sub-global scenarios · sub-global assessments differed greatly from the ma global scenarios....

Chapter 10

Sub-global Scenarios

Coordinating Lead Authors: Louis Lebel, Pongmanee Thongbai, Kasper KokLead Authors: John B. R. Agard, Elena Bennett, Reinette Biggs, Margarida Ferreira, Colin Filer, Yogesh

Gokhale, William Mala, Chuck Rumsey, Sandra J. Velarde, Monika ZurekContributing Authors: Hernan Blanco, Tim Lynam, Yue TianxiangReview Editors: Richard Moles, Fran Monks, Bernadette O’Regan

Main Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23110.1.1 Scenarios: Definition and Purpose10.1.2 Scenarios in the MA Global Assessment10.1.3 Scenarios in the MA Sub-global Assessments10.1.4 Assessing Work in Progress10.1.5 Analytical Framework and Organization of Chapter

10.2 Handling Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23610.2.1 Key Uncertainties Identified in Sub-global Assessments10.2.2 Describing Key Uncertainties after Identification10.2.3 Relationships of Uncertainty to Ecosystem Services and Human Well-

being

10.3 Telling Plausible Stories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23910.3.1 Diversity of Approaches10.3.2 Frameworks Used to Develop Storylines10.3.3 Systems Models, Quantification, and Spatial Explicitness

10.4 Incorporating Ecology into Scenarios . . . . . . . . . . . . . . . . . . . . . . . . 24410.4.1 Ecological Detail in Scenarios10.4.2 Ecological Interactions and Surprises

10.5 Dealing with Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24510.5.1 Multi- and Cross-scale Processes in Scenarios10.5.2 Spatial Extent, Heterogeneity, and Resolution10.5.3 Time Horizons10.5.4 The Linkage between Sub-global and Global Scenarios

10.6 Participating in Scenario Development . . . . . . . . . . . . . . . . . . . . . . . 24810.6.1 Goals of Participation10.6.2 Selection of Participants10.6.3 Mechanisms of Participation10.6.4 Roles of Participants10.6.5 Problems Encountered in Participation

10.7 Communication in Sub-global Scenarios . . . . . . . . . . . . . . . . . . . . . . 25010.7.1 Communication Strategies10.7.2 Scenarios in Policy Dialogues10.7.3 Storylines for a Wider Audience

PAGE 229

229

................. 11474$ CH10 10-17-05 15:45:41 PS

230 Ecosystems and Human Well-being: Sub-global

10.7.4 Spatial Representations of Scenarios10.7.5 Community Theater and Video Techniques

10.8 Findings of the Sub-Global Scenarios . . . . . . . . . . . . . . . . . . . . . . . . 25210.8.1 Commonalties among Scenario Findings10.8.2 Comparing Findings across Scales10.8.3 Summing Up: The Importance of Stakeholders

10.9 Conclusions and Recommendations for Future Assessments . . . . . . 25510.9.1 Strength in Diversity10.9.2 Lessons Learned

APPENDIX

10.1 Summary of Global Scenarios in the Millennium EcosystemAssessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258

BOXES

10.1 Developing Storylines: The Caribbean Sea Approach

10.2 Nested, Multiscale Design: SAfMA and Portugal

10.3 Obstacles to Communication: Papua New Guinea

FIGURES

10.1 Key Questions Asked about the Design and Implementationof Scenario Exercises in Sub-global Assessments andSelected Answers to Four of These Questions

10.2 Two Sets of Uncertainty Axes That Could Be Used toDescribe the Northern Highland Lake District, Wisconsin,Scenarios

10.3 Cascading Uncertainties in Social-ecological Systems ThatMight Be Captured in Scenarios

10.4 The Southern Africa Regional Assessment ScenarioDevelopment Approach

10.5 The Use of Soft-models as an Intermediate Step in the Forestand Agroecosystem Trade-offs in the Humid Tropics (TropicalForest Margins) Scenario Exercise

PAGE 230

10.6 Example of Spatially Explicit Scenario Outputs in theSouthern Africa Regional Assessment

TABLES

10.1 Summary of the Scenario Exercises in Selected Sub-globalAssessments

10.2 Major Exogenous and Endogenous Uncertainties across theSub-global Scenarios

10.3 Steps in the Scenario Development Process

10.4 Reasons Given by the Sub-global Assessments for Excludingor Including Multiscale Considerations in their ScenarioExercises: Selected Examples

10.5 Reasons for Seeking Participation in Sub-global AssessmentScenario Development Exercises

10.6 Key Ecosystem Services Addressed in the Scenarios ofVarious Sub-global Assessments

10.7 Cross-mapping of SAfMA, Caribbean Sea, and PortugalScenarios onto the MA Global Scenarios

................. 11474$ CH10 10-17-05 15:45:41 PS

231Sub-global Scenarios

Main Messages

Scenarios involve thinking about a wide range of futures, including bothwell-known trends and uncertainties. Developing scenarios is a fundamentalprerequisite of strategic thinking and planning. Scenarios facilitate strategy for-mulation and evaluation, improve understanding of the uncertainties inherentin ecosystems, and test the robustness of particular strategies against a set ofplausible futures. Unlike other decision-making techniques such as prediction,forecasting, and other single future outlooks, scenario-building is a cognitiveand imaginative mechanism for decision-making. It uses more holistic, inte-grated, and participatory approaches to aid understanding of the intrinsic heter-ogeneity and uncertainty of ecosystem management. It also extends predictionand forecasting methods to provide additional and relevant alternatives to helpdecision-makers think, talk, plan and act imaginatively in pursuit of a moresustainable society.

Sub-global assessments used scenarios for multiple purposes, whichoften extended beyond the rationale for scenarios developed at theglobal level. Besides being used as a tool for decision-makers to plan for thefuture, many sub-global assessments, such as Southern Africa and the North-ern Highland Lakes District of Wisconsin, also used scenarios as a means forcommunicating possible future changes and major uncertainties to stakehold-ers. In the assessments of San Pedro de Atacama, Chile, and Bajo Chirripo,Costa Rica, scenarios also have proved to be an important tool for acquiringdata about stakeholder preferences, perceptions, and values. In a few cases,including Wisconsin, Caribbean Sea, and SAfMA, scenarios had a role in de-fining the boundaries within which discussions about management and policyoptions relevant to ecosystem services and human well-being could be held.

Despite being based on the MA conceptual framework, scenarios in thesub-global assessments differed greatly from the MA global scenarios.Significant differences between global and sub-global assessments in terms ofkey uncertainties, stakeholders involved, and scales of analysis, resulted insets of sub-global scenarios that were different from the global scenarios.Nonetheless, a substantive link was maintained between global and sub-globalscenarios in the case of the SAfMA, Caribbean Sea, and Portugal assess-ments.

Most sub-global assessments limited the discussions on which scenar-ios to develop around only one or two key uncertainties. Uncertaintieswere typically related to issues of technology, markets, and economic develop-ment; over half of the scenarios identified institutional arrangements/gover-nance as a key uncertainty. Some unique examples of uncertainties includedHIV/AIDS for Papua New Guinea, mining in San Pedro de Atacama, and thelocal legal system in India Local. Somewhat surprisingly, the Mae Chaem,Thailand, component of the Tropical Forest Margins assessment was the onlyassessment to explicitly address uncertainties surrounding ecosystem feed-backs.

Qualitative rather than quantitative models were most often used to ex-plore interactions between major processes and structures, in order toprovide a framework within which scenario storylines were developed.There is little documented explanation of the methods by which narrative story-lines were developed in most sub-global scenarios. Many sub-global assess-ments noted the desire to quantify storylines, but time constraints and the lackof available models or expertise prevented all but Western China, TropicalForest Margins (Mae Chaem, Thailand component), and SAfMA Regional as-sessments from undertaking such analyses.

Important scientific advances have been made in constructing nestedscenarios at multiple scales. To meet the objectives of decision-makers and

PAGE 231

stakeholder groups with interests at different scales, the SAfMA assessmentundertook five local-scale assessments that were nested within two basin-scale assessments, which were in turn nested within a regional-scale assess-ment. The Portugal assessment was also undertaken at three scales—local,basin, and national—though the local case studies were not nested within thebasins that were assessed.

Scenario-building is an important method to involve stakeholders in pol-icy formulation, and to encourage citizens to adopt policies aimed atenvironmental protection. The relevance, significance, and influence of thescenarios that are constructed will ultimately depend on who was involved intheir development. Decision-makers will have difficulty introducing new policiesdesigned to alter behaviors without the support of the general population. Parti-cipants in scenario building can provide essential input on the relevance ofstorylines being developed, and the nature of uncertainties that are importantat sub-global scales.

The scale and context of a sub-global assessment are primary considera-tions when selecting media for communicating the findings of scenarios.Specific contextual factors include the size of the audience, their level of liter-acy or formal education, their religious and cultural beliefs, and the cost ofreaching the audience given available resources. Specific contexts have gener-ated some unique and creative solutions including the use of theater in SAfMALocal Livelihoods and cartoon animation in Wisconsin.

Identification of winners and losers in each scenario is an important stepin guiding future responses. The inclusion of stakeholders in the scenariodevelopment and validation process helps make explicit the circumstancesunder which winners and losers emerge. Sub-global scenarios highlighted theimportance of scale in determining winners and losers. In the Portugal assess-ment, one local scenario was characterized by the abandonment of agriculturalfields and rural–urban migration, which is undesirable to local policy-makers.However, this scenario could be nested within the MA’s Global Orchestrationscenario (developed as a part of the MA’s global assessment), which is char-acterized by economic growth and viewed by the policy-makers at higher levelsas having the highest net benefits for human well-being.

Future scenario activities need to pay even greater attention to ecosys-tem processes. Past and current scenario work has emphasized human activ-ities as the main drivers of change in the availability of ecosystem goods andservices, without much reflection on the implications of ecological feedbacksfor ecosystems and human well-being.

The large and growing number of sub-global assessment scenarios is aunique source of information within the multiscale assessment contextof the MA. The sub-global scenarios, employing the common conceptualframework of the MA, allow for the critical evaluation of local variation. It seemslikely that the incorporation of the findings from sub-global scenarios into theMA’s global assessment would have been valuable. However, because of tim-ing issues, the global MA scenarios did not have the opportunity to incorporatethe findings of the sub-global assessments.

10.1 IntroductionThe MA scenarios, unlike some earlier scenario efforts,were developed to integrate ecology into their design ex-plicitly (Bennett et al. 2003; Peterson et al. 2003a). Ecosys-tems are not treated solely as passive recipients of impactsresulting from changes driven by socioeconomic systems,but are understood to play an active role in jointly deter-mining the futures of humans and ecosystems. Changes in

................. 11474$ CH10 10-17-05 15:45:42 PS


the flow of ecosystem services are seen as having the poten-tial to alter future development pathways. This is a moreintegrated view of how human–environment systems un-fold over time than is typically assumed in scenario exerciseswhere the goal is to assess environmental changes. (See MAScenarios, Chapters 2 and 3.)

The purpose of this chapter is two-fold. First, the chap-ter aims to critically review how scenarios were used in theMA sub-global assessments. For those interested in con-ducting or designing their own sub-global assessments, itpoints to both the limitations and strengths of the differentapproaches taken. Second, the chapter aims to draw somemore general conclusions, for a broader audience, on howfuture assessments involving scenario exercises can incorpo-rate ecological considerations.

10.1.1 Scenarios: Definition and Purpose

A scenario is a story that offers an internally consistent andplausible explanation of how events unfold over time (Gal-lopın et al. 1997; Raskin et al. 2002). Sequences of eventsand interactions, rather than specific time scales, are usuallyemphasized. Decision-makers in the business communityand elsewhere have employed scenarios for several decadesas an approach to aid decision-making in the face of uncer-tainty (Chermack et al. 2001; Coates 2000; Davis 1998).Scenarios are generally useful for encouraging systematicplanning in situations of uncertainty (van der Heijden 1996)or for revealing a range of dynamic processes and causalchains that lead to alternative outcomes (Rotmans et al.2000). The intention of scenario planning is to consider avariety of possible futures that include important uncertain-ties, rather than to focus on the accurate prediction of asingle possible future (van der Heijden 1996, Peterson et al.2003b). Usually, scenario planners start by determining a setof focal questions or issues in conjunction with their pri-mary stakeholders. This is followed by an assessment of thecurrent state of a system, and identification of alternativepathways that the system might take (Peterson et al. 2003b).The focal questions often revolve around key uncertaintiesor unknowns in the system. The next step is to build story-lines by projecting these questions into the future, whichcan be done either qualitatively or quantitatively. For com-plex systems, various methods can be used in an iterativeprocess (Alcamo et al. 1998).

Scenarios were first used after World War II as a methodfor war game analysis (van der Heijden 1996). Their valuewas quickly recognized by Herman Kahn (Kahn andWeiner 1967) and others who developed the use of scenar-ios for other strategic planning applications. Scenarios wererefined at Royal Dutch/Shell by Pierre Wack in the 1970sand 1980s, and Shell became a leader in the scenario ap-proach to business planning. Today scenario development isused in a variety of different contexts ranging from politicaldecision-making (Kahane 1992, 1998), to business planning(Wack 1985; Schwartz 1996; Davis 1998), to local commu-nity management (Wollenberg et al. 2000; Peterson et al.2003b) and understanding global-scale environmental pat-terns and processes (Gallopın et al. 1997; Cosgrove and

PAGE 232

Rijsberman 2000; IPCC 2001; UNEP 2002; van Notten etal. 2003).

For ecosystem assessments, scenarios are seen as a funda-mental prerequisite for strategic thinking and planning. Inthis context, scenarios are used to facilitate strategy formula-tion and evaluation, develop an understanding of the uncer-tainty inherent in ecological systems, and test the robustnessof response strategies against a set of possible futures. Unlikeother decision-making techniques that focus on a single fu-ture outlook (such as prediction or forecasting), scenario-building is a cognitive and imaginative decision-makingtool. It emphasizes holistic, integrated, and participatoryapproaches to illuminate the heterogeneity and ambiguityinherent in ecosystem management. Scenarios extend pre-diction and forecasting methods, providing additional andrelevant alternatives to help decision-makers think, talk,plan, and act imaginatively in pursuit of a more sustainablesociety.

10.1.2 Scenarios in the MA Global Assessment

Scenarios are defined in the MA conceptual framework asplausible alternative futures, each an example of what mighthappen under a particular set of assumptions (MA 2003).Scenarios were a key tool in the overall MA process that,along with analysis of condition and trends, and responses,provided a comprehensive assessment of ecosystems andhuman well-being at the global level. The MA conceptualframework envisions scenarios serving as a tool in threeways: (1) to educate local stakeholders on possible futurechanges in ecosystem services and human well-being; (2) tocommunicate the overall results of an assessment to abroader audience; and (3) to facilitate decision-making atglobal and sub-global scales. In the MA global assessment,four global scenarios were developed. (See Appendix 10.1).These differed primarily in terms of the assumptions madeabout the drivers of change in ecosystem services, and howsociety reacts to such change.

In the global assessments, model and scenario develop-ment consider multiscale processes and heterogeneity bydisaggregating the globe into several multi-country regions(Alcamo et al. 1998; Nakicenovic et al. 2000). However,this top-down framework falls far short of embracingsmaller scale phenomenon in an interactive, multiscale way.

