Challenges and Sensitivities in Assessing Total Ecosystem Service Values: Lessons From Vanuatu for the Pacific

TESV for Vanuatu

The value of Vanuatu’s ESs to human society is considerable, delivering a median value of more than $2,901 m per year (min.: $1,963 m; max.: $11,537 m; Table 3). On a per capita basis, this is $10,423 per annum (min.: $7,054; max.: $41,452). By comparison, Vanuatu’s GDP for 2016 was $804 million (m; World Bank, 2018b)—smaller by a factor of 3.6—and Vanuatu’s GDP per capita is $2,890 per annum. This reflects the relative importance of ESs to the people of Vanuatu, when compared with goods and services traded through formal markets. Food provision, predominantly originating from the system of subsistence agriculture and agroforestry, is the most important ES by a considerable factor, representing 67.8% (median: $1,957 m yr^–1) of final ES value to Vanuatu. Air quality regulation (15%; median: $433 m yr^–1), water provision (7.3%; median: $210 m yr^–1), and climate regulation (5.5%; median: $158 m yr^–1) comprise the majority of the remainder. Provisioning ES provide $2,255 m yr^–1 (78.1% of total), regulating services provide $591 m yr^–1 (20.5%), and cultural services (mainly tourism) provide $40.7 m yr^–1 (1.4%). Of Vanuatu’s TESV, 87.6% derives from just two ecosystem types: subsistence gardens and tropical forests (Table 3).

TESV for Tanna

Similarly, the value of final ES to the people of Tanna is considerable, providing $325 m yr^–1 (min.: $368 m; max.: $681 m; Table 4). On a per capita basis, this is $10,053 per annum (min.: $6,973; max.: $21,100). Comparisons with Tanna’s gross island product are not possible, due to lack of data. For Tanna, $304 m yr^–1 (93.6%) of final ES value is delivered through provisioning services, $19 m yr^–1 (6.0%) is delivered through regulating services, and $1.4 m yr^–1 (0.4%) is delivered through cultural services. Approximately 99.0% of Tanna’s ES value derives from just two ecosystem asset types: subsistence gardens and tropical forests.

Sensitivity and Data Gaps

In data-poor environments, analysis risks significant sensitivity in valuation. Analysts must trade-off between “specificity, reliability, and applicability” to provide justifiable and replicable valuations (Troy & Wilson, 2006, p. 445). Our results show reasonable spread in the median, minimum, and maximum values. For example, on a dollar per hectare per year basis, for Vanuatu, coastal coral exhibits the greatest spread, with minimum and maximum values varying from the median (1.0) by 0.2 and 47.2, respectively (Table 3). The largest spread in value for individual ESs was for ornamental services (1.0, 0.1, 105.1). In general—and unsurprisingly—the greater the number of observations abstracted through our economic valuation strategy, the greater the spread in ES values. However, this relationship was a poor fit (R² of .0016). For example, the variance for food, which had the greatest number of observations, was relatively modest (1.0, 0.7, 1.4), though this is strongly influenced by the value attributed to food for subsistence gardens (Table 3). Overall, the variance across all ESs was between 0.7 (min.) and 4.1 (max.).

Variation in valuation estimates is undoubtedly a reflection of the heterogeneity in the socioeconomic-ecological environments of the original studies and the valuation methods employed rather than a systemic weakness in benefit transfer. However, the reasonable variances from the limited number of studies indicate that incorporation of data sets from new studies could significantly change our coefficients. The work completed for the TEEB valuation database by van der Ploeg and de Groot (2010), though highly cited, is now more than 10 years old and based on studies going back much further. Encouraging funding for more primary studies is the obvious answer; however, continually supported and curated valuation databases of existing studies would also improve this situation. More broadly, our study identified there is a lack of ES valuation data from research conducted in Vanuatu and Pacific SIDS, more generally. For example, our search revealed just two valuations for Vanuatu and relied on a single source for subsistence gardens in a Melanesian context (see Valuation Methodology section).

Our method also highlighted gaps in key data. For example, our case studies lack an important estimate for biological control for tropical forests and climate regulation for subsistence gardens. Such gaps result in an underestimation of the ES values for these habitats. These gaps also extend to spatial data. Our method’s data source for seagrass estimated zero extent for Tanna, yet from fieldwork conducted by the authors in 2018, we discovered and measured seagrass beds on Tanna’s south coast.

OPEN IN VIEWER

Table 5.

Variance in Final Ecosystem Service Value for Vanuatu (thousand 2016 US$ha^–1 yr^–1).

