proceedings of the National Academy of Sciences, USA, 106, 19723–19728. Zurell D., Berger U., Cabral J.S., Jeltsch F., Meynard C.N., Munkemuller T., Nehrbass N., Pagel J., Reineking B., Schroder B. & Grimm V. (2010). The virtual ecologist approach: simulating data and observers. Oikos, 119, 622-635.
Page 33 of 33
