Elasticity in ecosystem services: exploring the variable relationship between ecosystems and human well-being

Although ecosystem services are increasingly recognized as benefits people obtain from nature, we still have a poor understanding of how they actually enhance multidimensional human well-being, and how well-being is affected by ecosystem change. We develop a concept of “ecosystem service elasticity” (ES elasticity) that describes the sensitivity of human well-being to changes in ecosystems. ES Elasticity is a result of complex social and ecological dynamics and is context dependent, individually variable, and likely to demonstrate nonlinear dynamics such as thresholds and hysteresis. We present a conceptual framework that unpacks the chain of causality from ecosystem stocks through flows, goods, value, and shares to contribute to the well-being of different people. This framework builds on previous conceptualizations, but places multidimensional well-being of different people as the final element. This ultimately disaggregated approach emphasizes how different people access benefits and how benefits match their needs or aspirations. 
 
Applying this framework to case studies of individual coastal ecosystem services in East Africa illustrates a wide range of social and ecological factors that can affect ES elasticity. For example, food web and habitat dynamics affect the sensitivity of different fisheries ecosystem services to ecological change. Meanwhile high cultural significance, or lack of alternatives enhance ES elasticity, while social mechanisms that prevent access can reduce elasticity. 
 
Mapping out how chains are interlinked illustrates how different types of value and the well-being of different people are linked to each other and to common ecological stocks. We suggest that examining chains for individual ecosystem services can suggest potential interventions aimed at poverty alleviation and sustainable ecosystems while mapping out of interlinkages between chains can help to identify possible ecosystem service trade-offs and winners and losers. We discuss conceptual and practical challenges of applying such a framework and conclude on its utility as a heuristic for structuring interdisciplinary analysis of ecosystem services and human well-being.


INTRODUCTION
The Millennium Ecosystem Assessment (MA) conceptualized ecosystem services as contributions from nature to human wellbeing and drew on the "capability approach" proposed by scholars such as Amartya Sen (MA 2003, Robeyns 2005).This provided a broad conceptualization of well-being, including security, material assets, health, social relations, and freedom.In their recent review of well-being frameworks, Agarwala et al. (2014) describe how, in the context of ecosystem services, interest in the concept of multidimensional well-being evolved in response to limited mono-dimensional assessments, which were largely monetary and value-based.They argue that this is analagous to the development of a multidimensional understanding of poverty (Alkire 2007), which represents the opposite to well-being.They also acknowledge the challenge of engaging with such a more complex vision of people's lives.The concept of well-being is becoming increasingly central to research and policy on ecosystem services, and sustainability in general (Bizikova 2011).As Milner-Gulland et al. (2014) posit, engaging with human wellbeing may enable better outcomes for nature conservation by taking account of a broader spectrum of gains and losses as well as greater incorporation of people's views and needs.Although many different frameworks exist for the study of well-being (e.g., OECD 2011, Oxfam 2013), there seems to be consensus that any measure of well-being should include both an objective measure of a person's welfare, alongside a subjective evaluation of how a person thinks and feels about the quality of life they are experiencing (Gough andMcGregor 2007, UN 2012).This clearly calls for integrative approaches combining quantitative and qualitative data (McGregor et al. 2015).Despite this interest and Ecology and Society 21(2): 11 http://www.ecologyandsociety.org/vol21/iss2/art11/conceptual progress, significant gaps remain in our understanding of how ecosystems actually contribute to different people's wellbeing (Butler andOluoch-Kosura 2006, Carpenter et al. 2009) and to alleviation of multidimensional and dynamic poverty (Alkire 2007).This lack of understanding has particular implications for poor people who are often more reliant on ecosystems for their food, physical security, and livelihoods (Duraiappah 2004, Bizikova 2011) and disproportionately affected by changes in ecosystem services.
The ecosystem service concept alludes to a positive relationship between ecosystem quality and human well-being [1] .However, this relationship is complex and often indirect with the result that the well-being of people in particular places and times can be more or less coupled to ecosystem quality.For example, although it is clear at the aggregate global scale and in the long run, that humanity depends on the biosphere for survival (Dasgupta 2001), aggregate indicators of human well-being at the global scale appear to show that well-being has improved over recent years despite the ongoing degradation of ecosystems, a phenomenon referred to as the "environmentalist's paradox'" (Raudsepp-Hearne et al. 2010).At smaller scales, people may experience improvements in well-being in the face of ecosystem degradation, when this enhances the opportunities for human development (e.g., Wunder 2001).Conversely, conservation may enhance ecosystems with little benefit, or even harm to the well-being of local people, for example by exclusion from reserves (Dowie 2011, Kamat 2014) or where "ecosystem disservices" such as crop raiding by wild animals impact local farmers (Woodroffe et al. 2005).Such examples of a negative relationship between ecosystem health and human wellbeing may represent temporal or spatial effects, either that degradation will impact the well-being of people in the future, or in other places, or that benefits from ecological enhancement take time to materialize or are enjoyed by distant beneficiaries.However, what these examples demonstrate is that in particular places and times, and for particular people, the relationship between ecosystem quality and human well-being is variable and complex.
We can refer to the human-well-being impacts of ecological change using the concept of "elasticity," which captures the responsiveness of one variable to changes in another (York et al. 2003).For example, in economics, the price elasticity of demand captures how much demand will change in response to a change in price.Here we apply the concept of elasticity to ecosystem services and explore the elasticity of human well-being to ecosystem change (henceforth "ES elasticity"), i.e., how human well-being changes in response to increases or declines in ecosystem quality.We propose that studying ES elasticity challenges us to engage with the complexities and context dependency of the ecosystem-well-being relationship, and can facilitate a better understanding of the role of ecosystem services in human well-being.