10.1.3 Scenarios in the MA Sub-global Assessments

The sub-global assessments were strongly encouraged toundertake scenario analyses as part of their assessments.However, sub-global assessments varied greatly in howmuch emphasis they placed on scenarios in their overall as-sessment activities, and in the specific goals and focus ofthe scenario exercises undertaken. (See Table 10.1.) Thisvariation in emphasis led to a wide diversity of outcomesthat were not readily comparable.

Scenarios at the sub-global scale may yield results differ-ent from those that emerge from disaggregated global analy-ses for several reasons. First, the set of ecosystem servicesand the ecosystem changes of interest vary across scales. (SeeMA 2003, Chapter 5.). Second, the mixture of ecosystem

................. 11474$ CH10 10-17-05 15:45:43 PS


Table 10.1. Summary of the Scenario Exercises in Selected Sub-global Assessments (Information from specially designedquestionnaires, Knowledge Markets in KM1, and KM2)

Main Ecosystem Services andSub–Global Stated Goals of Scenario Human Well–being Aspects Main Methods Used to Develop TimeAssessment Analysis Considered Scenarios Spatial Scale Horizon

San Pedro de communication with stake- water, biodiversity, minerals, tour- workshops and expert work localAtacama holders ism, astronomical observation,

agriculture

Caribbean Sea stimulate thinking about the fisheries, tourism workshops and expert work regional 2000–50future sub-regional

Coastal BC food, biodiversity, fiber and tim- workshops and modeling regional –ber, runoff regulation, cultural sub-national

India Local assess influences of exter- food (hunting), firewood, biodiver- based on ‘‘what . . . if’’ questions local –nal forces on local commu- sity for management optionsnity

PNG change ways of thinking logging, coastal mining (heavy assessment and implications of national –about the future metals), position of women, birth the past; expert scenarios provincial

controllocal

community

Portugal for users and decision- food, biodiversity, water, soil, rec- workshops and expert work national 2050makers reation basin

local

SAfMA tool for planning/actions water, food, biodiversity, firewood participatory workshops including regional 2010–50particularly at local scales community theatre (local); model- basin

ing and expert work (basin andlocalregional)

Sweden KW and prepare for surprises, infor- KW: flood buffer, wetland, secur- expert work local 2050Sweden SU mation for planning; obtain ity from floods

stakeholder input SU: green area loss, biodiversity,CO2 sequestration, recreation,health

Northern Range stimulate thinking about the fresh water, forests, cultural, run- focus groups for developing sto- national 2000–2025future off regulation, and biodiversity rylines sub-national

Tropical Forest analyze natural resource biodiversity, hydrological function, expert work (Mae Chaem); partic- local bench- 2020–30Margins management options, future forest cover ipatory scenarios (elsewhere, mark sites

planning; enhance partici- planning) ecoregionpation; inform policy-

nationalmakers

Downstream visualize the future, infor- rice, shrimp, timber, firewood, localMekong mation for policy-makers, medicinal plants, fresh water, se-

input for models curity, social relation, freedom ofchoice

Western China information for the govern- urbanization, deforestation, water quantitative modeling regional 2010–20ment, input for models local

Sinai for local communication water (quality/quantity) workshops—qualitative local 2010–20

Bajo Chirripo get in touch with user needs culture, forest, biodiversity, inland workshops—qualitative local �2010/2020water

Eastern improve response options; forest, food, energy, water, bio- workshops—qualitative and sub-national 2010–15Himalayas inform policy-makers diversity, land use, loss of life, quantitative local

food security

Sao Paulo envision the future; change water, biodiversity, cultural ser- assessment and implication of the local riverbad situations vice, soil, livelihood, social conflict pilot expert scenario basin

(continues)

PAGE 233................. 11474$ CH10 10-17-05 15:45:44 PS


Table 10.1. continued

Main Ecosystem Services andSub–Global Stated Goals of Scenario Human Well–being Aspects Main Methods Used to Develop TimeAssessment Analysis Considered Scenarios Spatial Scale Horizon

India Urban share information with part- food, water, fuel, fiber, western individual consultations; literature local 1950–2050ners culture review national

Wisconsin Improve ecological man- Native American walleye spear initial expert assessment and local 2028agement fishery, tourism and sport fishing, scenario development; participa-

maintenance of the ‘‘Northwoods’’ tory scenarios workshop; scenar-landscape, habitats to increase ios redrafted by expertsresilience of fish population, nutri-ent cycling (reducing impact ofrunoff)

Benchmark Example

Goulburn Broken show broad range of possi- water, crop production, livestock, stakeholders involved in work- localCatchment, Aus- ble management options tourism shops organized as expert wit-tralia (CSIRO) ness jury; modeling

types varies greatly within and among larger regions of theworld, and it is far from certain that their resilience tohuman perturbations and natural disturbances is similaracross these regions. Third, the types of human perturba-tions vary widely within and across regions with, for exam-ple, the level of industrialization of production activities.Finally, interacting drivers, institutional responses, and envi-ronmental challenges are themselves heterogeneous at thesub-global scale and typically operate on shorter timeframes.

The rationale for undertaking scenarios in sub-global as-sessments varied considerably, often depending on the na-ture of stakeholder involvement. A number of sub-globalassessments directly engaged local decision-makers (for ex-ample, Caribbean Sea and India Urban), and scenariostherefore played a role in local decision-making. Anotherset of sub-global assessments focused on local communitiesand thus lent themselves to educational and communicationpurposes (for example, San Pedro de Atacama and Sinai,Egypt); their primary purpose was to start a process of stake-holder involvement in ecosystem management processes. Athird purpose of developing scenarios was to use them asinput into spatially explicit models (for example, WesternChina and Laguna Lake Basin, Philippines).

It is important to note that while the teams conductingsub-global assessments were provided with some method-ological guidelines for developing scenarios, they were en-couraged to develop their own methods where necessary.Consequently, the use of different methods, combined withthe focus on specific groups or issues, resulted in the devel-opment of a wide variety of scenario types.

10.1.4 Assessing Work in Progress

At the time of writing (January 2005), many sub-global as-sessments had yet to complete their assessment work. More-over, scenario analysis is normally one of the final steps inan assessment, as information on drivers, uncertainties, keyecosystem services, and current trends must be available be-

PAGE 234

fore scenario development can start. This is particularly truefor the quantification of scenario results by linking story-lines and models (discussed below). Therefore, this chapteroften reports on plans and intentions rather than on finalconclusions of the scenario exercises. Also, sub-global assess-ments that had completed their scenario activities usuallyindicated that theirs were iterative processes. Consequently,the chapter provides a snapshot in time of what is usually amuch longer analytical process.

To carry out these analyses, the authors were dependenton a combination of written background documents andresponses to queries provided by individual sub-global as-sessments, direct discussions with those involved in specificassessments, and the two Knowledge Markets (described inChapter 2, Box 2.1). Among the authors of this chapter wererepresentatives of 10 sub-global assessments. An examplefrom case studies in the Goulburn Broken Catchment as-sessment in Australia (CSIRO Sustainable Ecosystems2003) was used as a benchmark reference in this chapter.Though not part of the MA, this assessment is completeand was undertaken in a manner consistent with the MAconceptual framework.

10.1.5 Analytical Framework and Organization ofChapter

To help with the analysis in this chapter, an initial set ofquestions (Figure 10.1, left column) was developed and sentto all sub-global assessments. The list of questions was re-fined and additional questions were added in the course ofseveral iterations, with inputs from the sub-global assess-ments. The right column in Figure 10.1 shows a partial,illustrative list of responses from the sub-global assessmentsto those questions, highlighting that scenarios were devel-oped for varying purposes, that diverse approaches wereused to address the technical challenges of dealing withquantification and scale, and that practical constraints oftime, budget, and skills were often faced.

The rest of this chapter analyzes the sub-global assess-ment scenario experiences based on responses to the ques-

................. 11474$ CH10 10-17-05 15:45:45 PS


Figure 10.1. Key Questions Asked about the Design and Implementation of Scenario Exercises in Sub-global Assessments andSelected Answers to Four of These Questions

PAGE 235................. 11474$ CH10 10-17-05 15:45:50 PS


tions along the following dimensions: key uncertainties;methods used to develop storylines; how ecological consid-erations were incorporated into the scenarios; relationshipwith the global scenarios exercise; participation in the sce-nario development process; communication of scenarioprocess and findings; and insights from a comparison of sce-nario results across assessments. The chapter concludes witha summary of lessons drawn from the experiences of thesub-global assessments, which may be of particular value forfuture assessments.

10.2 Handling UncertaintiesScenarios are a useful method for improving decision-makingand stakeholder involvement in situations of high uncer-tainty. When the world is highly predictable, simpler decision-making techniques, such as prediction, forecasting, andother single-future outlooks may be useful (Peterson et al.2003b). However, such simplicity is rare in ecosystem man-agement. Instead, managers often find themselves con-fronted with complex situations and a high probability ofsurprise.

In describing uncertainty, clarity is essential since ‘‘anuncertain estimate’’ can mean different things to differentpeople, ranging from an estimate just short of complete cer-tainty to speculation (MA 2003, Chapter 7, Analytical Ap-proaches). Uncertainty can be the result of:• lack of information or a disagreement about what is

known or knowable;• ignorance or the possibility of surprise, both of which

derive from the impossibility of prediction in social-ecological systems;

• linguistic imprecision, statistical variation, measurementerror, variability, approximation, or subjective judg-ments.Uncertainty can be further compounded by long time

lags among driving forces, changes in ecosystem services,changes in human well-being, and responses. For ecosystemmanagement, uncertainties may be unknowns related to theecological system, such as how vulnerable a system is todisturbance; the social system, such as the strength of theeconomy over the next 50 years; or the interaction of socialand ecological systems, such as how the strength of theeconomy will affect tourism or logging.

Fundamental uncertainties to be addressed by scenariosrelate to key drivers of ecosystem change, and responsesby individuals or groups. While scenarios will not resolveuncertainties, they can help stakeholders make better deci-sions in the face of uncertainty. For example, if the keyuncertainty in a particular context is whether the economywill remain strong, decision-makers may choose a policyapproach that is likely to be successful whether the econ-omy is strong or weak. The scenario development processmay also help decision-makers prioritize the most importantuncertainties for further research.

10.2.1 Key Uncertainties Identified in Sub-globalAssessments

The main uncertainties that the sub-global assessments at-tempted to capture in their sets of scenarios varied greatly.

PAGE 236

While these uncertainties may be classified in many ways,Table 10.2 groups them according to whether they are ex-ogenous or endogenous. Exogenous uncertainties are thoseuncertainties related to drivers that operate primarily fromoutside the assessment region for which the scenario wasdeveloped. Endogenous uncertainties are related to driversthat are controlled primarily within the assessment region.For a small tourist area in Borneo, an endogenous uncer-tainty might relate to how much forest a community willcut, and an exogenous uncertainty might include how na-tional and global economic activity will affect the rate oftourism to ‘‘exotic’’ locations such as Borneo. Clarifyingthe uncertainties in this way can improve understanding ofwhat exogenous drivers the sub-global assessment stake-holders thought were important (but largely uncontrollableby them) and what endogenous drivers they considered im-portant (and somewhat more controllable by stakeholdersin the system).

Many sub-global assessments identified similar exoge-nous uncertainties, and not surprisingly, many highlightedbroad clusters of issues like governance and markets. In themultiscale Southern Africa sub-global assessment, where thescenarios at different scales were developed independently,governance emerged as a key uncertainty at all scales. Forsome assessments, resource use outside the assessment areawas also a key uncertainty.

Interestingly, the sub-global assessments shared severalendogenous uncertainties as well. Common endogenousuncertainties included institutional arrangements, wealthdistribution and equity, and governance. In the scenariosdeveloped by the Caribbean Sea assessment, the main un-certainty was whether the region would continue to relyon tourism as its main source of income or whether therewould be greater future diversification of the economy. Theissue of economic diversification in Caribbean Sea wasstrongly connected to a set of governance issues such as re-gional cooperation and trade negotiations.

Many other uncertainties were specific to individualsub-global assessments. For example, HIV/AIDS was iden-tified as a key uncertainty only for the Papua New Guineasub-global assessment. This does not mean that this factorwas not important in other sub-global assessments; it simplymeans that other scenarios did not identify it as critical forthe future. Decisions about key uncertainties are based notonly on what factors are unknown, but also on which fac-tors seem, at the time of scenario development, to be themost influential but uncertain determinants of the future.

For example, for scenarios in the Northern HighlandsLake District, Wisconsin, key uncertainties included the fu-ture impact of the national economy on tourism and immi-gration to the region, and local institutional arrangementsin the future for managing lakes and the ecosystem servicesthose lakes provide. Changes in the national economy willaffect how and where people vacation, which may affecttourism. Changes in institutional arrangements will affectthe impact of tourism and other activities in the area, butthe influence of these arrangements on tourism in future isuncertain given the many other factors related to the econ-omy or the availability of internal resources that would also

................. 11474$ CH10 10-17-05 15:45:51 PS


Table 10.2. Major Exogenous and Endogenous Uncertainties across the Sub-global Scenarios. Exogenous uncertainties are those thatare driven primarily from outside the assessment region for which the scenario was developed. Endogenous uncertainties are those that arecontrolled primarily within the assessment region.

Major UncertaintiesSub-globalAssessment Exogenous Endogenous

Tropical Forest Margins international markets for cocoa, oil palm, timber, and other human behaviortropical forest products adoption of new technologiesglobal environmental concerns and actions institutional arrangements

Western China central government policy rate of urbanization

population growth

SAfMA Regional governance (national and regional) governance (local)

equitability of wealth distribution equitability of wealth distribution

Wisconsin impact of state and national economies on tourism and institutional arrangements (for managing lakes andimmigration to the region ecosystem services)

resilience of ecosystems to changes in local use andmanagement

PNG balance of power between global and Asian economies population growth

HIV/AIDS

San Pedro de Atacama impact of state and national economies on tourism institutional arrangements (for managing tourism)

water use (by tourists, mining industry, and agriculture)

Caribbean Sea ecosystem vulnerability to change and thresholds governance mechanisms for ensuring equitable distributionof tourism incomeimpact of international economy on tourism

climate (impacts on tourism, ecosystems)

Northern Range climate variability distribution of wealth

ecosystem vulnerability to change and thresholds long-term economic prosperity

policy responsiveness

Sweden KW upstream water use institutional arrangements (bridging organizations)

India Urban national governance rate of adoption of new technology

empowerment of local communities rate of urbanization

Portugal societal attitudes toward the environment (proactive vs. societal attitudes toward the environment (proactive vs.reactive) reactive)

national agricultural practices and policies national agricultural practices

global connectedness (which also influences these practicesand policies)

Sinai regional governance awareness of environmental issues

demand for ecosystem services

local water pollution

Laguna Lake Basin governance population

international politics agricultural intensification

international markets industrialization

India Local markets institutional arrangements

legal system (especially with regard to criminalized nomadictribes)

PAGE 237................. 11474$ CH10 10-17-05 15:45:51 PS


influence the amount of tourism in the area. If the NorthernHighlands Lake District were to be one of the least devel-oped areas in the American Midwest in 25 years, it may bemore highly treasured and visited (and possibly then be-come more developed).

