Agent-based modeling of the effects of conservation policies on social-ecological feedbacks between cropland abandonment and labor migration.

Context: Payments for Ecosystem Services (PES) has increasingly been used in ecological conservation efforts through implementing forest policies worldwide. However, outcomes of forest policies with PES are often mixed due to the complex social-ecological dynamics. Objectives: In this study, we develop a spatially explicit agent-based model for cropland abandonment and labor migration (ABM-CALM) based on data from very high-resolution satellite images, the global positioning system, household surveys, and a population census. Methods: The ABM-CALM is used to simulate bidirectional social-ecological feedbacks within a coupled human and natural system under two large PES programs in China. Results: The simulation results show that labor migration and cropland abandonment follow nonlinear trajectories with feedback loops mediated by household capital endowment. Households tend to reduce labor allocation for migration to retain essential croplands after abandonment of marginal ones. For the reforestation program, participating households are more resilient in maintaining migration than non-participants. The model reveals the feedback mechanisms between the two concurrent PES programs through synergistic and offsetting interactions. Specifically, labor migration under higher payments of one program is more sensitive to payments of the other due to the feedback of additional cropland abandonment induced by the former program. Conclusions: Using the real-world landscape and in-situ socio-economic settings, the integrated model captures the dynamics of the social-ecological systems affected by PES across space and time. [ABSTRACT FROM AUTHOR]