Ecosystem service type	Ecosystem service	Mdn	M	Minimum	Maximum	Mdn	M	Minimum	Maximum
Provisioning	Food	1,956,787	1,970,990	1,377,965	2,675,476	1	1.0	0.7	1.4
	Water supply (drinking & irrigation)	210,457	249,244	132,882	404,645	1	1.2	0.6	1.9
	Raw materials and energy	32,467	301,752	400	2,519,180	1	9.3	0.0	77.6
	Genetic resources	5,415	28,859	4,770	76,393	1	5.3	0.9	14.1
	Medicinal resources	49,808	481,905	274	2,096,438	1	9.7	0.0	42.1
	Ornamental resources	224	7,921	16	23,522	1	35.4	0.1	105.1
Regulating	Air quality regulation	433,079	433,079	433,079	433,079	1	1.0	1.0	1.0
	Climate regulation	158,254	334,689	12,022	1,091,625	1	2.1	0.1	6.9
	Moderation of disturbance	59,083	293,205	16,568	1,460,170	1	5.0	0.3	24.7
	Water flow regulation	1,184	1,184	1,184	1,184,292	1	1.0	1.0	1.0
	Waste treatment	221	1,038	207	2,687,396	1	4.7	0.9	12.1
	Erosion prevention	102,747	371,182	2,534	1,382,190	1	3.6	0.0	13.5
	Soil fertility maintenance	24,600	174,631	11,260	488,034	1	7.1	0.5	19.8
	Pollination	39,815	39,815	35,581	44,049	1	1.0	0.9	1.1
	Biological control	18	18	18	18	1	1.0	1.0	1.0
Cultural	Aesthetic	232	232	232	232	1	1.0	1.0	1.0
	Cognitive	126	1,136	3	3,278	1	9.0	0.0	25.9
	Inspiration	5	5	5	10	1	1.0	1.0	2.0
	Spiritual	71	71	71	71	1	1.0	1.0	1.0
	Recreation	40,253	161,153	1,207	827,705	1	4.0	0.0	20.6

Ecosystem Integrity

It is not contentious to assert that the quality and thus quantity of ES flows reduce with declining ecosystem integrity. For example, 10 fragmented 10 ha blocks of tropical forest when compared with a single, intact 100 ha block will have a lesser potential for the provision of a number of ESs, including as a carbon stock, refuge for wildlife, and regulation of sediment and microclimates (Briant et al., 2010; Broadbent et al., 2008; Forman & Alexander, 1998; Hamilton et al., 2017; Kindu et al., 2018). Benefit transfer valuations commonly lack the local data needed to match the condition of ecosystems with either the potential or realized flow of benefits. We know from field transects carried out in 2017 and 2018 that, while Tanna’s coral reefs are generally in adequate condition (localized pressure in proximity to the main town, notwithstanding), there is wide variation in the condition of Tanna’s tropical forests, which range from young secondary regrowth to intact primary forest. The converse can be said for the condition of coral reefs across Vanuatu, which is not consistent with the relatively good condition of those around Tanna (Naviti & Aston, 2002). While it is probable that the values on which our ES coefficients are derived are from habitats that span a range of conditions, site-specific survey data are needed for these relationships to be quantified and appropriate adjustments made to the coefficient values of relevant ESs. This is an ongoing subject of further research, recognized in the SEEA-EEA framework (UN Statistical Division, 2018). SEEA-EEA tackles this challenge from two directions: first, by including only those ESs for which there is a measurable local demand, thus ensuring that the demonstrated demand can never exceed supply, and second, by reporting indicators of ecosystem asset condition that can be used to scale the ability of specific assets to deliver particular ES flows (Keith et al., 2019). In our case study, reflecting only services for which there is a local demand would ensure more robust final ESV for subsistence gardens and plantation forests but would come at very significant cost in terms of demanding detailed primary research.

Application of Data

Our inclusion of subsistence gardens and plantation cropping as “ecosystem types” (consistent with the SEEA-EEA framework) lends to results being interpreted in a way that suggest land-uses can be traded off to increase the TESV generated by a landscape. In our case studies, the trade-off between tropical forests, plantation forests, and subsistence gardens presents a pertinent tension. Across much of Vanuatu (mountainous areas, aside), each of these ecosystem types are interchangeable and are in competition. A forest, for example, provides different and often incompatible services to distinctive user groups who might wish to utilize it in competing ways—using a forest for timber extraction impacts on others who harvest nontimber forest products, yet both groups may have legitimate access to the habitat. The potential conflicts between incompatible land uses highlight the need to understand the role of synergies among and between regulating ESs that arise from the connectivity between ecosystems, such as the role of tropical forests in protecting coral reefs by reducing sediment runoff (Brown et al., 2017) and the role of secondary tropical forest regrowth in providing soil fertility. Ostensibly, at a per hectare per year level, the contribution of subsistence gardens to food is US$12,552, whereas tropical forest contributes US$917 in final services. This suggests conversion of tropical forest to subsistence gardens represents optimal policy for increasing well-being. It also, perhaps, explains the reality on the ground: On Tanna, the conversion of tropical forest to subsistence gardens is occurring at a significant rate (∼60 ha per year; Mackey, Ware, Nalau, et al., 2017). Simplistically, the greater the rate of conversion of tropical forest to subsistence gardens, the greater the level of welfare that can be obtained from the land.