10.2.2 Describing Key Uncertainties afterIdentification

Having identified their key uncertainties, some sub-globalassessment teams chose to explain their scenarios by usingperpendicular axes to describe two major uncertainties, andthen locating the scenarios along those axes. For example,the three Wisconsin scenarios can be arranged, post-hoc,along two axes, one from ecological crisis through to grad-ual ecological change, and the second from local growth toexternally driven growth from retirement settlers and tour-ism. (See Figure 10.2.) The scenarios of the Tropical ForestMargins–Mae Chaem assessment can also be describedalong axes. In fact, most sets of scenarios could be describedalong multiple axes of uncertainty and the sub-global assess-ment teams could choose to explain the scenarios using axesmost appropriate to the situation in which each set of sce-narios were described.

It is important to note that many sub-global assessmentsfound that the use of axes, while useful post-hoc to explain

Figure 10.2. Two Sets of Uncertainty Axes That Could Be Usedto Describe the Northern Highland Lake District, Wisconsin,Scenarios

PAGE 238

and describe the scenarios, tended to limit creativity whenused as a scenario development tool. In particular, Carib-bean Sea participants explained that thinking about the un-certainties in terms of clusters was more useful and allowedthem to consider relationships among uncertainties. Thereis always a wide range of possible scenarios that can be de-veloped, and choosing only two axes tends to narrow thisrange of possibilities. Therefore, at least at the early stagesof scenario development, it may be more productive to em-brace the widest range of possible uncertainties and avoidthe use of axes.

10.2.3 Relationships of Uncertainty to EcosystemServices and Human Well-being

In general, sub-global assessments selected uncertainties byfocusing attention on socioeconomic drivers, with muchless attention on how these uncertainties might cascadethrough impacts on ecosystem functions, services, andhuman well-being. Conceptually, the cascade could lead toeither amplified or reduced uncertainties about humanwell-being. Figure 10.3 illustrates how this could poten-tially happen. In many sub-global assessments, surprises andshocks to the system and the impact on ecosystem servicesand human well-being are included in the scenarios. Yetthe human responses to these changes are, by and large,not included, although these indirect effects could eitherdampen or amplify uncertainties.

If resilience is high and there are many redundancies inthe system, uncertainties may affect ecosystem functionwithout having much impact on the flow of ecosystem ser-vices. On the other hand, if there are few redundancies,small changes to ecosystem function may have a large im-pact on the flow of services. Management, technology, andother social factors may affect the relationship betweenchanges in ecosystem services and human well-being. Ifpeople can use technology to substitute one service for an-other (for example, hunting or eco-tourism instead of tim-ber products from a forest) or to substitute the provision ofa service in the future for provision now (for example, or-ganic farming to preserve the environment instead of con-ventional farming), it may be possible to adjust to a declinein ecosystem services. However, if these devices are lacking,small changes in ecosystem services may have a large impacton human well-being.

In developing the Wisconsin scenarios, it was recog-nized that tourists bring income to the area, but also placehigh demands on ecosystem services. The amount and typeof tourism in the future—including whether it will involvemotorized water vehicles, require the development of infra-structure, or value quiet natural areas—are examples of thekind of uncertainties that will largely determine the use ofecosystem services and affect the ecosystems’ ability to pro-vide those services. In some of the Wisconsin scenarios, in-creased tourism leads to higher demands on local ecosystemsfor nutrient cycling, sewage management, and pollution di-lution. At the same time, some aspects of human well-beingare improved by increased flow of money into the regionaleconomy. Scenario development in the Northern High-

................. 11474$ CH10 10-17-05 15:45:53 PS


Figure 10.3. Cascading Uncertainties in Social-ecological Systems That Might Be Captured in Scenarios

lands Lake District was useful in thinking about trade-offssuch as these and encouraging stakeholder discussion aboutimportant future decisions related to managing tourism andecosystem services.

10.3 Telling Plausible Stories

10.3.1 Diversity of Approaches

A detailed methodology of how to develop scenarios waspresented to all sub-global assessments and discussed withrepresentatives of the global MA Scenarios Working Groupduring a scenario training workshop. (See Table 10.3, andBennett and Zurek 2004.) Although the ‘‘scenario cook-book’’ provided a detailed guide, sub-global assessmentswere encouraged to adapt the methods to their local cir-cumstances. Unfortunately, not all sub-global assessmentswere completely clear on what type of scenarios should bedeveloped and this, combined with specific local interestsand limited financial resources, led to a wide variety of sce-nario products that, more often than not, were less elaboratethan initially envisioned.

Step-by-step approaches to developing scenarios weredocumented by the SAfMA assessments and by CaribbeanSea. SAfMA Regional identified governance as the majoruncertainty, leading to bifurcation in the storylines for theregion over the next three decades. The two resulting sce-narios were fleshed out by filling in the MA conceptualframework diagram as a causal diagram. These key bifurca-tions were translated from the regional scale to the GariepBasin scale to local studies in a ‘‘cascade’’ of progressivelyrefined and locally relevant scenarios (Biggs et al. 2004).The first scenario, named ‘‘African Patchwork,’’ relied onprojecting current trends (that is, a business-as-usual sce-

PAGE 239

nario). The second scenario, ‘‘African Partnership,’’ pre-sented a vision of the future based on the strong politicalsupport for the New Partnership for Africa’s Development.(See Figure 10.4).

Many countries have similar politically endorsed fore-sighting initiatives, which usually use either forecasting (Wol-lenberg et al. 2000; Rotmans et al. 2000) or backcasting(Dreborg 1996; Robinson 1982; Robinson 2003) scenariodevelopment approaches to develop a desirable vision of thefuture based on expected or desired policy directions. Oneexample of a forecasting scenario is the Stockholm Urbanassessment, in which a newly developed governmental pro-gram for protected areas was evaluated in terms of how itwill affect key organizations (for example, NGOs or politi-cal bodies) and local stakeholders active in the managementof urban green areas. In the case of the Caribbean Sea assess-ment, both regional and global experts were involved inscenario development. (See Box 10.1 for a summary of theCaribbean Sea scenario development approach.)

10.3.2 Frameworks Used to Develop Storylines

In developing scenarios at the sub-global level, assessmentsmade use of different frameworks to structure their work.Some assessments, such as the Eastern Himalayas, started offby trying to map their local situation directly to one of theMA global scenario archetypes. Other assessments used dia-grams with perpendicular crossed axes representing majordirect and indirect drivers. The four resulting quadrantsthen represented the major storyline permutations (exam-ples include Portugal, Sweden SU, and Caribbean Sea). An-ticipated changes in drivers were then assessed qualitatively,and storylines developed.

................. 11474$ CH10 10-17-05 15:45:55 PS


Table 10.3. Steps in the Scenario Development Process

When Step Who

Before first workshop Decide on purpose of scenario and stakeholder Core team in consultation with main stakeholder groupsinvolvement

At first workshop Icebreaker: Getting people to think about the long-term Stakeholder group with core team facilitationfuture

Introduce the concept of scenarios Core team

Back casting exercisea Stakeholder group with core team facilitation

Identify main areas of uncertaintya Stakeholder group with core team facilitation

Develop focal questions to be addressed by scenariosa Stakeholder group with core team facilitation

Identify main drivers of changea Stakeholder group with core team facilitation

Develop first set of storylines Stakeholder group with core team facilitation

Decide on modeling capacity Stakeholder group and core team

Between workshops Prepare documentation material of first workshop Core team

Model runs Modeling teams

At the second workshop Critically assess storylines and incorporate model resultsa Stakeholder group with core team facilitation

Identify important surprisesa Stakeholder group with core team facilitation

Identify implications of scenarios for the main stakeholder Stakeholder group with core team facilitationgroups in the area

Decide on final storylines Stakeholder group with core team facilitation

Evaluate the implications of each scenario for addressingidentified uncertainties

Optional Wider stakeholder feedback session and scenario Core team facilitationiterations

After workshop Final write up of scenario storylines and their implications Core team

Dissemination of scenarios write-up Core team

a The sequencing of these steps can vary.

10.3.2.1 Portugal

In the first phase of the national scale Portugal assessment,one of the two major axes of change was related to society’sattitudes toward environmental change—reactive versusproactive. The second axis related to agricultural practicesin Portugal, particularly the effects of agricultural policieson those practices. In these independently created scenarios,the societal attitudes axis was shared with the global MAscenarios. This convergence between global and sub-globalscenarios meant that some of the quantitative modeling al-ready done in the global scenarios could be ground-truthedin Portugal. It also provided the opportunity to scale downthe global scenarios, where in the second phase of the as-sessment, the global scenarios were used to provide bound-ary conditions for the national scenarios in Portugal. Thelocal assessment of Sistelo in Portugal developed scenariosindependently, the results of which were later integratedinto the national scenarios (Pereira et al. in review).

10.3.2.2 Tropical Forest Margins–Mae Chaem, Thailand

The Tropical Forest Margins–Mae Chaem scenarios alsostarted with two contrasting perpendicular axes. The firstcaptured the character of linkages to outside regions (local-

PAGE 240

ized versus globalized), and the second captured the sectoralcomposition of economic development (agricultural versusdiversified). Taken together these two axes were seen as de-termining a third composite axis of ‘‘connectivity’’ (or clus-ter of processes), which appeared to capture an essentialcluster of features in each of the quadrants that were subse-quently developed into each of the contrasting scenarios.

10.3.2.3 Southern Africa Assessments

The SAfMA nested assessments adopted an experimentalapproach to multiscale scenario development. Each SAfMAcomponent assessment selected methods designed to answerthe questions relevant at its particular scale, while also at-tempting to retain multiscale comparability. The local scaledata were largely collected using participatory methods,while the basin and regional scale studies primarily madeuse of published studies, national and international data-bases, and modeling approaches.

At the basin scale, existing scenarios previously producedfor the region were first explored. Subsequently, the feasi-bility of adapting elements of the MA global scenarios wasinvestigated. After reviewing existing scenarios, the choicewas made to develop a set of scenarios with input fromstakeholders and the other Southern Africa component as-

................. 11474$ CH10 10-17-05 15:45:56 PS


Figure 10.4. The Southern Africa Regional Assessment Scenario Development Approach. The Southern Africa Regional assessmentmade use of the MA conceptual framework to synthesize several existing regional scenarios. The African Partnership scenario was based onan action plan called the New Partnership for Africa’s Development, a politically endorsed initiative for the economic and social developmentof the continent. By mapping the main aspects of the action plan into the framework, it was possible to identify targets that may be comprisedby the degradation of ecosystem services.

PAGE 241................. 11474$ CH10 10-17-05 15:46:15 PS


BOX 10.1

Developing Storylines: The Caribbean Sea Approach

The Caribbean Sea assessment pursued a two-part approach to devel-oping scenarios, first encouraging stakeholder involvement, and thenrefining and testing scenarios.

Part I: Stakeholder involvement (conducted separately in smallbreakout groups)

1. Assess current knowledge and the current state of the system• List key historical eras, key actors, external forces affecting the

region, and ecosystem services.2. Identify a focal issue or issues

• List hopes and fears for the future to determine focal questionsthat might drive scenario development.

3. Identify alternative trajectories and branch points• Identify key sources and threats to the resilience of the region

(the ability to adapt to future surprises). Focal questions for Ca-ribbean marine ecosystems were developed based on: What isthe biggest concern for the region in the future? What are thefactors that can make this region better able to cope with what-ever the future brings?

4. Build scenarios• Ask breakout groups to come up with a set of storylines, based

on the focal questions and considering the major historical erasand vulnerabilities of the region. Groups were asked to do this in1 hour to create a sense of urgency. Each group was then askedto present their set of scenarios in a plenary session.

Part II: Refine and test the scenarios (conducted with a single, smallfocus group)

5. Assess the scenarios• Are there certain themes emerging? Can we come up with a set

of four or so that seem to make sense together? What are thecritical contrasts among the scenarios? What are the recurringthemes? Are the trends/events plausible? What trends/events areuseful for illustrating key themes or concepts?

Does the set of scenarios address the focal questions?6. Shocks and surprises

• Come up with a list of shocks and surprises that might happen andask how the world would respond under each of the scenarios.

7. Use this analysis to refine a final set of scenarios.

sessments at the regional and local scale. At the local scale,future scenarios for three local sites within the Gariep Basinwere constructed during a workshop for project teammembers using the following steps:• Static drivers of change were identified. These are

known drivers and trends that will not change withinthe time horizon considered but will have an influenceon the future that unfolds.

• Assumptions were identified and listed, especially thoserelating to dynamic variables that have the potential tochange during the time horizon considered, but werenot expected to change.

• The key uncertainties were identified. These are dy-namic variables that were expected to change during theforecast period and to influence the outcome of the

PAGE 242

scenarios. Changes in the influence of these drivers, orpossible options around key uncertainties, were thenelaborated for each driver.

• Drivers of change were ranked from potentially the mostinfluential to the least.

• Correlations between drivers were identified, and wheredetected, the least influential driver was omitted.

• A suite of proto-scenarios was then developed fromcombinations of different pathways for each of the mostinfluential drivers.

• Each resultant scenario was then assessed in terms of itsfeasibility and probability. Those scenarios deemed un-feasible or improbable were then omitted.From within the final suite of scenarios, a subset was

selected, so that each scenario represented a cluster at differ-ent stages along the primary iteration. These were thenfleshed out in narrative form with respect to the state ofissues/descriptors of interest under such combinations ofdrivers of change.

In summary, most sub-global assessments used somevariant of the MA methodology to develop largely qualita-tive stories based on major drivers of change. There waslittle evidence of the use of causal diagrams, but the SanPedro de Atacama and SAfMA Regional assessments dem-onstrated a relatively simple approach to filling in the MAconceptual framework to accomplish the same end.

10.3.3 Systems Models, Quantification, and SpatialExplicitness

The most common approach among sub-global assessmentswas to opt for qualitative scenarios that relied either on un-structured narratives (for example, India Urban and EasternHimalayas) or more structured sequences of events (for ex-ample, Coastal BC, Sinai, and San Pedro de Atacama). Thisis a logical first step in scenario development and the mosteffective way to engage a wide variety of stakeholders.Many sub-global assessments noted the desire to quantifystorylines, yet in many cases, time constraints and the lackof a calibrated and validated quantitative model for the as-sessment area limited the possibilities for linking to models.Quantitative scenarios based on computable model-basedexercises (such as Western China and SAfMA Regional)were therefore less common. However, the rarity of quanti-tative analysis could well be a consequence of the relativelyshort period of sub-global assessment work considered inthis chapter. In the future, many sub-global assessmentsmight engage in quantitative analyses of the various sce-narios.

