How can ecological systems be efficiently managed at different scales? Data to understand the variation of management practices and to quantify interactions and linkages among ecosystems are increasing in both their resolution and coverage at an unprecedented rate. But ecosystem management is only partly about understanding ecological dynamics. It is also about accounting for human interests, and communities vary in what they want from ecological systems and how they interact with those systems at different spatial scales. The use of fine-grained management strategies allows policy-makers to incorporate socio-ecological heterogeneity more precisely. But doing so risks organizations designing strategies in a piecemeal, site-by-site fashion that fails to account for spatial interactions and feedbacks between components of larger ecological and social systems. Moreover, implementing fine-grained management entails higher administrative costs than most management agencies are willing to bear. On the other hand, acting over broad spatial scales can obscure ecological heterogeneity and disconnects policy-makers from local resource users, which undermines management effectiveness. To balance this trade-off, conservation organizations often try to work over large spatial scales, while embedding institutional hierarchies in order to resolve ecological dynamics more finely and to better incorporate the actions and needs of local communities. In this interdisciplinary session, we will examine how best to apportion responsibility for managing ecological systems among natural resource management agencies and conservation organizations working over different scales. Speakers will present innovations in optimization methods that can be used to design management interventions that better reflect the institutional layering of the different actors involved. They will illustrate the underlying theories and methodologies with applications to marine fisheries, invasive species management, epidemiology, biodiversity conservation, and the management of recreation and hunting pressure. Applications will span developing- and developed-world contexts and include data from the nearby Appalachian region as well as tropical regions at the deforestation frontier. Some speakers will examine whether state or federal governments are better positioned to manage ecological systems; others will examine behavior of conservation NGOs. Some will downscale and examine how best to account for local resource users in regional management plans. The session will showcase how ecological models and data can be integrated using approaches fromenvironmental law, economics, mathematics and computer science and will include speakers from each of these disciplines. Key advances that will be presented in the session include combining game theory and metapopulation dynamics, bi-level optimization and biodiversity data, externality theory and epidemiological modeling.