Our data assume a linear relationship between ecosystem asset area and ES delivery. Despite an overall increase in the area of cultivated land, population growth (2.3% per annum) has reduced the cultivated area per capita (however, it is possible urbanization might have offset the per capita loss of cultivated area for rural subsistence households). This trend suggests that increasing demand for food is resulting in the overexploitation of current agricultural land, leading to encroachment on forests in the search for new, more fertile soil for subsistence agriculture (Reardon et al., 1996). While tropical forests provide important ESs, subsistence gardens provide the most pressing of human needs: food. Reported reductions in agricultural yields, most likely from root crop diseases and falling soil fertility from insufficient fallow periods, coupled with population growth that increases the demand for food, will likely lead to further deforestation pressures until forest scarcity equalizes the opportunity cost of more clearing. This trend has implications for the ongoing protection of tropical forests on Tanna and their associated ESs, including the regulating services they in turn provide to subsistence gardens, as apparent, observed ecological state and thus, the signal of scarcity, might not be in equilibrium.

On Tanna, external pressures, such as climate change impacts and the increased demand for food from population growth, will undoubtedly heighten pressures on habitats, increase the likelihood of degradation in their condition, and potentially invoke irreversible changes. Experience elsewhere across a range of ecosystems highlights that the risk of reductions in future yields of ESs (particularly provisioning and regulating) may not become evident before thresholds have already been breached and state change has become inevitable (Koch et al., 2009). Current ES flow is thus not necessarily the final equilibrium outcome of present drivers and current state. For example, it is unclear if current fishing effort on Tanna is within long-term sustainable thresholds, or the “safe operating space” for coastal coral reefs (Hughes et al., 2017). Community consultations on Tanna reveal the shore-based reef fishery is both important to local food supply and has been subject to significant changes in catch, in terms of quantity and fish species. Time-series data collection should be a priority to detect any changes in ES delivery. However, we currently lack the information to identify thresholds in ecosystem asset areas below which specific ESs begin to degrade nonlinearly.

Incorporating Kastom Conceptions of ESs

Governing services are those “co-produced by humans using societal institutions and nature’s components and processes” (Spangenberg et al., 2015, p. 202). In this instance, the authors were discussing developed world governance structures—laws, policies, governments that protect private property—but similarly, Mackey et al. (2017) introduced the concept of “Kastom services” to articulate the close coupled socioecological systems, such as how Kastom land and reef resource management is an enabler for people and communities to harvest and benefit from ESs. Estimates of these contributions to aspects of Kastom in the Pacific SIDS context have been attempted (Kenter et al., 2011; Paudyal et al., 2016). Although Ni-Vanuatu communities undoubtedly experience significant benefits from habitats and landscapes providing ESs to culture, including sense of communal identity, sacred sites for ceremonies, and other aspects of traditional knowledge (Nalau, Becken, Schliephack, et al., 2018), there are very few studies from which benefit estimates could be transferred. Furthermore, benefit transfer is likely not an appropriate method for Kastom services as, by their nature, they are highly situation-specific (Kenter et al., 2011). We, therefore, did not attempt to derive a monetized valuation of one of the most important benefits of tropical forest ecosystems in the Vanuatu context, namely, their Kastom use. Kastom use is largely of a nonextractive nature, including cultural practices and harvesting of some non-timber-based products. The cloud forest on Tanna, in particular, has both high biodiversity and Kastom values, as have numerous places across Vanuatu and Pacific nations. While it is not essential to obtain monetized valuations for cultural ES, it is worthwhile investigating whether there are feasible approaches to qualifying Kastom values in a way that can integrate with ESV through multi-criteria assessment, particularly as, in our case studies, these forests are under pressure from poorly regulated timber extraction and expansion of subsistence agriculture.

While the demand for food, raw materials, and water is direct, immediate, and well understood in the community, regulating and cultural services, which undoubtedly contribute essential elements of value, are more difficult to evaluate (Fischer & Eastwood, 2016; Rasmussen et al., 2017). This has implications for data capture methods in future research into the ESV in SIDS and least developed countries and perhaps suggests that discursive elicitation techniques should be used in parallel to traditional valuation methods to better capture social and shared benefits provided by the different ES types (Jacobs et al., 2016; Kenter et al., 2011, 2016). For example, the value of the Tanna reef fishery is not simply a function of the catch rate but also an outcome of an understanding of the role of cultural relations within and between tribes and their relationship with Kastom natural resource management (Foale et al., 2011; Lindstrom, 2011; Stoeckl et al., 2018).