Structural or soft models and event sequence diagramswere used in the Tropical Forest Margins–Mae Chaem sce-narios to help clarify the logical assumptions behind each ofthe stories. (See Figure 10.5.) These were prepared beforethe text was written, and then partly revised and adjusted asthe storylines became richer and more specific. Within theIndia Urban assessment, a qualitative, structural diagramwith feedbacks based in implicit data and relationships wasused. Several case studies in the Goulburn Broken Catch-ment assessment in Australia (CSIRO Sustainable Ecosys-

................. 11474$ CH10 10-17-05 15:46:15 PS


Figure 10.5. The Use of Soft-models as an Intermediate Step in the Forest and Agroecosystem Trade-offs in the Humid Tropics(Tropical Forest Margins) Scenario Exercise

tems 2003) made use of explicit system dynamics modelsquantifying ecological processes and, in some cases, finan-cial returns.

Some scenario exercises produced high-resolution mapsof land use and land cover. In Western China, scenarioswere developed specifically to link to a Geographic Infor-mation System-based model, and scenarios were thereforequantitative rather than qualitative. Several other sub-global

PAGE 243

assessment groups indicated that they plan to use spatiallyexplicit models at a later stage (Laguna Lake Basin, Down-stream Mekong) using rather simple but easily applicablemodels like CLUE-S (Verburg et al. 2002). In SAfMA Re-gional, extensive use was made of various models (Biggset al. 2004) to assess the effects of different scenarios onbiodiversity. Figure 10.6 shows the results of runs of theIMAGE model (Alcamo et al. 1998) for two regional sce-

................. 11474$ CH10 10-17-05 15:46:18 PS


Figure 10.6. Example of Spatially Explicit Scenario Outputs in the Southern Africa Regional Assessment. Projections of land coverchange using the IMAGE model for the (b) African Patchwork and (c) the African Partnership scenarios. (Scholes and Biggs 2004)

narios, African Patchwork and African Partnership. In thisapplication, a downscaled version of an existing globalmodel was used, showing one way in which scenario quan-tification can be done.

As a rule, scenario development exercises that used quan-tification did so for only a restricted part of the overall issuesthey set out to explore and assess. This is a sensible approachsince frequently the objective of scenario development is toallow exploration of uncertainties, surprises, and thresholds,not all of which are easy to model quantitatively.

10.4 Incorporating Ecology into ScenariosThe use of scenarios in the MA was intended to differ fromearlier exercises in that the MA explicitly set out to inte-grate ecology into the scenarios design (Bennett et al.2003). Ecosystems were not to be treated solely as passiverecipients of the impacts of changes in socioeconomic sys-tems, but also as having a role in jointly determining socialfutures. Changes in the flow of ecosystem services are seenas potentially altering future development pathways. Thisview of how integrated social-ecological systems unfoldover time is much more co-evolutionary than in scenarioexercises where the goal is only to assess environmentalchanges.

10.4.1 Ecological Detail in Scenarios

Despite the MA’s focus on ecology, the ecological detail inmost sub-global assessment scenarios was relatively weak.While the impact of external and internal shocks to ecosys-tems and the consequences for ecosystem services andhuman well-being were considered by the sub-global as-sessments, the scenarios developed did not explicitly includeecological feedbacks (see also Figure 10.3). Usually, thefocus was on socioeconomic factors and the consequences

PAGE 244

of these for variables such as land use change, biodiversity,or food production. The examples below draw on twocompleted assessments that did attempt to include ecologi-cal feedbacks.

Several case studies in the Goulburn Broken Catchmentassessment (CSIRO Sustainable Ecosystems 2003) made useof explicit system dynamics models to explore alternativeland management scenarios. For example, in the GoulburnRiver Floodplain, Australia, case study, dynamic models ofbiophysical processes and a land management model weredeveloped to explore alternative land management strategies.In this landscape, the alternative land uses were combina-tions of cropping, grazing, and conservation. The scenariosidentified strategies to maximize financial returns and con-servation value, meet targets for protection of differentvegetation types, and create mixed land uses to achievemultiple goals. Through a model interface, this approachhelped users to explore trade-offs and spatial interactionsamong ecosystem services.

The Northern Highlands Lake District of Wisconsin hasbeen a field laboratory for ecological research for severaldecades, and a site for long-term ecological research fundedby the U.S. National Science Foundation since 1982. Theassessment analysis benefited from a tremendous baseline ofecological (as well as demographic and economic) informa-tion. Thus the scenario exercise, focusing on lake ecosys-tems, was able to bring substantial ecological explicitnessinto the analysis. For example, narratives on how differentactors would use various ecosystem services, what maydrive changes in these services (for example, biological in-vasions, climate, human migration), and the linkages be-tween these, were constructed.

The India Urban assessment included the negative feed-back of the constant drain of soil nutrients, water, and car-bon to the sea and atmosphere due to urbanization and

................. 11474$ CH10 10-17-05 15:46:24 PS


declining annual productivity, as well as overall scarcity offood, water, and fuel services. In SAfMA, modeling exer-cises were used that included feedbacks, while in the Carib-bean Sea assessment, the negative feedback on the economyof the entire region that emerges from ecosystem (coralreef ) destruction by overfishing and tourism was explored.

10.4.2 Ecological Interactions and Surprises

Several sub-global assessments considered the possibility ofecological surprises. For example, the India Urban assess-ment addressed potential surprises such as global increasesin flooding and the collapse of transport systems due to un-anticipated abrupt climatic change; scenarios discussed howcities affected by such calamities might sink, starve, and per-ish, while some upland subsistence farmer communitiescould survive.

The Wisconsin assessment has the best example of usingscenarios to consider the implications of ecological surprises(Carpenter et al. 2003; Peterson et al. 2003a). In the ‘‘Wall-eye Commons’’ scenario, conflict over resource use inter-acts with a series of unexpected environmental changes,including climate shifts that affect winter tourism, fisherycollapses, and zoonotic diseases. These unexpected eventsprovoke conflicts as different groups blame each other forthe resulting losses. The net effect of the social and ecologi-cal breakdowns is an exodus of many people from the re-gion.

10.5 Dealing with ScaleThe MA design process placed strong emphasis on gainingbetter understandings of interactions between human activ-ities and ecosystem services at multiple scales. In the light ofChapter 4 of this volume, which examines the scale-relatedissues in conducting an assessment, and of Chapter 5 of theMA conceptual framework (MA 2003), the discussion inthis chapter addresses the specific challenges in developingmultiscale scenarios.

10.5.1 Multi- and Cross-scale Processes in Scenarios

Multiscale processes were taken into account in all sub-global assessment scenarios, but often this was limited toincorporating ‘‘exogenous’’ drivers in the construction ofscenarios and making general comparisons with global sce-nario exercises. Such a limited approach makes sense wherecross-scale interactions are primarily top-down or relativelyunimportant in comparison with the processes that takeplace at the scale of primary interest of an assessment. Aboveall, this keeps the analysis simpler—one of the main reasonsgiven by several sub-global assessments for declining to in-vest more resources for including multiscale considerationsin their scenarios. (See Table 10.4.)

Some sub-global assessments used the global scenarios toset boundary conditions for the scenarios developed in theirassessments. The SAfMA local assessments used basin-levelscenarios to set their boundary conditions; and the basin-level scenarios in turn considered regional scenarios for thesame purpose. In the Caribbean Sea assessment, scenarios

PAGE 245

Table 10.4. Reasons Given by the Sub-global Assessments forExcluding or Including Multiscale Considerations in theirScenario Exercises: Selected Examples

Reasons for Excluding Multiscale Reasons for Including MultiscaleConsiderations Considerations

Regional scenarios that bridgeThe importance of the local issuethe local and the global alreadyexceeds any possible global de-existvelopmentExample: SAfMA. Scenarios in thisExample: Sweden KW. Recentassessment built on a number offlooding events have put the ques-sub-Saharan scenarios that alreadytion of protection against flood waterexisted.high on the agenda. The main issue

is whether dikes should be raised or Links between regional andnew wetlands should be created. global levels are strong and evi-Global issues like climate change dentare important, but the issue is clear Examples: Caribbean Sea. Thisand mostly local. area of the Caribbean Sea includes

parts of around 30 countries at theNational government is dominantregional level, which facilitates thein the organization of national re-link to global scenarios. Additionally,sources to implement sustainabil-many issues are the same at bothity goalsregional and global levels.Example: Western China. China’sPortugal. Scenarios were first devel-economy is no longer centrallyoped independently at the nationalplanned, and local government andlevel, and proved so similar to thepeople have significant roles in tak-global scenarios that the global sce-ing decisions on local affairs. At thenarios were taken and adapted tosame time, China is affected byPortugal instead.global change and the international

market economy, for example The link between the local and thethrough the accession of China to global is strong.the WTO. Nevertheless, because Example: Sweden SU. In general,China is very large and has many less isolated local areas have morenationalities, the national govern- connections with the regional andment is dominant in the organization global levels and, consequently,of national resources and the West- translation of issues across scales isern China sub-global assessment easier.was structured accordingly.

Information at the regional level isinaccessible and an analysis isimpossible for security reasonsExample: Sinai. The Sinai desert isa strategically important region; in-formation at any level above thelocal is scarce and often not acces-sible.

It is difficult to link the local levelto global scenariosExample: Bajo Chirripo. This as-sessment focused on the needs andconcerns of the indigenous people inthe area. The user needs here areso different that linking to globallevel scenarios was difficult.

were primarily targeted at the regional level, and the globalscenarios were explicitly used as boundary conditions.

To incorporate more complexity into their scenarios,however, a few assessments—notably SAfMA and Portu-gal—used more sophisticated multiscale scenario methodol-ogies. Although these two assessments were exceptions

................. 11474$ CH10 10-17-05 15:46:24 PS


rather than the rule, they demonstrate the potential of em-ploying multiscale methodologies by developing internallyconsistent stories at more than one scale. In the SouthernAfrica set of assessments, there were three levels of assess-ment, each involving at least some scenario activities.Though linked, these separate scenario exercises primarilyaddressed the particular needs and goals of each of the com-ponent assessments. The Portugal assessment also worked atmultiple scales, but cases at the local scale were not com-pletely nested in the larger-scale scenarios. (See Box 10.2.)

The Tropical Forest Margins–Mae Chaem scenarioswere planned as one of six sets of scenarios to be developedfor the benchmark sites of the Alternatives to Slash-and-Burn consortium in six countries worldwide (Brazil, Peru,Cameroon, Indonesia, Thailand and the Philippines). At thesame time, these scenarios were also linked with two addi-tional new exercises in northern Thailand and the MekongRiver Basin (across the five countries of southwest China,Laos, Thailand, Cambodia, and Viet Nam). The geographi-

BOX 10.2

Nested, Multiscale Design: SAfMA and Portugal

Within the Southern Africa assessment, five local scale assessments, The Portugal assessment was undertaken at three scales: national,each covering the area of a community or local authority, were nested basin, and local. There were two basin assessments and four local as-within two basin-scale assessments, which in turn lay within an assess- sessments. The local case studies were not within the basins studied, andment of the greater SADC region. All contributed to the MA global assess- covered different reporting units (systems) of the national assessment.ment, but were not linked to the global scenario storylines (dotted lines in These reporting units included a very small rural community (mountainthe figure). system), two farms (cultivated systems), and a biological research station

Scenario development followed the same general design. SAfMA de- (Montado system). National scenarios were first developed independentlyrived two scenarios by synthesizing information from several recent sce- and in a second phase compared with the global ones.nario exercises in the region, rather than developing its own scenarios. One of the two axes of the Portugal scenarios was similar to one ofThe classification of existing regional and global scenarios into a limited the global ones, and congruence between the Portugal and global scenar-number of archetypes provided the basis for interaction and comparison ios could be established. Thus the Portugal assessment decided to adoptwith the MA global scenarios working group. The Gariep Basin assess- the global scenarios as boundary conditions for the national scenarios,ment used four scenario archetypes and explored what it would mean for and developed storylines accordingly. This also allowed the use of theecosystem services and human well-being if the conditions of each of the quantitative modeling work already done in the global scenarios, and pro-archetypes prevailed in the basin. vided a test for the scaling down and regional calibration of the global

The local community assessments within the Gariep Basin focused on scenarios.identifying the key drivers in the different study areas. The likely permuta- The Portugal assessment plans to develop scenarios for basin andtions of these drivers were considered and led to the development of local assessments also using the global scenarios. Nevertheless, the localthree scenarios. The local Gorongosa-Marromeu assessment (within the community assessment of Sistelo developed independent scenarios thatZambezi Basin) used scenarios in the sense of ‘‘wind-tunnels’’ for testing were subsequently integrated with the national scenarios.the robustness of policy and management responses in alternate plausibleworlds.

PAGE 246

cal scope of the scenarios in each case was set at a scalebroader than the focus area for assessment and simulation oflandscape trajectories. This was considered extremely im-portant in these assessments because of the major assump-tions on historical, cultural, trade, investment, andinformation links, and the structure of the regional econ-omy. At the time of writing of this report, scenarios for arural area (the Mae Chaem watershed) have been com-pleted, and those for an urbanizing area (greater ChiangMai) were underway.

Explicit nesting of scenarios is advantageous, but it alsointroduces increased complexity in scenario work. Nestingallows for some preliminary consideration of the plausibilityof particular scenarios unfolding at one scale in the face ofchanges at another. When these mixtures are cross-tabulated,many ‘‘discordant’’ combinations can be eliminated as un-stable and unlikely to persist. For example, it is hard fornorthern Thailand to pursue a strong, locally oriented de-velopment trajectory while the wider region unfolds in a

................. 11474$ CH10 10-17-05 15:46:26 PS


highly globalized and interconnected manner. Althoughthis approach of examining ‘‘discordance’’ among scenarioscreated at different levels is pragmatic, it still falls short ofallowing for dynamic feedback between scenarios at variousscales.

10.5.2 Spatial Extent, Heterogeneity, andResolution

The sub-global assessment scenario exercises varied substan-tially with regard to spatial extent, boundary setting, andhandling of spatial heterogeneity.

The spatial extent at each level was typically defined ei-ther by socioeconomic (for example, village, city, nation)or biophysical (for example, watershed) boundaries. Forexample, even within the various SAfMA component as-sessments, different criteria were used to set boundaries.SAfMA Regional was defined by a political grouping ofcountries (the Southern African Development Commu-nity); SAfMA Gariep and SAfMA Zambezi were definedby large drainage basins (the Gariep and Zambezi basins,respectively); SAfMA Gorongosa-Marromeu and SAfMALivelihoods had boundaries that were politically defined.

To understand the implications of spatial heterogeneity,several sub-global assessments used baseline maps of bio-physical or basic demographic features to help understandhow scenarios unfold differently at different locations. Inthese cases, the resolution often differs. For example, thesimulated landscapes in the Tropical Forest Margins scenar-ios at the Mekong Region scale were based on 10x10 ki-lometer gridded input data and processes at this and higheraggregated scales; in contrast, the Tropical Forest Margins–Mae Chaem simulations used a 1x1 kilometer grid as thebasis for land use change. The latter landscape may in futurebe modeled hydrologically on a finer 30x30 meter digitalelevation grid.

The issue of spatial resolution and heterogeneity isclosely related to decisions about nesting. In practice, devel-oping full scenarios for more than 2 or 3 scales is difficult toenvision, but additional detail for some ecosystem servicesmay nevertheless be available through (dis-)aggregation andmapping techniques.