INTRODUCING ES ELASTICITY
Elasticity describes how one variable changes in relation to changes in another, and thus ES elasticity can be formally represented as the following: ES Elasticity = change in well-being / change in ecosystem stocks Thus high, positive elasticity means that human well-being is strongly coupled to changes in ecosystem quality and people strongly benefit when ecosystem quality improves and suffer when quality declines.Examples of high, positive elasticity include when the loss of storm protection services from mangroves severely impacts coastal people's security (e.g., Das and Vincent 2009), or when successful conservation of charismatic species promotes the livelihoods of tour guides.Meanwhile, low elasticity means that changing ecosystem quality has minor or negligible impacts on well-being, for example, when people are decoupled from the ecosystem because of alternatives or substitutes.Negative ES elasticity describes situations, like the environmentalist's paradox, in which people's well-being actually improves as ecosystem quality declines or people suffer as ecosystem quality improves (Fig. 1a).For the rest of this paper we use the term "high ES elasticity" to refer to high positive ES elasticity.ES Elasticity will vary depending on the nature of the ecosystemwell-being link such that explaining or even forecasting ES elasticity challenges us to better understand both ecological and social dimensions, and context of such linkages.Elasticity has important implications for environmental management and poverty alleviation.Prescriptions for community-based conservation presuppose a self-evident and universally experienced positive elasticity that creates the necessary incentives for conservation (Roe et al. 2012).However, this may overlook or underplay the importance of delayed effects, trade-offs, and social disparities in terms of who benefits.Where conservation efforts are characterized by low or negative ES elasticity for some people, they may resist or sabotage such efforts.An example is the widespread low compliance with fisheries regulations and marine protected areas in the tropics (Wood et al. 2008).Thus failure to recognize or address low or negative ES elasticity can ultimately lead to failures in conservation initiatives (McClanahan 1999, Pascual et al. 2014).
ES Elasticity is also core to understanding human vulnerability to environmental change and factors that are critical to adaptation.Low elasticity for example, would be desirable, for example, where climate change impacts degrade ecosystems, to minimize the harm experienced by people as a result of this degradation.Conversely, high elasticities would be desirable in situations of ecological rehabilitation and recovery, so that people benefit as much as possible from improving ecosystem services and are also incentivized to support continued rehabilitation.
Although ES elasticity is defined as the sign and slope of a linear relationship between ecosystem quality and human well-being http://www.ecologyandsociety.org/vol21/iss2/art11/(Fig. 1a), there is little evidence to suggest that this relationship will be a straight line.Rather, research into social-ecological systems has shown nonlinear relationships and thresholds to be common (Liu et al. 2007, Barbier et al. 2008, Barrett and Constas 2014).For example, people's well-being may be relatively unaffected by initial losses of ecosystem quality, but below a certain threshold, further degradation leads to the breakdown of critical functions impacting well-being.This results in a nonlinear relationship in which elasticity is low when ecosystem condition is good but increases sharply (steeper line) when ecosystem condition is below a certain threshold (Fig. 1b).
Hysteresis and path dependency are also common features of complex social-ecological systems (Scheffer et al. 2001).For example, Figure 1c could illustrate a situation in which well-being benefits from a cultural ecosystem service, which in turn depends on skills and other intellectual assets that are lost when ecosystem quality falls below some threshold level (E 1 ).Subsequent recovery of ecosystem quality would not contribute to improving wellbeing in the absence of these assets, which may only recover (if at all) at higher levels (E 2 ) of ecosystem quality.Thus the expectation of nonlinear changes in social-ecological systems emphasizes the importance of context and history.ES elasticity may depend not only on the current state of ecosystems but also recent history.
ES elasticity can be applied to different units of analysis and at different scales.The environmentalist's paradox has been investigated as an aggregate system-level attribute at a global scale, but given the importance of individual experiences and circumstances for the understanding of well-being (Coulthard et al. 2011), research will also be needed that can capture the differences between groups of people, based in terms of their relationship to, and dependence on, different processes or components of the ecosystem.For example, ES elasticity will be higher for people with ecosystem-based livelihoods than people not directly dependent on local ecosystems.These individual elasticities will be affected by the range of individual assets and the institutional arrangements that determine how individuals and groups benefit from nature (e.g., Leach et al. 1999, Ribot and Peluso 2003, Hicks and Cinner 2014).When there are improvements in ecological quality people endowed with "mechanisms of access" (Ribot and Peluso 2003) stand to benefit more than those denied access.Thus disaggregated analysis of ES elasticity is likely to show contrasting patterns for different groups of people.Such that high positive and negative ES elasticity may be found among different stakeholders within the same system creating trade-offs and dilemmas in policy choices.
Although much research has focused on factors affecting the supply of ecosystem services, less attention has been given to the demand side (Lele et al. 2013, García-Nieto et al. 2013).Elasticity will be affected by these contextual factors, such as the availability of and access to alternatives to ecosystem services, such as nonecosystem-based employment, imported foods, technological innovations, or nonecosystem based social practices, that may compensate for declines in local ecosystem services (MA 2003) [2]  and will reduce ES elasticity for people who have access to them.Additionally, many ecosystem services may only contribute to well-being under particular circumstances (Andersson et al. 2015).For example, wild foods may only be important during times of food shortage, storm protection during storms, and cultural services in the context of a culture that regards them as significant.In such cases, ES elasticity would only be high when circumstances supporting the demand side of ecosystem services are present.