However, the very concept of “cultural services” is critiqued. For example, Satz et al. (2013) argue cultural benefits are more conceptual rather than material, which has tended to lead the category being used as a miscellaneous basket into which unallocated, nonmaterial services get placed when no better basket can be determined. Instead, the authors suggest cultural ESs be revised and the term intangible/immaterial ESs be used to clarify boundaries. Further critiques have looked at the relational aspects of cultural services, arguing cultural services are coproduced by nature and society (Díaz et al., 2018) or the value lays in the state of the “stock” of relational strength between people and nature (Jackson & Palmer, 2015).

Willingness to Pay as a Measure of Value

Economic valuations of ES in low-income contexts, such as those from low-income countries in the TEEB database, will typically yield very low willingness to pay values. In predominantly cashless, subsistence communities, where land is majority customarily owned, there is often lower values associated with the utility of money (as there is often little to buy with cash) when compared with communities for whom exchange via cash is dominant (Bush et al., 2020; Kenter et al., 2011). Kastom communities in the Pacific also have a tendency to accumulate and share common wealth (Lindstrom, 1982; Nalau, Becken, Schliephack, et al., 2018) rather than retain it only at the individual or household level—the standard unit of neoclassical economic analysis. From initial fieldwork on Tanna, it is evident common ownership principles do not extend to the full gamut of material goods (fishing gear, boats, and household items are exclusively “owned” by households or families), but clearly land and coral reefs and the provisioning of goods from these habitats are customarily owned and therefore do not lend to individualistic notions of willingness to pay estimation of value.

Monetary values of ESs elicited through economic valuation methods assume that demand curves are derived from opportunity costs in efficiently functioning markets, not one where the price is opaque or the market is distorted, which is often the prima facie case for undertaking the economic study in the first place. For example, basing the marginal ES value of a subsistence garden on the cost of goods purchased at a nearby market (in this instance, the value from Anderson, 2006) assumes the prices at those markets have already accounted for any externalities (positive and negative) in their production, such as forest loss and depletion of localized regulating ESs to grow cash crops. This is never likely to be the case. Vinde Folkersen (2018) suggests a range of value-indicators, such as time, labor, geographical distance, and collective community efforts be explored in the evolution of social and nonmonetary valuation of ecosystems in Pacific SIDS contexts, which reflect complex, social norms and cultural value systems, such as Kastom (Costanza et al., 2017; Kenter et al., 2016; Stoeckl et al., 2018; Vinde Folkersen, 2018). Further research that reveals more locally explicit exchange values for the contribution of subsistence agriculture (e.g., based on replacement cost) would fill an important research gap at the national or regional scale; however, overlaying these values with complex, social values at the local scale should also be included where budgets allow.

Last, nonmarket valuation methods assume that social welfare (or the total value of a good, or the net social benefit of a project) is the sum of individual willingness to pay values. There is growing recognition that benefits associated with the environment are frequently interdependent (i.e., “complex”; Stoeckl et al., 2018). Researchers who aggregate partial values to estimate TESV risk both double counting overlapping values and precluding interdependent values that may be synergistic; that is, the whole may comprise more than the sum of its parts. Furthermore, simple aggregation of individual estimates of “value” may be inappropriate in some contexts, particularly when both the ecosystem asset and their values are held at the community level (Adamowicz et al., 1998). This suggests that “integrated valuation” or “participatory valuation” methods could be both fruitful and useful (Costanza et al., 2017, p. 8).

The Political Realm

Finally, ESV as a decision support tool should be placed in a political context. Vanuatu has international commitments to the Convention on Biological Diversity and the Aichi Targets, which binds its government to achieving biodiversity protection outcomes. These targets provide impetus to policy from the top down to the community level to support sustainable management of ecosystems and Kastom resources through its own National Biodiversity Strategy and Action Plan (Republic of Vanuatu, 2018), which includes nationally appropriate interpretations of the Aichi Targets by referencing sustainable management of natural resources, protection of biodiversity for future generations, and protection of the legal rights of Ni-Vanuatu as resource custodians and users. ES value estimates should be used as merely one type of evidence among many (e.g., habitat assessments, ecological field studies, biological inventories, socioeconomic research, etc.) in supporting natural resource allocation and management decisions (Troy & Wilson, 2006), particularly to avoid risks of commodifying nature (Gómez-Baggethun & Ruiz-Pérez, 2011).