10.5.3 Time Horizons

In general, one human generation was the longest temporalscale considered by stakeholders, and therefore scenarios inthe various sub-global assessments usually limited them-selves to a 15–25 year time horizon, though a few extendedthe time period considered to 2050. However, the SAfMAlocal assessments noted that traditional communities mightuse much longer time horizons in their planning. Some alsoincluded a longer historical time period. For example, theIndia Urban assessment analyzed the rise and fall of the Brit-ish Empire over the past centuries. In several assessments,the temporal dimension was not made explicit. ‘‘What if. . . scenarios’’ (Coastal BC) and (un)desirable futures (Kar-nataka village cluster in India Local) are two examples. Theaim in these assessments was to reason through the effect ofcertain policy measures and their ability to deliver desirable

PAGE 247

futures, rather than to discuss particular time frames. Boththese approaches are particularly appealing options whenstakeholders are heavily involved in the process; discussionstend to be livelier and results more creative when partici-pants are asked to develop visions for a desirable future orto respond to certain policy measures, without being con-cerned with questions about exactly when things will hap-pen (Robinson 2003; Kok et al. 2006).

Immediate issues of concern to stakeholders can limitpeople’s perceptions about what is a useful time frame forscenario development. As was stated by the Papua NewGuinea assessment in a Knowledge Market session, ‘‘Evenat the national level, there is a strange unwillingness on thepart of national bureaucrats to think beyond the next fiveyears’’ (Colin Filer, personal communication, KM2).

10.5.4 The Linkage between Sub-global and GlobalScenarios

Sub-global scenarios often targeted specific user groupsand/or addressed specific questions. Therefore, it was notalways possible to link the sub-global scenarios to the globalscenarios. To harmonize the global and sub-global scenarioexercises as much as possible, the following steps weretaken:• Representatives of several sub-global assessments partici-

pated in the global scenarios working group.• Participants of the global scenario team participated at

various occasions in meetings of the sub-global assess-ments, explaining both the preliminary global scenarioresults and the procedure that was followed in develop-ing the global scenarios.

• At several MA Sub-global Working Group meetings,representatives of the various sub-global assessmentswere given the opportunity to discuss methodologieswith each other. For example, the two Knowledge Mar-kets facilitated the interaction among representatives ofmost sub-global assessments.The sub-global assessments had to make practical deci-

sions about how they would make use of the still-evolvingglobal scenarios in their own work. Most groups explicitlyopted to focus first on the needs and issues in their sub-global assessments, rather than be overly bound by the di-rections taken by the global scenarios. This made sensegiven that sub-global scenarios need to have relevance andexplore the options available to decision-makers at sub-global scales.

Nevertheless, exploring commonalities and differencesbetween individual sub-global assessment scenarios and theglobal scenarios may highlight gaps in logic at either scaleor, in some cases, genuine regional differences that are notapparent in more aggregated analyses. These comparisonsmay also reveal ways in which the global scenarios could beimproved to take into account regional differences. Particu-larly relevant are the sub-global body of information onquantitative, scale-dependent variables such as biodiversityand land degradation. The spatially explicit informationgathered by the SAfMA Regional assessment (Biggs et al.2004) on changes in ecosystem services is a good example.

................. 11474$ CH10 10-17-05 15:46:27 PS


Different possibilities exist for how scenarios can be nes-ted within each other to create a set of multiscale scenarios.We distinguish five ways to link global and sub-global sce-narios:• The global storylines can be played out at sub-global

scales. This would likely be the strongest link, where theglobal scenarios are downscaled to finer scales, butwhere the main storylines are similar.

• The global scenarios can be used as boundary conditionsfor finer scale scenarios, which then develop their ownstorylines. For example, the trends described for themajor driving forces at the global scale can be used fordescribing the range of variability these drivers exhibitat a regional level.

• The global scenarios can be used to create scenariosabout policy and management options being discussed atthe sub-global level.

• The underlying assumptions and worldviews played outin the global scenarios can be applied to developing sce-narios at sub-global scales.

• Scenarios can be developed independently at global andsub-global levels and can then be compared and mappedagainst each other afterwards.These potential methods of linkage were identified in

consultation with the sub-global assessments. The fourthand fifth methods of translating worldviews and mappingglobal and sub-global scenarios onto each other post-hocwere most frequently adopted. It can be expected that withmore sub-global assessments developing scenarios, the othermethods will be increasingly employed.

Several sub-global assessments (Portugal, PNG, Carib-bean Sea, and SAfMA Regional) compared their indepen-dently developed scenarios with the four global storylines.The Portugal assessment independently developed scenariosat the national scale, and when compared to the global sce-narios, one axis of change was found to be shared in com-mon. Because of this similarity with the global scenarios,the Portugal scenarios were replaced by the four globalstorylines adapted to Portugal. This replacement also madesense in that it allowed for the quantitative modeling al-ready done in the global scenarios to be applied to the Por-tugal scenarios work. Furthermore, sub-global use of globalscenarios could test methods for scaling down scenarios andprovide regional calibration of the global scenarios. ThePortugal assessment plans to further downscale the globalscenarios to basin and local levels. Sistelo, a local commu-nity assessment within the Portugal assessment (Pereira etal. in review), developed independent scenarios that werelater integrated into the national scenarios. Most othergroups did not find such a close overlap in objectives, keyuncertainties, or trends and processes between their scenar-ios and the MA global scenarios. The PNG assessmentnoted that the ‘‘MA is itself an activity which belongs towhat we call the Globalization scenario’’ (Colin Filer, per-sonal communication, KM2).

10.6 Participating in Scenario DevelopmentAn important reason for developing scenarios is to constructa more comprehensive analysis of the social-ecological sys-

PAGE 248

tem. Some stakeholders know a tremendous amount abouthow systems work ecologically as well as socially, and theirknowledge is more integrated compared to traditional aca-demic analyses. Most sub-global assessments acknowledgedthe need for participation but indicated that limited timeand budgets put constraints on what could be achieved.However, closer inspection reveals that sub-global assess-ments differed widely with regard to the reasons for seekingstakeholder participation, who was invited to participate,and the roles they were assigned in the process.

10.6.1 Goals of Participation

Within the context of the MA emphasis on engaging withassessment users, sub-global assessments gave very diversereasons for seeking stakeholder participation and, after theirexperiences, revised their perceptions about the value ofparticipation. Some assessments wanted to communicatepolicy options to stakeholders, while others sought to en-courage communication among them. For others, the pri-mary interest was to create ‘‘buy in’’ to the assessmentprocess. Buy-in refers to the process of first establishing mu-tual trust among scientists, decision-makers, and otherstakeholders, and subsequently establishing a common un-derstanding of the aims of the assessment, the value of par-ticipatory processes, and the possible benefits for localstakeholders. For example, the Wisconsin assessment triedto involve indigenous people, which proved difficult dueto a history of conflict and lack of trust between the state,non-native local people, and native people. Few assessmentsemployed specific methods to differentiate the perceptionsof various local stakeholders on future developments. Thefocus of the work remained on communication with stake-holders regarding their opinions about the future, withoutanalyzing differences in perceptions among various peopleor groups.

For simplicity, we synthesized the reasons for seekingstakeholder participation into several main themes. (SeeTable 10.5). Most of the reasons given were instrumental,that is, aimed at improving the quality of the assessment orat selling its message. However, normative reasons couldalso have been part of the rationale and some examples ofthese are listed at the bottom of the table. Leaders of sub-global assessment scenario exercises and stakeholders whoactually participated in the exercises probably had differentreasons for participating. Of course, the extent to whichtheir various goals were met depended greatly on the meth-ods actually used to engage stakeholders, and which stake-holders actually participated.

10.6.2 Selection of Participants

Unfortunately, most sub-global assessments provided littledocumentation of the criteria, reasons, and processes bywhich the participants in scenario work were selected. Thesub-global assessments had received detailed guidelinesfrom the MA on the stakeholder selection process (see Ben-nett and Zurek 2004), emphasizing that stakeholder selec-tion is a crucial step. Stakeholders can strongly influence the

................. 11474$ CH10 10-17-05 15:46:27 PS


Table 10.5. Reasons for Seeking Participation in Sub-globalAssessment Scenario Development Exercises

Instrumental (often expressed)Persuade and advocate environmental concerns. To raise environmental

awareness and increase the ecological knowledge of stakeholders.Capture an audience. To help achieve buy-in to the assessment process.Facilitate communication. Discussions around scenarios were used to help

explain assessment findings to various stakeholder groups in a flexibleform that related easily to their planning and decision-making concerns.

Instrumental (less frequently expressed)Increase the diversity of perspectives. Scientists have only a small range of

experiences and interests. Stakeholder participation helps broaden theperspectives that can be included in the scenarios, making them morerealistic and robust.

Improve understanding of social processes. To more accurately capture pref-erences, values, and possible response behaviors.

Ensure relevance. To ensure scenarios relate to the priorities of managersand decision-makers.

Normative (not usually expressed)Rights of stakeholders. It should be the right of stakeholders to have their

diverse interests represented at the table.Experts are shortsighted and biased. Experts and facilitators of scenario

exercises should not let their biases predetermine the direction and em-phasis of scenarios. Participation helps to mitigate such shortsightedness.

outcome of a scenario activity, because they can differgreatly in substantive judgments and/or worldviews.

The SAfMA assessments established a set of advisorycommittees for governance as well as for types of involve-ment in the various scenario exercises conducted at differ-ent scales. In contrast to the Goulburn Broken Catchmentscenario exercise described below, SAfMA found that it waseasier to engage stakeholders at the local scale than at theregional scale. Regional decision-makers at the supra-nationalscale (for example, SADC officials) were particularly diffi-cult to involve.

Sub-global assessments carried out at finer scales (Vilca-nota, Bajo Chirripo) placed special emphasis on engaginglocal and indigenous people, who would experience the ef-fects of scenario outcomes directly. In other sub-global as-sessments (Sweden SU, Sweden KW, and Western China),the scenarios were very narrowly focused, and therefore as-sessment teams could target a relatively narrow group ofstakeholders for participation. Probably the broadest stake-holder involvement was in SAfMA Livelihoods, SAfMAG-M, and India Local, each of which involved a broad sam-pling of civil society, including young people in particular.

For various reasons, including budget constraints, theGoulburn Broken Catchment assessment in Australia fo-cused on the participation of key political stakeholders. Inthis context, representatives from local, state, and common-wealth governments, the Goulburn Broken CatchmentManagement Authority, community leaders, and a few non-governmental organizations were targeted. The assessmentthen made sure that these groups participated in regularworkshops, as well as being part of steering and manage-ment committees of the assessment (CSIRO SustainableEcosystems 2003).

PAGE 249

10.6.3 Mechanisms of Participation

Although workshops were the most common tool for defin-ing or refining existing scenarios, a wide range of mechanismswere employed to engage stakeholders in the scenario-building process. Some sub-global assessments that were stillin the scenario planning process at the time of writing (forexample, Sao Paulo) intended to employ more innovativefacilitation methods to improve engagement of participants,including role-playing games, computer games, and citi-zens’ juries (also the main method used in the GoulburnBroken Catchment study). Most sub-global assessmentgroups acknowledged that workshops with stakeholders areexpensive, and take effort and skill to prepare and imple-ment. The number and size of these activities were clearlyconstrained by time and financial resources in most sub-global assessments. The Caribbean Sea assessment followedstakeholder workshops with expert group meetings to buildmultiscale scenarios.

Interview techniques were also employed (India Local,Sinai), as well as a range of other more informal meetings,mostly in the early phases of the assessment. In India Local,researchers were stationed in the villages, thus effectivelymonitoring and interviewing the stakeholders for longerperiods of time.

Within the Wisconsin assessment, questionnaires weredistributed and a website was maintained for public discus-sion and feedback. Individual interviews to discuss the sce-narios were conducted with leaders of local government,Native American tribes, and lake associations. People in theregion were surveyed randomly to gather their opinions ofthe scenarios. A computer game for managing the regionwas developed and tested with small groups of stakeholders;eventually the computer game will be used in workshops torevise the scenarios both qualitatively and quantitatively.The scenarios have been widely covered in the media. Al-though this project began as a largely expert-driven exer-cise, it has matured over time into providing more andmore avenues for public participation. As a result, the sce-narios are likely to be iteratively refined, as acknowledgedat the end of the first phase by the authors (Peterson et al.2003a).

10.6.4 Roles of Participants

Participation in scenario-building activities ranged frompassive roles, where stakeholders served simply as an audi-ence, to highly interactive experiences, where participantswere called upon to directly shape, design, and validate thescenarios. The ability to have direct input into the creationof scenarios, in addition to scenario validation, represents amore fully participatory approach. Coastal BC, India Local,India Urban, Tropical Forest Margins–Cameroon, Wiscon-sin, SAfMA Livelihoods, SAfMA G-M, and Portugal, allemployed workshops at some stage of the assessment togather valuable input on scenario design, as well as to re-view the scenarios.

10.6.5 Problems Encountered in Participation

Although most sub-global assessments lack documented in-formation on stakeholder involvement, the two Knowledge

................. 11474$ CH10 10-17-05 15:46:28 PS


Markets provided many insights on some of the problemsencountered in the scenario-building process. It is difficultto assess the underlying reasons for the problems that werementioned, but a few insights can be learned from thesediscussions.

Sub-global assessment teams acknowledged that the verynature of stakeholder involvement is a creative and stimu-lating process, although the outputs can be highly unpre-dictable in terms of their value and usefulness to scenarioactivities. The main conclusion is that involving stakehold-ers, especially through workshops and meetings, is a veryslow process that is resource-intensive. Other problemsmentioned included:• initially not all relevant stakeholders were selected (Ca-

ribbean Sea);• serious cultural obstacles to effective scenario-building

(PNG);• stakeholders’ unfamiliarity with radically different tech-

nologies (India Local);• exogenous driving forces not of interest to participating

communities (Trinidad);• stakeholders have limited access to, and/or limited un-

derstanding of, information on processes at other scales(Caribbean Sea, India Local); and

• preoccupation with tangible outputs (Portugal).These problems highlight the necessity of careful prepa-

ration before workshops and other information-gatheringexercises.

10.7 Communication in Sub-global ScenariosThe issues of communication and participation are closelyintertwined. If communication is viewed as advertising,then it is largely one-way, but if the aim is genuine andinterested dialogue, then the possibilities of more symmetri-cal two-way exchanges, among stakeholder groups or be-tween scientists and policy-makers, become real.

The development of scenarios involves two distinctforms of communication. One is the process of communi-cation between scientists and other stakeholders that enablesthe scientists and other stakeholders to formulate alternativestorylines. The other is the process by which the scientiststhen communicate these alternatives to an audience. Thislatter form of communication includes, but may not be re-stricted to, the stakeholders involved in the first process.The first process is communication for scenarios; the secondprocess is communication of scenarios. The first process wasaddressed in the preceding section on participation, thoughprimarily with an emphasis on the question of who partici-pates, rather than the form of communication throughwhich they participate. These two forms of communica-tion—among experts and between experts and others—theoretically could be connected in an on-going process ofdialogue by which scenarios are continually modified as aresult of feedback from the stakeholders who participate inthe initial construction, but this may not be feasible in prac-tice because of resource constraints.