A CONCEPTUAL FRAMEWORK TO INTERROGATE ELASTICITY
The previous section highlighted that ES elasticity can be determined by a range of social, ecological, historical, or contextual factors.Thus advancing understanding of ES elasticity requires an interdisciplinary effort that integrates the strengths of existing literature on production, trade-offs, and valuation of ecosystem services and the growing literature on human well-being and its connection to natural resources.Taken alone, each of these branches of the ecosystem services literature is insufficient to explain ES elasticity because of a partial view over the ecosystem-well-being relationship.For example landscape-scale modeling of ecosystem services has led to an improved understanding of generation of different ecosystem services (e.g., Goldstein et al. 2012) but has generally not evaluated how these services actually contribute to people's well-being.Ecosystem service valuations based on aggregate measures of willingness-to-pay such as market prices, provide indications of the welfare value of ecosystem services, but are criticized for downplaying nonmonetary aspects of well-being, failing to disaggregate to reflect how benefits are distributed in society, and being based on market prices that are skewed to reflect preferences of the wealthy.Meanwhile frameworks for the study of well-being are now being applied to assess specific ecosystem service contributions to different aspects of life (e.g., Abunge et al. 2013), and the impacts of conservation (Milner-Gulland et al. 2014), but the majority of well-being research has limited appreciation for ecological dynamics and change.Meanwhile, studies that explicitly relate well-being to ecological change are often correlational and lack a specific theory of causation or experimental evidence of how particular ecosystem services actually contribute to well-being (e.g., McNally et al. 2011).
This section presents a conceptual framework that explicitly maps out the social and ecological links between ecosystems and wellbeing for different beneficiaries, and aims to understand how the well-being contribution of ecosystem services is shaped by people's individual condition and context over time.We acknowledge that a plethora of ecosystem service conceptual frameworks (e.g., Costanza et al. 2007, Bateman et al. 2011, Fisher et al. 2013, Reyers et al. 2013, Díaz et al. 2015) have been developed to facilitate thinking about the complex reality of ecosystem services.However, most frameworks generally have focused on specification of the ecological generation of ecosystem services to the detriment of understanding how they actually contribute to well-being (Fisher et al. 2013).
Our framework builds on the "cascade model" (Haines-Young and Potschin 2010), which was developed and adopted by TEEB (The Economics of Ecosystems and Biodiversity; Braat and de Groot 2012) and aims to conceptually clarify the steps by which ecosystems generate value.We augment this approach by returning to the MA's focus on multidimensional well-being rather than aggregate value as the end point.Thus we address some additional complexities highlighted by the more holistic model of http://www.ecologyandsociety.org/vol21/iss2/art11/Fisher et al. (2014) but incorporate these into a more linear flow like the cascade, while maintaining an emphasis on the disaggregation of beneficiaries (Daw et al. 2011).This framework shown in Figure 2 represents a chain of elements (a-f) that link from Ecosystem Stocks to Well-being Contribution.These are listed in Table 1.(For clarity, references to elements and multipliers are formatted as bold and italics, respectively).Each element has an influence on the next and the degree of this influence can be conceptualized as a component of ES elasticity.This breaks down overall ES elasticity between ecological stocks and well-being into subelasticities between each pair of elements.Elasticity between the elements (a-f) is influenced by five "multipliers" (i-v) that represent the things that explain or parameterize the relationship between adjacent elements.For example Flows (b) interact with Human Inputs (ii) to determine the quantity of Goods (c): Thus the multipliers are critical for determining the elasticity between adjacent elements and together, the ES elasticity of the overall chain between Ecosystem Stocks and Well-being Contribution.
The framework follows the cascade and MA frameworks, with arrows and an implied causality running from left (ecology) to right (well-being).This reflects our focus on elasticity and how ecosystem change can affect well-being.However, as emphasized in more social-ecological conceptualizations of ecosystem services (e.g., Reyers et al. 2013), causal influences in the other direction ("feedbacks" in Fig. 2), are important for the development of ecosystem quality and human well-being over Processes that determine who can access Goods and benefit from their value e.Share The amount of Value that each person/group actually benefits from v. Needs, Gaps, and Asprirations The contextually and personally determined needs and aspirations of a group/person that can be satisfied by ecosystem services f.Well-being Contribution The improvement in well-being experienced by a group/person as a result of their Share meeting their Needs Goods (c) are the things or experiences from ecosystems that are valued by people (Bateman et al. 2011).Goods are not solely http://www.ecologyandsociety.org/vol21/iss2/art11/biophysical, but are coproduced from Flows by Human Inputs (ii) (Lele et al. 2013) such as labor, capital, or the presence of people to appreciate or benefit from them.
The Value (d) of Goods is an aggregate indication of the total value of benefits generated by an ecosystem service, as calculated by standard valuation, although it could also be calculated in nonmonetary terms.Value is determined by societal process of Valorization (iii), such as the structure of markets, or cultural norms (Bateman et al. 2011).As such, Valorization represents the demand side of ecosystem services at an aggregate level and the contextual conditions that lead Goods to be of more or less interest and/or use to society in general.
Ecological Stocks, Flows, Goods, and Value (a-d) generally reflect the cascade model used for ecosystem service valuation (de Groot et al. 2010, Bateman et al. 2011) and are aggregate quantities.In contrast to the cascade model, Value (an aggregate quality) is not the final outcome, but is further mediated by Share and Well-being Contribution that focus on who accesses Value and how it contributes to their well-being (Coulthard et al. 2011, Daw et al. 2011).Thus the aggregate processes of Valorization are distinguished from the disaggregated processes of Access, and Needs, Gaps, and Aspirations that determine the Well-being Contribution of ecosystem services for different kinds of people.
Valorization may in some cases reflect the Needs, Gaps, and Aspirations of the same local people who benefit from Well-being Contribution.For example, in an assessment of local cultural ecosystem services Valorization is linked to both Access, and Needs, Gaps, and Aspiration of the same local people benefiting from the cultural ecosystem service (see sense of place example in Table 2).In contrast, Valorization of traded ecosystem services reflects motivations and aspirations of a different and often distant group of people.For example the market price of aquarium fish is determined by the aspirations of distant consumers as well as the availability of substitutes from other regions or aquaculture.This case illustrates why, for a disaggregated analysis, Valorization is considered independently from the Needs, Gaps, and Aspirations, of local people.

