Considering the land-cover elasticity of ecosystem service value coefficients improves assessments of large land-use changes

Economic development often impacts on ecosystem services. Previous studies have raised public and political awareness of the costs associated with such impacts and the benefits of ecosystem services. In cases where empirical information on the value of ecosystem services is lacking, benefit transfer (BT) approaches that use value estimates from a previously studied site to estimate the economic values of a new target area have been established. One of the most popular BT approaches is unit value transfer, where constant ecosystem service value coefficients are used to assess a given land-use/land-cover (LULC) change. In several case studies assessing LULC changes, such unit value transfers with constant value coefficients are biased when nonmarginal changes are involved. Theoretical considerations suggest that large changes in land allocation should alter the opportunity costs of gaining or losing natural capital because the marginal costs of additional losses increase as some LULC types become scarcer (e.g. natural ecosystems). In contrast, marginal benefits shrink as other LULC types become more abundant (e.g. agricultural replacement systems).

Here, we propose an improved method for assessing larger scale (i.e., at national levels and beyond) LULC changes using endogenous value coefficients that account for the size of the land cover allocated to each LULC type and derive an equation for calculating these coefficients. The extent to which the value coefficient changes with variations in the land cover area depends on the land-cover elasticity of the value coefficient. Using a hypothetical numerical example of an area of tropical forest converted into grassland, we show that the bias caused by neglecting this land-cover elasticity can be considerable. We also demonstrate how the elasticity needed to correct the value coefficient can be estimated empirically. Finally, we suggest some modifications for future studies assessing large LULC changes.