PAGE 250

10.7.1 Communication Strategies

In relation to strategies adopted, the sub-global assessmentscan be divided into four groups:• those that adopted specific forms of communication for

both the development and dissemination of scenarios,and separated the communication of scenarios from thecommunication of the overall results of the assessmentin order to address the needs and interests of specific usergroups. Examples include the SAfMA local assessmentsand the Tropical Forest Margins assessments;

• those that used specific forms of communication withrelevant stakeholders in the development of their scenar-ios, but did not separate the communication of thesescenarios from the overall results of their assessments.The majority of sub-global assessments fall into this cate-gory;

• those that developed their initial scenarios without theactive participation of relevant stakeholders, but stillpropose to develop distinctive forms of communicationfor the modification and dissemination of these scenar-ios. The PNG assessment seems to be the only one inthis category (see Box 10.3);

• those that made little or no use of scenarios in their as-sessment—mostly because of budget constraints—andhave not therefore adopted any specific forms of com-munication for this purpose. A good example of a sub-global assessment that did not have the resources to in-vest in communication was Coastal BC, which intendedto use the interactive computer game QUEST (Tanseyet al. 2002).A minimal communication strategy would be one that

embeds the results of scenario development in written re-ports addressed to an audience of scientists and policy-makers. Most assessments proposed to do this. Some wentfurther than this, by communicating the results of the assess-ment (including scenarios) in a form that was suitable for‘‘consumption’’ by other user groups, such as schoolchil-dren or illiterate members of local communities. The restof this chapter focuses primarily on those cases in whichscenarios have been, or will be, treated as a separate outputof the assessment process.

10.7.2 Scenarios in Policy Dialogues

A number of sub-global assessments (for example, SAfMA,Tropical Forest Margins–Pan-Tropics, Western China) em-ployed similar modes of communication with policy-makersand/or decision-makers at the beginning of their scenariosexercises and at the end when reporting the results. Report-ing may be directed either to the same group of stakehold-ers involved in the original exercise, or to other groups atdifferent levels in the policy-making process. This type ofiterative communication normally involves a series ofworkshops or seminars in which scenarios are designed oranalyzed with the aid of computer technology. For exam-ple, the Caribbean Sea scenarios will be used to guide thedeliberations of ministers and senior officials engaged in theprocess of negotiating an intergovernmental managementregime.

................. 11474$ CH10 10-17-05 15:46:29 PS


BOX 10.3

Obstacles to Communication: Papua New Guinea

The first (national) phase of the PNG assessment did not engage in a proc- vation project in which the assessment is embedded. The choice of tech-ess of communication for scenario development, partly because of the limited niques for communication will depend partly on the lessons alreadybudget allocated for work in this phase and partly because some relevant learned from other sub-global assessments and partly on the lessonsscenarios had already been developed in previous forms of policy dialogue learned from experiments with similar forms of communication in theat the national scale. These earlier scenarios were modified by the scientists southwest Pacific region. A well-known community theater group (basedengaged in the assessment in order to take account both of the global sce- in Vanuatu) has produced a film about the effects of one driver (increasingnarios developed by the MA Scenarios Working Group and of the changing population pressure) on ecosystem services and human well-being in onetenor of national debate about the country’s political, economic, and environ- (imaginary) coastal local community. The national assessment report in-mental future, as reflected in the pages of the national press. cludes a brief evaluation of that process because of its potential applicabil-

The scenarios developed in the national phase of the assessment will ity in future assessments.be used as an input to the communication process for further scenario PNG also includes the interesting observation that popular or commu-development at the provincial and community scales in the second phase nity reactions to the process of scenario construction may be influenced,of the assessment. In this second phase, various communication tech- in some areas, by the prevalence of millenarian religious beliefs, espe-niques will be applied to the development of models and storylines at the cially when people are invited to include uncertainties or surprises or ca-national and community scales, and then to the task of communicating tastrophes in their storylines. This point underlines the importance ofscenarios through an education and outreach program. paying careful attention to the cultural context when designing a communi-

The outreach program is part of the community-based marine conser- cation strategy for the development and dissemination of scenarios.

The Tropical Forest Margins assessments illustrate theway in which this type of communication is varied to ac-commodate the needs and interests of decision-makers atdifferent levels in a policy process. For example, at thepan-tropic level, scenarios were developed to analyze theimplications of biodiversity loss on hydrological functions.Constant iteration with a World Bank expert helped to de-lineate the focus of the results in order to effectively informpolicy-makers. At the end of the project, the results werepresented in a one-day policy seminar at the World Bankand are available at: http://www.asb.cgiar.org/BNPP/phase2/bnpp_phase2_general.htm. At the regional scale inBrazil, an economic equilibrium model was used to test theeffects of devaluation on forest area in the Brazilian Ama-zon. The published results were mainly targeted at the re-gional government, as well as the national and internationalscientific communities. At the basin scale, Tropical ForestMargins–Mae Chaem tested different implications of hy-drological functions by applying expert-driven scenarios;the results were communicated at the World Bank seminar,but also will be part of a wider strategy to translate the re-sults to policy-makers throughout Southeast Asia. In Tropi-cal Forest Margins–Western Brazil, data obtained at thelocal level were used to inform farm-level ‘‘bioeconomicmodels’’ in which the effects of different mixes of techno-logical innovations were tested at different time horizons.The results were communicated to policy-makers throughEMBRAPA, a Brazilian institute that is one of the partnersin the Tropical Forest Margins consortium.

Stakeholder workshops and other types of face-to-facecommunication were used to disseminate the overall resultsof the SAfMA assessments at a number of different scales.SAfMA G-M presented verbal descriptions of its scenarios,first to local communities to elicit their responses (whatthey would do in each scenario) and then to local policy-makers to also elicit responses (what policies should be putin place). According to Tim Lynam (personal communica-

PAGE 251

tion), who participated in these exercises, the communitypresentations were very successful and their responses wereused in the assessment, whereas the policy presentationswere not as successful.

10.7.3 Storylines for a Wider Audience

Stories about the future can be the most interesting andexciting way of communicating the results of an assessmentto an audience that does not consist exclusively of scientistsand policy-makers. However, there could be some incom-patibility between the divergent needs of establishing thescientific credibility of alternative scenarios and communi-cating these scenarios effectively to a wider audience.

The SAfMA assessments followed the common practiceof incorporating the analysis of scenarios into written mate-rials that represented the overall results for the benefit of thedifferent users represented on the advisory committees. Thisassessment went further than most in producing differenttypes of documentation, from brochures to scientific jour-nal articles, in order to address the needs and interests ofdifferent user groups, and hence to convince them of theusefulness of the assessment. However, while the communi-cations strategy was thought to have been appropriate forthe needs of policy-makers and planners in the public sec-tor, it was recognized that communication with the generalpublic was more problematic.

San Pedro de Atacama was the only assessment that de-veloped specific media for communicating scenarios toschoolchildren. The assessment devised a process of famil-iarizing local children with the ecosystem that was beingassessed, partly by engaging indigenous elders to provideinformal lessons on recent ecosystem history. This exercisein oral history was expected to provide the basis for con-structing plausible futures. In India Local, schoolchildrenwere also engaged in the process of data collection andmonitoring, although the assessment did not include a sepa-rate communication of scenarios. PNG also used specific

................. 11474$ CH10 10-17-05 15:46:30 PS


media for communication with schoolchildren, but againnot in the context of developing scenarios. In the Wiscon-sin assessment, booklet cartoon maps and illustrations wereused to communicate the scenarios, though not necessarilyonly to young readers.

10.7.4 Spatial Representations of Scenarios

Spatial mapping or modeling techniques may be used as adistinctive form of communication in the development anddissemination of scenarios, but most sub-global assessmentsappear to have treated these as one aspect of the policydialogue with a sophisticated audience of scientists andpolicy-makers, rather than a technique for securing popularparticipation or understanding.

At the local level, the Tropical Forest Margins–Cameroon assessment had villagers actively participating inthe formulation of scenarios, using historic transects andcreating graphics of the landscape futures across time. Thiswas a form of participatory rural appraisal, which has beenwidely used in developing countries. In this particular case,the results were inconsistent with the scientific researchfindings on the investment strategies of local farmers.

In the case of Coastal BC, various stakeholders proposedemphasizing the use of mapping techniques to communi-cate the scenarios, as well as to record baseline data aboutthe condition of ecosystem services and human well-being.Stakeholders asserted that mapping products would allowthe users, including members of indigenous communities,to grasp the spatial and temporal scope of consequencesemerging from their decisions. Investment in clear and ef-fective techniques for the communication of scenarios wasalso thought to be a useful way of connecting the scientificand policy communities involved with the assessment. Theapplication of these mapping techniques proved to be dif-ficult, firstly because of technical problems with the soft-ware chosen for this purpose, and secondly because themethodology was not clearly understood by some of thestakeholders engaged in the initial process of communica-tion. In other contexts, there might have been some resis-tance from indigenous communities to the idea of mappingtheir resources and placing these maps in the public domain,but this does not seem to have been an important obstaclein this case.

10.7.5 Community Theater and Video Techniques

The representation of scenarios by means of theatrical per-formances or film and video recordings is almost certainlythe most powerful way of communicating ‘‘plausible fu-tures’’ to a popular audience, but the human and financialresources available to the sub-global assessments limited theuse of such techniques.

In the SAfMA Livelihoods assessment, a series of shortplays representing alternative scenarios were performedbefore an audience drawn from one of the communitiesinvolved in the assessment process. Members of the com-munity had some opportunity to participate in the processof scripting these plays before their performance, but hadnot been directly involved in the initial process of scenario

PAGE 252

development, which was based on communications withstakeholders at a higher level.

In this case, the communication of scenarios managed toaccomplish the difficult task of dramatizing the content of ascientific report, but was essentially a feedback mechanismadded on to the end of the assessment process. Like someother stakeholders in the assessment process, the audiencewas left to wonder what would happen after the scenarioshad been communicated, but had at least been convincedthat their own future was uncertain and that they could playa role in shaping it.

For SAfMA Livelihoods, the choice of media was dic-tated by the low levels of literacy in the community, and bythe observation that local people were more likely to en-gage with messages conveyed through live performancethan those conveyed through film. This still leaves open thequestion of how best to communicate scenarios to multiplecommunities living in similar circumstances, given the hightransaction costs involved in the medium of communitytheater.

10.8 Findings of the Sub-Global ScenariosMost of this chapter has focused on how the sub-globalscenarios were developed. This section considers the find-ings they produced. It is an analysis of a selection of theresulting scenarios, concentrating on the main drivingforces identified during scenario development, key proc-esses described, the main ecosystem services included, and across-scale comparison of the scenarios results. The sectionfirst describes a few examples of how uncertainties and eco-system services were treated, comparing among sub-globalscenario findings. It then analyzes the scenario findings withrespect to geographical scale and examines whether scalehad any influence on the interpretation of scenario out-comes.

This section focuses only on those scenario exercises thatwere well advanced in at least their first iteration and thathad therefore started to draw initial conclusions at the timeof writing. The comparison mainly draws from the multi-scale SAfMA assessment, Caribbean Sea, India Local, andthe first draft of scenarios from the Bajo Chirripo and Por-tugal assessments. It should be noted well that when moreresults become available, the conclusions drawn here mightchange slightly.

10.8.1 Commonalties among Scenario Findings

A comparison of similarities among major driving forces ofchange, key ecosystem services, and processes described bythe sub-global scenarios can help to enhance understandingof how these forces work and interact. While the scenarioexercises in many cases differed in their conclusions, a fewcommon themes emerged. Governance was the one drivingforce that surfaced in almost all the scenario exercises con-sidered. Biodiversity was another issue that a number of sce-narios addressed because it relates to a number of ecosystemservices.

................. 11474$ CH10 10-17-05 15:46:30 PS


10.8.1.1 Key Uncertainties: Governance

Governance issues were the only major uncertainty thatsurfaced in almost all scenario exercises considered. The na-ture of different governance regimes was used by many as-sessments as a major driving force to differentiate betweenscenario trajectories. In the multiscale SAfMA sub-globalassessment, where the scenarios at different scales were de-veloped independently, governance was identified as a keybifurcation at all scales. Social equity and the distribution ofwealth was the second major axis of uncertainty in SAfMA.

In the scenarios built by the Caribbean Sea assessment,the main bifurcation that emerged centered on the issue ofeconomic diversification: continued reliance on tourism asthe main source of income for the region versus economicdiversification to increase the resilience of economic sys-tems. The issue of economic diversification in the Carib-bean was strongly connected to a set of governance issues,such as regional cooperation, trade negotiations, and insti-tution building.

The first set of national scenarios developed by the Por-tugal assessment identified the attitude of society toward theenvironment and its understanding of environmentalchange as one of the main driving forces. Here, an environ-mentally friendly attitude was translated into lifestyles andgovernance systems and institutions that try to find a bal-ance between environmental protection and development.The second main driving force for these scenarios, land usemeasures and agricultural practices, were seen to be influ-enced by institutional structures and polices such as theCommon Agricultural Policy of the European Union.

The draft scenarios for Bajo Chirripo also used gover-nance as one major factor affecting the future of the indige-nous communities in the area. Clarifying their rights undera new law that was discussed with the Costa Rican govern-ment at the time of the scenario exercise was described asone of the main bifurcation points for the future.

Besides governance, other factors, such as climatechange at the global level or land use change at more local/regional levels, were also considered in various sub-globalassessments. However, there was a much higher degree ofvariability among the other drivers considered by the sub-global assessments.

10.8.1.2 Key Ecosystem Services

The specific ecosystem services addressed varied widelyamong the various sub-global assessments. Table 10.6 listssome ecosystem services that were mentioned as being im-portant. Two services immediately stand out: biodiversityand water quantity and quality. The prospects for biodiver-sity were clearly an important issue in many sub-global sce-narios, being explicitly mentioned by five out of the eightsub-global assessments examined here. Water quantity andquality was also an important service, mentioned by six outof eight assessments. However, the water issue covered alarge variety of different aspects including flooding,drought, irrigation, and salinity, drawing a picture that isless coherent than for biodiversity. A number of serviceswere mentioned by only one or two sub-global assessments,including soil protection, tourism, and landscape aesthetics.

PAGE 253

Table 10.6. Key Ecosystem Services Addressed in the Scenariosof Various Sub-global Assessments

Ecosystem Service Sub-global Assessment

Biodiversity (pollination, pest control, SAfMA, Caribbean Sea, Portugal,genetic resources, habitat Bajo Chirripo, India Localregeneration)

Water quality and quantity Goulburn-Broken Catchment,Wisconsin, SAfMA, Portugal, BajoChirripo, San Pedro de Atacama

Soil protection Portugal

Landscape aesthetics Wisconsin, Portugal, San Pedro deAtacama

Recreation/Tourism San Pedro de Atacama, CaribbeanSea

10.8.1.2.1 Biodiversity

Introducing policy and legal measures to protect biodiver-sity was seen in some regions as having a major impact onhow biodiversity will be used in the future. Draft scenariosin both Portugal and India Local described this kind ofresponse option to current biodiversity decline as major de-cision points. In Portugal, a move toward payments for en-vironmental services to farmers due to changes in the EUCAP was seen as having a positive impact on biodiversity.In India, the introduction of a National Biodiversity Actopened up new prospects for biodiversity conservation, al-though it was recognized that the measure needed to becoupled with strengthening the institutional capacities forenforcing and monitoring the act and for collecting andmanaging data. At the same time resolving conflicts be-tween developmental and biodiversity conservation goals atthe local level were seen as an important challenge in thefuture.