Share (e) is an explicit expression of the absolute amounts of
Value received by different people, and is determined by the Access (iv), which represents the assets, institutions, laws, social norms, and structures (such as class, gender, ethnicity) that give people the ability to benefit from ES (Leach et al. 1999, Ribot andPeluso 2003).Access reflects power dynamics, for example as people compete to capture Shares or when governance structures impose limits on Access to reduce impacts on Ecosystem Stocks.
Well-being Contribution (f), reflects how an ecosystem service actually translates into well-being outcomes for an individual or group of beneficiaries.Well-being Contribution incorporates an assessment of the well-being of different beneficiaries (size of the circles in Fig. 2) as well as the proportional contribution to wellbeing from ecosystem services.In this framework, well-being is portrayed as the combination of needs met (a quantitative measure of the extent to which an individual, within a household, is able to sufficiently meet a predetermined list of basic needs) and life satisfaction (the person's subjective evaluation of how they are doing); the latter is captured methodologically through life satisfaction scoring, and more in-depth interviews.
Assessment of a person's well-being, even combined with an assessment of Shares is not sufficient to evaluate how ecosystem services actually contribute to well-being.This requires understanding of the individual circumstances that mean that a Share translates into a Well-being Contribution.As emphasized by the capability approach of Sen and others, "different people need different amounts of and different kinds of goods to reach the same levels of well-being" (Robeyns 2005:97).Hence Needs, Gaps, and Aspirations (v) constitute the multiplier between Share and Well-being Contribution and aims to capture the individual circumstances that determine whether a Share contributes to wellbeing.Needs, Gaps, and Aspirations incorporates objective and subjective dimensions that are affected by internal factors, such as personal illness or disability and individual preferences, as well as external factors such as substitutes, e.g., imported food, and events, e.g., storms, that affect what an individual or group needs from ecosystems.Needs, Gaps, and Aspirations thus reflect the demand side of ecosystem services on a more individual basis than Valorization.
The distinction between Access and Needs, Gaps, and Aspirations are illustrated by comparing a rich person with Access to wild foods but who does not benefit because they have no Need for them to a poor person who has a Gap in their nutritional security, but lacks Access to wild foods and hence has no Share.In both cases, wild foods make a limited contribution to well-being, but this is due to Needs, Gaps, and Aspirations in one case and Access in the other.In this sense Access represents the ability of people to benefit from Share, whether or not that ability is realized, and echoes Sen's (2001) concept of capabilities as distinct from functionings.Where Needs, Gaps, and Aspirations are low, patterns of Access may not be obvious.However, if Needs, Gaps, and Aspirations increase, e.g., during a food shortage, patterns of Access may become more obvious as people struggle to benefit from the available Value.

