The maintenance of less-disturbed vegetation in agricultural landscapes provides habitat for beneficial insects, promoting a class of services such as pollination and pest control that are collectively referred to as mobile agent-based ecosystem services (MABES). Patches of land on which MABES are enhanced can provide spatially far-reaching benefits, but a problem arises in fragmented smallholder farming landscapes as to whose land should become the host. The land available to a farmer, along with the set of available inputs, shape the opportunity cost of keeping land in productive use; further, the structure and security of institutions for land tenure – held privately, collectively, or by the state – shape the incentives farmers have to make investments in farm structure. The goal of our research is to improve MABES management and reduce the use of chemical pesticides through coordinated action across farmers at the landscape scale. Our key research questions are: 1) How do farmer objectives, and incentives to coordinate, vary across property size and institutional arrangements for tenure? 2) What are the most appropriate interventions to optimize MABES management across these different size and institutional contexts? We propose a three-year, three-country (Cambodia, Vietnam, China) study to design and evaluate incentives for MABES management, taking into account heterogeneity in property size and institutional arrangements for property rights that characterize China and Southeast Asia. Our choice of countries leverages existing research in Northern China, and covers landscapes i) where smallholder properties are unvaryingly small (> 2ha; Cambodia, China, and areas in North Vietnam) as well as where some consolidation has occurred (> 10ha; Southern Vietnam); and ii) with different prevalence of private property rights (Southern Vietnam and China), collective farming systems (Northern Vietnam), and threats to tenure security (Cambodia). Our research will combine modeling study with household survey. Drawing on existing MABES research in China and published models of MABES provision, a spatially explicit agent-based model of agricultural households will be developed to i) propose candidate incentives for managing MABES and reducing pesticide use, and ii) identify key variables in farmer decision making for the household survey. Survey research in chosen agricultural contexts will follow a 2-stage random sampling design of agricultural communities and households within them, post-stratifying by wealth, property size, land tenure system, and current reliance on pesticides. This survey research will inform the understanding of household level objective functions, decision-making criteria, and community network structure, and will feed back into a detailed agent-based model of farming decisions within the agricultural landscape. Coupled to existing ecological models of MABES provision, we will apply this modeling framework to evaluating and comparing alternative incentives for MABES management and draw implications on reducing pesticides use.