Climate change was seen to hold significant uncertain-ties for biodiversity in a number of scenario exercises.Changing sea temperatures in the Caribbean are likely toinfluence primary production and may have surprising ef-fects on populations of fish, corals, and other keystonespecies. Climate change is also an important stressor forsouthern African ecosystems. But even without accountingfor its impacts, in all SAfMA Regional scenarios, biodiver-sity will decline over the next three decades, although therate of decline and expected level of stabilization differ de-pending on the scenario assumptions.

Technological development was described in CaribbeanSea as having the potential to change the regional statusquo for biodiversity in novel and possibly unexpected ways.Better bioprospecting technologies could lead to increasedpharmaceutical involvement in several poorer regions of theworld. In the Caribbean, the implications of better fishingtechnologies and fish-tracking devices, such as allowinggreater and more selective exploitation of fish stocks forfood, were explored. Improvements in monitoring and en-forcement technology make it easier to manage fish stockssustainably.

................. 11474$ CH10 10-17-05 15:46:31 PS


10.8.1.2.2 Water

Water quality and quantity are other examples of ecosystemservices addressed explicitly in many sub-global assessments.While in Wisconsin, for example, water quality was of par-ticular importance, the San Pedro de Atacama and theSAfMA scenario exercises focused mainly on water avail-ability, with each identifying a threat to the availability ofwater resources in their respective areas.

These are but examples of the variety of ecosystem ser-vices that were considered. Even when categorized underthe same common denominator, the focus was often differ-ent in different places, while in most sub-global assessmentsat least one ecosystem service was unique. The list of eco-system services will undoubtedly grow as more sub-globalassessments produce results, but it is also anticipated thatmost sub-global scenario exercises will include commonecosystem services such as water, biodiversity and possiblytourism.

10.8.2 Comparing Findings across Scales

10.8.2.1 Differences in Scenarios across Scales

Table 10.7 shows how three sub-global scenarios exercises(SAfMA, Caribbean Sea, and Portugal) can be mapped ontothe MA global scenarios exercises based on the commonelements among them. The process of how the specific sce-narios were developed differed between the exercises; how-ever, they all contain specific aspects that also occur in theglobal scenarios. These include either common basic story-line elements across the exercises (that appear for examplein both the global MA scenarios and the Portugal scenarios)or similar assumptions and ideas (for example, a globalizedversus a fragmented world), such as in the global MA,SAfMA, and Caribbean Sea scenarios. Our analysis showsthat a number of common themes can be found acrossscales. These themes could be interpreted and played outdifferently at each geographical level, as the comparison

Table 10.7. Cross-mapping of SAfMA, Caribbean Sea, and Portugal Scenarios onto the MA Global Scenarios

SAfMA Scenarios Portugal ScenariosCaribbean Sea

MA Global Gorongosa- National National Sistelo

Scenarios

Scenarios Regional Gariep Livelihoods Marromeu (first phase) (second phase) (local)

Global African Policy Betterment Devolution Neo-plantation- Evolution of Global AbandonmentOrchestration Partnership Reform Economy Continuity Orchestration

(downscaled)Market Green PatronageForces Engineering

Techno Quality over Quantity Celestial Portugal TechnoGarden Garden

(downscaled)

Adapting African Local Stagnation Diversify Together Bucolic Portugal Adapting RejuvenationMosaic Patchwork Resources Mosaic

(downscaled)

Order from Fortress Growing Portugal at its Order fromStrength World Asymmetries Worst Strength

(downscaled)

PAGE 254

between the Caribbean Sea and the MA global scenariosdemonstrates. Both exercises used the degree of intercon-nectedness with the world as one of the main differencesbetween the portrayed future worlds.

10.8.2.2 Differences in Perceptions across Scales

One interesting outcome of comparing the scenarios acrossscales is that the analysis reveals how the interpretation ofscenario outcomes by stakeholders can differ across scales.What is seen as a beneficial outcome by stakeholders at onescale can be interpreted differently by stakeholders at an-other scale.

The Portugal assessment describes this tension in a com-parison of its national and local scenarios: One local scenarioin Portugal is characterized by the abandonment of agricul-tural fields and rural outmigration. Agricultural terraces arereplaced by oak forests, with the loss of local provisioningservices but a likely improvement in biodiversity. For localpolicy-makers this land abandonment scenario is undesir-able. However, in the national and global context, the sce-nario could be part of a Global Orchestration type scenario(described in Appendix 10.1), which is viewed by policy-makers at these coarser scales as having the highest directbenefits to human well-being.

The Caribbean Sea scenarios demonstrate this differencein scenario interpretations as well. The Caribbean region isconsidered particularly vulnerable to external forces becauseof its current and historically high reliance on internationalmarkets and capital, and the effects of economies of scale onsmall islands and developing states. While a well-connectedworld that emphasizes the equitable distribution of wealthmay hold benefits for poorer regions of the world, it alsoholds substantial risks. For the Caribbean region, which isalready highly connected to international markets and isparticularly dependent on international tourism, a GlobalOrchestration-type scenario would most likely result in fur-ther dependence on these external markets. A change in

................. 11474$ CH10 10-17-05 15:46:32 PS


tourism activities for any reason, and particularly if it wereabrupt, would have significant negative impacts on the re-gion’s economy.

Another example can be taken from SAfMA. Themultiscale structure of the SAfMA assessments showed thatcertain responses or developments at coarser scales are expe-rienced as surprises or shocks at local scales (for example, ifmega parks and large irrigation schemes are implementedwithout adequate local stakeholder participation and con-sideration of impacts). The SAfMA regional and local scalescenarios also suggest that general trends in ecosystem ser-vices at the regional and basin-scale may be reversed inparticular local situations. The SAfMA results are also inter-esting with respect to comparing global and regional inter-pretations of scenario outcomes. The greatest direct benefitsto human well-being are generally expected under scenarioscorresponding to the Global Orchestration scenario. How-ever, a key uncertainty in southern Africa is the degree ofsocial equity that would prevail under this scenario. Whileeconomies are expected to strengthen, the degree to whichthe benefits of economic growth are distributed within so-ciety is seen as an important bifurcation point, as evidencedby the different variations of the Global Orchestration sce-nario in SAfMA Regional. Without specific emphasis onpolicies that ensure social equity, it appears unlikely thateconomic growth will hold substantial benefits for the poor.The SAfMA results here reinforce one of the messages ofthe Global Orchestration scenario.

These examples show that the comparison of scenariointerpretations across scales can reveal useful insights for de-veloping response options for different stakeholders underthe various scenarios developed. They show where differ-ences between stakeholder priorities or worldviews lie, andcan therefore be used to analyze potential areas of conflictbetween them.

10.8.3 Summing Up: The Importance ofStakeholders

The comparison of findings across the sub-global scenarioexercises reveals more differences than commonalities inspecific results. Given the high degree of variability amongthe sub-global assessments that were considered here, thiswas to be expected. Nevertheless a number of emergingcommon themes can be used to ground-truth coarser scalescenarios. In particular, the importance of governance as amain driver in all sub-global assessments should be checkedagainst assumptions in the global storylines.

More importantly, the set of scenarios for which resultswere available showed that each strategy has winners andlosers, which vary depending on the scale examined. Sce-narios that appear beneficial at the national scale may holdsubstantial losses and threats for certain communities at finerscales. Identification of winners and losers in each scenariois an important step in guiding future responses. The inclu-sion of stakeholders in scenario development and validationprocesses helps make explicit the circumstances underwhich winners and losers emerge. The comparison of sce-nario findings across scales also demonstrates that scenario

PAGE 255

interpretations differ across scales. In the context of the pur-poses and methods employed to develop scenarios, directlyinvolving end-users and stakeholders in the process, theseconclusions suggest a need for more comprehensive en-gagement with users and a more systematic analysis of theirperceptions.

10.9 Conclusions and Recommendations forFuture Assessments

10.9.1 Strength in Diversity

The MA sub-global assessments comprise a unique collec-tion of scenario-building experiences covering several con-tinents and a wide variety of local, national, and regionalcontexts. The most outstanding feature of this collection isthe diversity of goals and of approaches to building, refin-ing, and communicating scenarios. At the same time, thevarious assessment groups have kept the goals of assessingecosystems services and human well-being largely in focus.The lesson is that there is no single clearly superior way toconduct scenarios. As noted, constructing storylines usingone or two main uncertainties as axes worked well for someassessments, but inhibited creativity for some others. Pro-spective scenario-builders and users can reflect upon a listof questions, such as the ones shown in Figure 10.1, toguide their processes. This final section highlights some im-portant points not fully considered in previous sections.

10.9.2 Lessons Learned

The most significant strength of the entire endeavor is thecommon methodology that links all sub-global assessments.Given the variety of assessments, important conclusions onhow best to construct scenarios can be drawn. Already dis-cussed were recommendations on how linking global andsub-global scenarios could be made successful, and whenthere are good reasons not to attempt a complete link. Similarrecommendations follow in this section for other method-ological aspects, such as stakeholder involvement. Engaginglocal stakeholders is another important strength of mostsub-global assessments. In particular, the use of participatoryapproaches, and thus the incorporation of key decision-makers in the process, was attempted in most assessments.

A problem encountered by all assessments was the rela-tively poor link between the various sub-global scales. All as-sessments recognized the importance of nesting across scales,and all devoted some attention to key factors at differentscales. However, few assessments focused on cross-scale in-teractions and feedbacks, with the notable exceptions ofSAfMA Regional and planned work in Portugal. In mostassessments, the emphasis was at the level of decision-making, in other words at the political or social scale, ratherthan the ecological scale (scales of time and space). It has notbeen possible in this chapter to cover all aspects of political/social scale and ecological scale, and it is expected that thetwo will not always coincide.

................. 11474$ CH10 10-17-05 15:46:33 PS


10.9.2.1 Integrating Scenarios with other AssessmentComponents

The degree to which the scenario-based activities in theassessments were linked to efforts to assess conditions andtrends, or explore responses, varied greatly among sub-global assessments. Integration was excellent for a numberof sub-global assessments, such as San Pedro de Atacama(management options) and SAfMA (condition and trends),but could have been improved for many others. A numberof sub-global assessments, including SAfMA Regional, ac-knowledged specifically that scenarios were a relativelysmall part of the assessment. Some assessments regarded sce-narios primarily as a useful tool to communicate with localdecision-makers rather than as a result in their own right.Elevating the importance of scenario development couldhave consequences for methods employed, such as stake-holder selection, scale selection, and dealing with uncer-tainties.

Nevertheless, the use of narrative scenarios developedthrough participatory methods is relatively new to the sci-entific world, and therefore full realization of the potentialgains should not yet be expected. It should be noted thatscenario development is being assigned increasing impor-tance in the process of testing the robustness of policy mea-sures, and was prominent in many sub-global assessments(including Portugal, SAfMA Regional, and Sweden). Theconnection between this type of scenario analysis and otheranalyses, however, is still rather weak. A method to linkqualitative stories to quantitative models is clearly needed,although some progress is being made in the sub-global as-sessments in Western China, Downstream Mekong, and La-guna Lake Basin. More attempts and examples are neededto convince future sub-global assessments that scenario de-velopment extends the quantitative or forecasting methodsin a positive and meaningful way, and that it also helpsdecision-makers to think, talk, plan, and act imaginativelyin pursuit of a more sustainable society.

10.9.2.2 Improving the Link between Global and Sub-globalScenarios

The greatest challenge in trying to link global and sub-global scenarios is increasing understanding in many (verylocal) assessments of how such large-scale developments canbe translated to the local level. There are many solutions tobridge this gap between global and local. Most of them in-volve a ‘‘translation’’ of the global scenarios to make themmore applicable to the local situation:• Translate ‘‘facts’’ to ‘‘feelings.’’ Local stakeholders will be

less overwhelmed by global developments if they are notpresented as facts but as underlying assumptions and(changes in) worldviews. For instance, the presentationof the Techno Garden scenario could focus more on‘‘society’s ability to invent’’ (see Appendix 10.1) and lesson technical solutions that are imposed on a reactivesociety.

• Translate all multiscale aspects. Developments have to bepresented to local stakeholders not only for a smallergeographical region, but also for shorter time scales and

PAGE 256

for shorter human decision-making scales. Local stake-holders do not relate as easily to the effects of globalwarming, as they would to three consecutive years ofdrought.The need for translation relates directly to the primary

goal for which scenarios are constructed. At the global level,scenarios are mainly used as a communication tool and toinform decision-making on important future trends and un-certainties. At the local level, scenarios have been utilizedmore as a direct tool for engaging people from differentsectors of society, and leveraging expertise in decision-making processes. In general, the more local the scale, themore scenarios become a tool for empowerment.

The above implies that the challenge lies both with thesub-global assessments that need to make an effort to down-scale global storylines, and with the global assessments thatcould improve the manner in which their scenarios are pre-sented. Scenarios developed both at global and sub-globalscales could have benefited from some additional iteration.Only with these methodological efforts can we begin tolink the rich variety of local and regional stories and globaldevelopments.

Despite the complexity that these mismatches introduceand the fact that ‘‘multiscale’’ considerations are not alwaysimportant for each sub-global assessment, the MA effort hasstimulated renewed interest in the potential for multiscalescenario development. It potentially has much greater localpolicy relevance than most previous global environmentalassessments.

10.9.2.3 Regional Diversity, Multiple Scales, and Ecology

The diversity of local and regional scenarios being devel-oped by the sub-global assessments represents a major chal-lenge to the global assessment community and it may oftenbe difficult, and even counter-productive, to tightly con-strain (or link) scenarios developed at these sub-global levelswith those at higher or lower levels. On the other hand,checking for inconsistencies in major assumptions (‘‘dis-cordances across scale’’) may help interpret regional varia-tion and improve scenarios at both global and sub-globalscales. How much attention needs to be paid to these cross-scale concerns ultimately depends on how inconsistent thekey processes of change are for the particular ecosystem ser-vices assessed.

In most sub-global assessments, the interactions betweenecosystems, their services, and human well-being were ratherpoorly documented, and their incorporation in the scenar-ios was relatively insignificant. There are difficult decisionsto be made in the design of scenario exercises, betweenfocusing on a few services that can well be understood andtrying to realize a better appreciation of interaction betweena wider range of services in a more comprehensive, but lessdetailed assessment. The currently advocated methods ofdeveloping holistic, integrated, participatory scenariosmight not be the best to address the objective of the sce-nario exercises. When the main objective of developingscenarios is direct decision support, and the uncertaintiesand possible futures are pre-defined, complex participatoryprocesses may, in particular, be unnecessary.

................. 11474$ CH10 10-17-05 15:46:34 PS


10.9.2.4 Concluding Remarks

There are four main conclusions from this chapter, based onthe unique source of information that the large and growingnumber of sub-global assessment scenarios represent:• Of all the approaches to conducting scenarios, none is

clearly superior to the others.• Sub-global scenarios allow for the critical evaluation of

local variations that cannot be properly assessed at theglobal scale alone.