EMPIRICAL APPLICATIONS OF THE CHAIN FRAMEWORK TO UNDERSTAND ELASTICITIES
We applied the chain framework to five different regulating, provisioning, and cultural coastal ecosystem services studied in East Africa by the SPACES (Sustainable Poverty Alleviation from Coastal Ecosystem Services, http://www.espa-spaces.org)project (Table 2, Figure 3, Box 1).The cases illustrate a range of biophysical and social factors that lead to high or low ES elasticity.Examples of low ES elasticity include shoreline protection from reefs where biophysical dynamics mean that wave attenuation is not sensitive in the short term to changes in ecological quality (low Stock-Flow elasticity).In the example of livelihoods for female traders, increasing Value of landed fish is unlikely to improve female traders' livelihood if they cannot access that fish due to competition with male traders with higher buying power (low Value-Share elasticity due to limited Access).The aquarium fishery provides examples of high ES elasticity deriving from ecological and social factors.Many aquarium fish species are dependent on temperature-sensitive corals and thus sensitive to impacts of coral bleaching (high Stock-Flow elasticity), while connection to global markets generate a high price for the fish (high Goods-Value elasticity).The shade from mangrove trees example suggests that gleaners would be sensitive to loss of mangrove trees (high Share-Well-being Contribution elasticity) because of their Needs, Gaps, and Aspirations for shade and a lack

Improves subjective wellbeing
Improves relational WB through belonging http://www.ecologyandsociety.org/vol21/iss2/art11/ of alternative shade.Identity and sense of place of fishers would be sensitive to an inability to maintain catches of large fish because of Valorization by the community (high Goods-Value elasticity) particularly for fishers whose identity and standing within the community depends on fishing (Access) and who lack alternative sources of pride or respect (lack of alternatives leading to high Share-Well-being Contribution elasticity).

Shoreline protection service
Between three and four million people on the East African coast likely benefit from shoreline protection because of coral reefs (Ferrario et al. 2014).The degree of wave attenuation is affected by the height of the reef as well as the species composition and abundance of live coral, relative to the abundance of eroding organisms, such as parrotfishes or urchins (Perry et al. 2008).The balance between these determines if a reef grows or erodes such that impacts such as coral bleaching can result in reefs starting to erode away.Nevertheless, ES elasticity of shoreline protection is low in the short term (Sheppard et al. 2005) because a high proportion of wave energy is dissipated by historically accreted reef (Ferrario et al. 2014), and because skeletons of complex corals remain after they have died.Over time coral skeletons do break down as a result of biological erosion or storms (Graham et al. 2007), leading to increased wave energy reaching the shore (Sheppard et al. 2005).ES elasticity would therefore be low at first, and increases in the elasticity may be avoided through management of bioeroding populations on the reef, or through natural recovery of corals.

Livelihood support for female traders
The roles of people in the coastal Kenyan fishery are highly gendered.Women do not participate in catching finfish and their primary role is buying, frying, and selling low-value fish to local consumers (Yang 2013, Matsue et al. 2014).In sites around Mombasa, competition between traders to access fish can be high, particularly when catches are low.Male traders, who tend to have greater capital and access to transport such as bicycles (Yang 2013) have priority access to larger and higher value species.Heavy fishing effort by illegal beach seines, although damaging to the ecosystem, generate high volumes of low-value fish (McClanahan et al. 2008), while protected areas, enforcement against illegal gears, and lower fishing effort increases catch rates, fish size, and individual fishers' revenues (McClanahan 2010).However, the gendered access conditions suggests that such ecological improvements might negatively impact the livelihoods of female traders because of lower total catches and a shift in catch composition to high-value species favored by male traders (Daw et al. 2015).

Shade from mangroves service
The way that mangroves are valued by people is influenced by culture, and the level of dependence on mangroves for livelihoods and well-being (James et al. 2013).Even within households, men and women value resources differently based on gendered resource use practices (UNEP-WCMC 2014).The apparently insignificant small stand of mangrove trees (low Ecosystem Stock) adjacent to Lalane village in northern Mozambique contributes to the well-being of certain groups.Women gleaners for example, spend many hours in the sun at low tide collecting shells.With little or no alternative shade in the vicinity, the mangroves improve gleaners' subjective well-being by providing a cool sanctuary from the heat, which increases enjoyment of the activity and improves social relations.Thus this service demonstrates a high ES elasticity despite the very limited Ecosystem Stocks; the removal of these trees would negatively impact on the well-being of gleaners because the shade they offer is accessible to gleaners (Access) and because of gleaners' need for, and lack of alternatives to (Needs and Gaps) the shade from these mangroves.

Aquarium trade:
The ornamental coral reef wildlife trade supports a multimillion dollar industry (Grey et al. 2005), and can provide significant levels of income to local fishers on coral reefs in developing countries (Máñez et al. 2014).Many of the finely branched coral species and coral dwelling fish prized by aquarium collectors are highly susceptible to coral bleaching and other stressors (Loya et al. 2001).Thus the aquarium trade demonstrates high ES elasticity; overcollection, or shifts in coral community composition due to bleaching can result in large reductions in provision of species to the aquarium trade (Flows) despite relatively small changes in overall ecosystem health (Ecosystem Stocks).

Identity
People often develop an identity associated with a place as a result of the activities and relationships that people engage with there (Stedman 2002).For example, the emergence of place-based identities may be associated with the benefits and activities that are realized from the presence or utilization of ecosystem services, such as fisheries goods or through the use of the environment for recreation (supported by the coastal protection service) or shelter (provided by mangroves).For example, individuals from a fishing community may strongly identify with the reputation of having skilled fishers, from a close and supportive community, who engage in successful community conservation initiatives, and boast a productive and reliable fishery.The identity in this case is based on the same Ecosystem Stocks, Flows and Goods as a fisheries livelihood service.But, identity is also dependent on specific attributes of the Good, i.e., consistently large landings, and gain meaning (Valorization) through a history of social or community engagement.Individuals who are fishers and who belong to the community have Access to this identity.How important fishing, community, and recognition is to them and the extent to which they feel a sense of identity or belonging to another social group (Needs, Gaps, and Aspirations) will determine how this pride, identity, and sense of belonging contributes to their subjective and relational well-being.Because of the dependence of this identity on environmental quality and social engagement, a drop in environmental quality or social participation can rapidly affect the validity of a person's placebased identity, suggesting a high ES elasticity.
The linear arrangement of this conceptual framework, draws specific attention to the connections and elasticity between ecosystems and human well-being.However it also risks a linear and isolated representation of different ecosystem services and blindness to complexity created by feedbacks and interactions between different ecosystem services as Norgaard (2010) warns.In reality ecosystem services are interlinked and bundled as shown by Figure 3, which illustrates the differential impact of coral bleaching on different fisheries and how food fish landings generate different kinds of Value (income, food) and how different groups Access different Shares of that Value and benefit to different extents.
Mapping how different types of benefits accrue to different people in this way is essential for understanding linkages and trade-offs.For example an increase in the price of fish in Figure 4 would represent an increase in Valorization of fish for income but would also decrease Access of poor consumers to fish for food.
Multiplier effects in local economies may indirectly benefit multiple beneficiaries, such as tourism enhancing prices for local producers who can in turn hire more labor.In some cases such indirect benefits may be important for the well-being of many people (e.g., Ashley and Mitchell 2007) in which case they would need to be included in the aggregate calculation of Value.Value chain analysis can then identify how different people have Access to different shares of this Value.