• Scenarios developed at global and sub-global scalescould have benefited from additional iteration and inter-action across scales. Only with additional methodologi-cal efforts can we begin to link the rich variety of sub-global and global scenarios.

• In relation to global assessment frameworks, the MA isunusual in its emphasis on ecology. Even so, future sce-nario activities should pay even greater attention to eco-logical processes.Finally, we must acknowledge that the level of skill in

conducting scenario exercises and then integrating thesewith assessments of conditions and trends, and response op-tions, is still growing. The sub-global assessments representa heterogeneous collection of approaches, focal issues, andscales, which is an encouraging start for much more experi-mentation with scenarios and related future-exploringmethods.

Appendix 10.1. Summary of Global Scenarios inthe Millennium Ecosystem AssessmentThe Global Orchestration scenario depicts a globally con-nected society in which policy reforms that focus on globaltrade and economic liberalization are used to reshape econ-omies and governance, emphasizing the creation of marketsthat allow equitable participation and provide equitable ac-cess to goods and services. These policies, in combinationwith large investments in global public health and the im-provement of education worldwide, generally succeed inpromoting economic expansion and lifting many peopleout of poverty into an expanding global middle class. Supra-national institutions in this globalized scenario are wellplaced to deal with global environmental problems such asclimate change and fisheries decline. However, the reactiveapproach to ecosystem management makes people vulnera-ble to surprises arising from delayed action. While the focusis on improving the well-being of all people, environmentalproblems that threaten human well-being are only consid-ered after they become apparent. Growing economies,expansion of education, and growth of the middle class leadto demands for cleaner cities, less pollution, and a morebeautiful environment. Rising income levels bring aboutchanges in global consumption patterns, boosting demandfor ecosystem services, including agricultural products suchas meat, fish, and vegetables. Growing demand for theseservices leads to declines in other ones, as forests are con-verted into cropped area and pasture and the services theyformerly provided decline. The problems related to increas-ing food production, such as loss of wildlands, are not ap-

PAGE 257

parent to most people who live in urban areas. Theytherefore receive only limited attention. Global economicexpansion expropriates or degrades many of the ecosystemservices poor people once depended on for survival. Whileeconomic growth more than compensates for these lossesin some regions by increasing the ability to find substitutesfor particular ecosystem services, in many other places, itdoes not. An increasing number of people are affected bythe loss of basic ecosystem services essential for human life.While risks seem manageable in some places, in other placesthere are sudden, unexpected losses as ecosystems crossthresholds and degrade irreversibly. Loss of potable watersupplies, crop failures, floods, species invasions, and out-breaks of environmental pathogens increase in frequency.The expansion of abrupt, unpredictable changes in ecosys-tems, many with harmful effects on increasingly large num-bers of people, is the key challenge facing managers ofecosystem services.

The Order from Strength scenario represents a regionalizedand fragmented world that is concerned with security andprotection, emphasizes primarily regional markets, and payslittle attention to common goods. Nations see looking aftertheir own interests as the best defense against economicinsecurity, and the movement of goods, people, and infor-mation is strongly regulated and policed. The role of gov-ernment expands as oil companies, water utilities, and otherstrategic businesses are either nationalized or subjected tomore state oversight. Trade is restricted, large amounts ofmoney are invested in security systems, and technologicalchange slows due to restrictions on the flow of goods andinformation. Regionalization exacerbates global inequality.Treaties on global climate change, international fisheries,and trade in endangered species are only weakly and hap-hazardly implemented, resulting in degradation of theglobal commons. Local problems often go unresolved, butmajor problems are sometimes handled by rapid disaster re-lief to at least temporarily resolve the immediate crisis.Many powerful countries cope with local problems by shift-ing burdens to other, less powerful ones, increasing the gapbetween rich and poor. In particular, natural resource–intensive industries are moved from wealthier nations topoorer, less powerful ones. Inequality increases considerablywithin countries as well. Ecosystem services become morevulnerable, fragile, and variable in Order from Strength. Forexample, parks and reserves exist within fixed boundaries,but climate changes around them, leading to the unin-tended extirpation of many species. Conditions for cropsare often suboptimal, and the ability of societies to importalternative foods is diminished by trade barriers. As a result,there are frequent shortages of food and water, particularlyin poor regions. Low levels of trade tend to restrict thenumber of invasions by exotic species; ecosystems are lessresilient, however, and invaders are therefore more oftensuccessful when they arrive.

In the Adapting Mosaic scenario, regional watershed-scaleecosystems are the focus of political and economic activity.This scenario sees the rise of local ecosystem managementstrategies and the strengthening of local institutions. Invest-ments in human and social capital are geared toward im-

................. 11474$ CH10 10-17-05 15:46:35 PS


proving knowledge about ecosystem functioning andmanagement, which results in a better understanding of re-silience, fragility, and local flexibility of ecosystems. Thereis optimism that we can learn, but humility about preparingfor surprises and about our ability to know everything aboutmanaging ecosystems. There is also great variation amongnations and regions in styles of governance, includingmanagement of ecosystem services. Some regions exploreactively adaptive management, investigating alternativesthrough experimentation. Others use bureaucratically rigidmethods to optimize ecosystem performance. Great diver-sity exists in the outcome of these approaches: some areasthrive, while others develop severe inequality or experienceecological degradation. Initially, trade barriers for goods andproducts are increased, but barriers for information nearlydisappear (for those who are motivated to use them) due toimproving communication technologies and rapidly de-creasing costs of access to information. Eventually, the focuson local governance leads to failures in managing the globalcommons. Problems like climate change, marine fisheries,and pollution grow worse, and global environmental prob-lems intensify. Communities slowly realize that they cannotmanage their local areas because global and regional prob-lems are infringing on them, and they begin to develop net-works among communities, regions, and even nations tobetter manage the global commons. Solutions that were ef-fective locally are adopted among networks. These net-works of regional successes are especially common insituations where there are mutually beneficial opportunitiesfor coordination, such as along river valleys. Sharing goodsolutions and discarding poor ones eventually improves ap-proaches to a variety of social and environmental problems,ranging from urban poverty to agricultural water pollution.As more knowledge is collected from successes and failures,provision of many services improves.

The TechnoGarden scenario depicts a globally connectedworld relying strongly on technology and highly managed,often engineered ecosystems to deliver ecosystem services.Overall efficiency of ecosystem service provision improves,but it is shadowed by the risks inherent in large-scale hu-manmade solutions and rigid control of ecosystems. Tech-nology and market-oriented institutional reform are used toachieve solutions to environmental problems. These solu-tions are designed to benefit both the economy and theenvironment. These changes co-develop with the expan-sion of property rights to ecosystem services, such as requir-ing people to pay for pollution they create or paying peoplefor providing key ecosystem services through actions suchas preservation of key watersheds. Interest in maintaining,and even increasing, the economic value of these propertyrights, combined with an interest in learning and informa-tion, leads to a flowering of ecological engineering ap-proaches for managing ecosystem services. Investment ingreen technology is accompanied by a significant focus oneconomic development and education, improving people’slives and helping them understand how ecosystems maketheir livelihoods possible. A variety of problems in globalagriculture are addressed by focusing on the multifunctionalaspects of agriculture and a global reduction of agricultural

PAGE 258

subsidies and trade barriers. Recognition of the role of ag-ricultural diversification encourages farms to produce a va-riety of ecological services rather than simply maximizingfood production. The combination of these movementsstimulates the growth of new markets for ecosystem ser-vices, such as tradable nutrient runoff permits, and the de-velopment of technology for increasingly sophisticatedecosystem management. Gradually, environmental entre-preneurship expands as new property rights and technolo-gies co-evolve to stimulate the growth of companies andcooperatives providing reliable ecosystem services to cities,towns, and individual property owners. Innovative capacityexpands quickly in developing nations. The reliable provi-sion of ecosystem services as a component of economicgrowth, together with enhanced uptake of technology dueto rising income levels, lifts many of the world’s poor intoa global middle class. Elements of human well-being associ-ated with social relations decline in this scenario due togreat loss of local culture, customs, and traditional knowl-edge and the weakening of civil society institutions as anincreasing share of interactions take place over the Internet.While the provision of basic ecosystem services improvesthe well-being of the world’s poor, the reliability of theservices, especially in urban areas, become more critical andis increasingly difficult to ensure. Not every problem hassuccumbed to technological innovation. Reliance on tech-nological solutions sometimes creates new problems andvulnerabilities. In some cases, societies seem to be barelyahead of the next threat to ecosystem services. In such casesnew problems often seem to emerge from the last solution,and the costs of managing the environment are continuallyrising. Environmental breakdowns that affect large numbersof people become more common. Sometimes new prob-lems seem to emerge faster than solutions. The challengefor the future is to learn how to organize socioecologicalsystems so that ecosystem services are maintained withouttaxing society’s ability to implement solutions to novel,emergent problems.

ReferencesAlcamo, J., R. Leemans, and E. Kreileman (eds.), 1998: Global Change Scenarios

of the 21st Century: Results from the IMAGE 2.1 Model. Elsevier, Kidlington,UK.

Bennett, E.M., S.R. Carpenter, G.D. Peterson, G.S. Cumming, M. Zurek,and P. Pingali, 2003: Why global scenarios need ecology. Frontiers in Ecologyand the Environment, 1, 322–329.

Bennett, E.M. and M. Zurek, 2004: Proposal for scenario development. Inter-nal report, updated draft, MA Scenarios Working Group.

Biggs, R., E. Bohensky, P.V. Desanker, C. Fabricius, T. Lynam, et al. 2004:Nature Supporting People. The Southern African Millennium Ecosystem Assessment.Integrated Report. Council for Scientific and Industrial Research, Pretoria,South Africa.

Carpenter, S.R., E.A. Levitt, G.D. Peterson, E.M. Bennett, T.D. Beard, J.A.Cardille, and G.S. Cumming, 2003: Scenarios for the Future of the NorthernHighland Lake District. Center for Limnology, Madison, University of Wis-consin.

Chermack, T., S. Lynham, and W. Ruona, 2001: A review of scenario plan-ning literature. Futures Research Quarterly, 17(2).

Coates, J., 2000: Scenario planning. Technological Forecasting and Social Change,65, 115–123.

Cosgrove, W. and F. Rijsberman, 2000: World Water Vision: Making WaterEverybody’s Business. Earthscan Publications, Ltd., London.

................. 11474$ CH10 10-17-05 15:46:36 PS


CSIRO Sustainable Ecosystems, 2003: Natural Values: Exploring Options forEnhancing Ecosystem Services in the Goulburn Broken Catchment. CSIRO Sus-tainable Ecosystems, Canberra, Australia.

Davis, G., 1998: Creating scenarios for your company’s future. In: CorporateEnvironmental, Health, and Safety Excellence: Bringing Sustainable DevelopmentDown to Earth, New York.

Dreborg, K.H., 1996: Essence of backcasting. Futures, 28, 813–828.Gallopın, G., A. Hammond, P. Raskin, and R. Swart, 1997: Branch Points:

Global Scenarios and Human Choice, Pole Star Series Report No. 7, StockholmEnvironment Institute, Stockholm.

IPCC, 2001: Climate Change 2001: Mitigation. Third Assessment Report of theIntergovernmental Panel on Climate Change. Cambridge University Press,Cambridge, UK.

Kahane, A., 1992: The Mont Fleur scenarios. Deeper News, 7. Available atwww.gbn.org.

Kahane, A., 1998: Changing the winds. Whole Earth, 96, 77–81.Kahn, H. and A. Weiner, 1967: The Year 2000: A Framework for Speculation on

the Next Thirty-Three Years. Macmillan, New York.Kok, K., M. Patel, D.S. Rothman, and G. Quaranta (2006): Multi-scale narra-

tives from an IA perspective: Part II. Participatory local scenario develop-ment. Futures, 38(3).

MA (Millennium Ecosystem Assessment), 2003: Ecosystems and HumanWell-Being: A Framework for Assessment. Island Press, Washington, DC, 245pp.

Nakicenovic, N., J. Alcamo, G. Davis, B. de Vries, J. Fenhann, et al. 2000:Special Report on Emissions Scenarios: A Special Report of Working Group III ofthe Intergovernmental Panel on Climate Change. Cambridge University Press,Cambridge, UK.

Pereira, H.M., T. Domingos, and L. Vicente, 2004: Assessing ecosystem ser-vices at different scales in the Portugal Millennium Ecosystem Assessment.Paper presented at Bridging Scales and Epistemologies: Linking Local Knowledgeand Global Science in Multi-Scale Assessments, March. Alexandria, Egypt.

Peterson, G.D., T.D. Beard Jr., B.E. Beisner, E.M. Bennett, S.R. Carpenter,G.S. Cumming, C.L. Dent, and T.D. Havlicek, 2003a: Assessing future eco-

PAGE 259

system services: A case study of the Northern Highlands Lake District, Wis-consin. Conservation Ecology, 7(3), 1. Available at www.consecol.org/vol7/iss3/art1.

Peterson, G.D., G.S. Cumming, and S.R. Carpenter, 2003b: Scenario plan-ning: A tool for conservation in an uncertain world. Conservation Biology, 17,358–366.

Raskin, P., T. Banuri, G. Gallopın, P. Gutman, A. Hammond, R. Kates, andR. Swart, 2002: Great Transition: The Promise and Lure of the Times Ahead.Pole Star Series Report No. 10, Stockholm Environment Institute, Boston.

Robinson, J., 1982: Energy backcasting: A proposed method of policy analysis.Energy Policy, 10, 337–345.

Robinson, J., 2003: Future subjunctive: Backcasting as social learning. Futures,35, 839–856.

Rotmans, J., M. van Asselt, C. Anastasi, S. Greeuw, J. Mellors, et al. 2000:Visions for a sustainable Europe. Futures, 32, 809–831.

Schwartz, P., 1996: The Art of the Long View. Doubleday, New York.Tansey, J., J. Carmichael, R. VanWynsberghe, J. Robinson, 2002: The future

is not what it used to be: Participatory integrated assessment in the GeorgiaBasin. Global Environmental Change, 12, 97–104.

UNEP, 2002: Global Environmental Outlook 2002. United Nations EnvironmentProgramme, Nairobi.

van der Heijden, K., 1996: Scenarios: The Art of Strategic Conversation. JohnWiley and Sons, Inc., New York.

van Notten, P.W.F., J. Rotmans, M.B.A. van Asselt, and D.S. Rothman, 2003:An updated scenario typology. Futures, 35, 423–443.

Verburg, P.H., W. Soepboer, R. Limpiada, M.V.O. Espaldon, and M. Sharifa,2002: Land use change modeling at the regional scale: The CLUE-s model.Environmental Management, 30, 391–405.

Wack, P., 1985: Scenarios: Uncharted waters ahead. Harvard Business Review,September-October, 73–89.

Wollenberg, E., D. Edmunds, and L. Buck, 2000: Using scenarios to make deci-sions about the future: Anticipatory learning for the adaptive co-managementof community forests. Landscape and Urban Planning, 47, 65–77.

................. 11474$ CH10 10-17-05 15:46:37 PS

sub-global scenarios · sub-global assessments differed greatly from the ma global scenarios....

Documents