POLICY RELEVANCE OF ANALYSING CHAINS AND ELASTICITY
We suggest ES elasticity can be applied to understanding vulnerability of different social actors to change in ecosystems, and to identify and assess the opportunities and potential interventions to improve well-being.The policy implications depend on the particular circumstance, specifically if the ES elasticity is high or low (or negative) and whether ecosystem quality is declining or improving (Table 3).Vulnerabilities of people result from high elasticities to ecological decline, or from processes that change multipliers and reduce elasticity, e.g., loss of access.Opportunities occur in situations of high elasticities when ecological enhancement is possible or where interventions can change multipliers to increase the Well-being Contribution of ecosystem services (Table 3).
First, the chain framework applied to any individual ecosystem service can identify important multipliers (and points of ES elasticity) as opportunities to reduce vulnerability or enhance or maintain the contribution of ecosystem services to well-being,  4.
Second, understanding the elasticities within different chains and distribution of benefits as illustrated in Figure 4 facilitates an assessment of trade-offs linked to different interventions.This information can be analyzed with decision makers, for example, through use of scenarios exercises or "toy models" and other interactive techniques (e.g., Daw et al. 2015), to discuss the tradeoffs associated with different courses of action.This is particularly relevant to interventions such as Payments for Ecosystem Services, or Community-based Conservation, where simplistic assumptions about benefits and their distribution may lead to negative outcomes in terms of compliance and negate "success" in ecological and social terms (Pascual et al. 2014).
Application of this framework clearly poses challenges.The data requirements for populating the chains to analyze elasticities are considerable, requiring a coordinated interdisciplinary effort incorporating ecologists and a diversity of social scientists.In addition, practical application of the elasticity concept may be challenging in highly dynamic settings where elements and multipliers change very rapidly.In such circumstances, the framework may be best used as a frame to guide rapid, qualitative, Provision of sanitation infrastructure so people do not depend on mangroves http://www.ecologyandsociety.org/vol21/iss2/art11/and participatory diagnoses of ES elasticity and the resulting opportunities.An outstanding challenge remains in linking ecosystem quality and well-being at different scales.Although dependency and ES elasticity may be reduced at one scale, alternatives are frequently based on ES from further afield so that well-being is sensitive to aggregate Ecological Stocks and multipliers at larger scales beyond the scope of most place-based assessments.
Modeling, particularly recent innovation and development of "end-to-end" models (Fulton 2010), may provide an opportunity to synthesize diverse types of information (from biophysical to social), to explore how patterns of elasticity emerge and can be modified.Modeling approaches range from simple empirical relationships between Ecosystem Stocks, Flows, and Goods and some measure of well-being, to simulation models that have explicit representation of socioeconomic and ecological processes and their temporal and spatial dynamics.Existing modeling approaches mostly focus on the biophysical dynamics, with only simple representation of Goods, e.g., catches, and Values, e.g., net economic profits, and occasionally Share for different groups or sectors, e.g., along the value chain (Christensen et al. 2011).The subjective and context-dependent nature of well-being suggests that participatory modeling may provide a means to better represent the right-hand side of the chain from Value to Wellbeing Contribution.For example, to explore the trade-offs between different stakeholders under scenarios of fisheries management in coastal Mombasa, outputs from ecological models were linked to a rule-based model describing the interactions between different groups of fishers, fish traders, and their well-being (Daw et al. 2015).Other modeling approaches such as agent-based modeling (Murray-Rust et al. 2013) also offer potential for disaggregating individual Shares and Well-being Contribution.

CONCLUSION
Given the complex and as yet poorly understood linkages between ecosystem services and human well-being, ES elasticity is proposed as a core concept that informs these critical relationships.These relationships are expected to be nonlinear, complex, and context dependent.
Our heuristic framework of an ecosystem service chain is designed to help analysis of elasticities of different ecosystem services.This framework promotes and facilitates a structured assessment of ecosystem services-well-being relationships as a common starting point for interdisciplinary analysis and should provide important insights into trade-offs and possible policy levers.It can facilitate a pro-poor and well-being-explicit approach to understanding different mechanisms along the chain that determine who is able to derive what benefits from changes in ecosystems.This adds to existing ecosystem services frameworks that may overlook critical issues of access, needs, and aspirations or ecological dynamics.Applying this framework across different ecosystem services, contexts, and for different people to explore patterns of ES elasticity could advance both ecosystem services theory and provide practical guidelines for pursuing environmental management for human well-being.__________ [1] We acknowledge that "ecosystem quality" implies a value-laden assumption of quality that depends on what aspects of an ecosystem are desirable (Robards et al. 2011).We do not discuss this in this paper and assume that some dimension of ecosystem quality is applicable in practise, as indicated by the frequent use of proxies such as indicators of forest cover, biodiversity, or fish biomass. [2]In practise these substitutes are often also ecosystem services, e.g., imported food, from elsewhere but in many cases this represents a subsidy from ecosystems outside the scope of the system under study.
Responses to this article can be read online at: http://www.ecologyandsociety.org/issues/responses.php/8173

Fig. 1 .
Fig. 1.The ecosystem quality elasticity of human well-being (ES elasticity).(a) Examples of high, low, positive, and negative elasticities.(b) A threshold (shaded) in the ecosystem qualitywell-being relationship leading to different elasticities at different levels of ecosystem quality.(c) Hysteresis in the ecosystem quality-well-being relationship such that elasticity is different when ecosystem quality is declining from when ecosystem quality is increasing.
For example, elasticity of Flow (b) to Ecosystem Stocks (a) can be represented as the change in Flows (b) divided by change in Ecosystem Stocks (a).

Fig. 2 .
Fig. 2.A framework to interrogate ecosystem-well-being relationships and shed light on factors affecting ecosystem service elasticity.Ecosystem services are represented as a chain of elements (a-f) that link ecosystem stocks to well-being (See Table1for definitions, Table2for examples, and text for further elaboration).
E b,a = Δb / Δa and the product of the elasticities between elements ultimately determines how people's well-being is influenced by changes to Ecosystem Stocks, i.e., overall ES elasticity: ES elasticity (E f,a ) = E b,a x E c,b x E d,c x E e,d x E f,e = Δf / ΔaThis explicitly acknowledges that ES elasticity can be affected by any of the biophysical or social processes along this chain.
time.For example, extraction of Goods impacts Ecosystem Stocks, while Valorization incentivizes and thus drives Human Inputs.Ecosystem Stocks (a) represent natural capital or ecological assets(Bateman et al. 2011), affected by a wide range of impacts that may be external to this particular chain, such as climate changedriven disturbance or infrastructure development, or that represent feedback from within the chain, such as extraction of Goods.Stocks could be represented by a single indicator of ecosystem quality, e.g., live coral cover, but it may be more appropriate to include a range of processes and interactions within Stocks.For example, supporting services that maintain structure and functionality of ecosystems (Mace et al. 2012) would be included within Ecosystem Stocks because they do not directly contribute to well-being.Flows (b), the biophysical processes that are potentially directly useful for humans are equivalent to Mace et al.'s (2012) "final services."They are generated by, but are importantly distinct from Ecological Stocks (Vira and Adams 2009).The complex and multifaceted relationship between the two (Mace et al. 2012) is represented by the multiplier Ecological Dynamics (i), which encompasses variable production functions or biodiversityecosystem service relationships(Mora et al. 2014).

Fig. 3 .
Fig. 3. Images representing the five ecosystem services in Table 2. (A) The wave attenuation provided by the coral reef offshore at Bamburi beach, Mombasa, provides calm waters and beachfront for recreation in the evenings.(B) Female fish traders buying fish from fishers in Kenya.(C) Colorful corals and associated fishes of high value in international aquaria trade.(D) People resting in the shade of a tree in Vamizi, Mozambique.(E) Fishermen bringing large reef fish ashore.

Fig. 4 .
Fig.4.The branching of an ecosystem services-well-being chain by consideration of multiple flows from a given stock, multiple values of a good, and multiple groups of people receiving different shares of that value that contributes to each's well-being dependent on their needs, gaps, and aspirations.The aquarium fish chain is only illustrated as far as the flow.Arrows between each element highlight high or low elasticity between elements along the chain.
Table 1 for definitions, Table 2 for examples, and text for further elaboration).

Table 1 .
Explanations and examples for the elements and multipliers.For clarity, references to elements and multipliers are formatted as bold and italics, respectively.See Table2for examples.
Human factors that combine with Flows to coproduce Goods c. Good Things and services directly experienced or used by, and valued by people iii.Valorization Processes that determine the societal value of Goods d.ValueAggregate worth of benefit from goods produced, regardless of distribution iv.Access

Table 2 .
Five coastal ecosystem services from East Africa (see Box 1) mapped out using the chain framework, with examples of low and high ES elasticity deriving from different parts of each chain.Multipliers contributing to low or high overall ES elasticity are marked with †.

Table 3 .
Implications of high and low ES elasticities for policy and environmental management.CBNRM, community-based natural resource management.

Table 4 .
Interventions to improve well-being suggested by the ecosystem-well-being chain conceptual framework.