The worldÆs marine ecosystems are changing due to the impacts of climate change, harvest of natural resources, nutrient pollution, and contamination. In recognition of these threats, The Joint Subcommittee on Ocean Science and Technology of the National Science and Technology Council of the White House Office of Science and Technology Policy (JSOST) recommended a new course for the nationÆs ocean policy. Within this framework, JSOST identified a near-term need of undertaking comparative analyses of the structure and function of marine ecosystems to improve indices of ecosystem health and the stewardship of natural resources. In support of this goal, our research team conducted a series of mathematical and statistical modeling studies to understand the sources, patterns and, consequences of connectivity among nine marine fishery ecosystems that together from the northwest Atlantic coastal shelf ecosystem from Cape Hatteras, NC to the Canadian border. Together these ecosystems support production of over 6.5 million metric tons of finfish and shellfish, provide $1.3 billion of economic benefit to local communities and provide recreational opportunities and direct experience of marine ecosystems for the 52.6 million people who lived in coastal states from Maine to Virginia as of 2003. Our analyses were conducted in a hierarchical fashion with smaller ecosystems nested spatially within larger ecosystems. By taking this approach, we worked to identify the key ecological processes that describe the growth and survival of valuable fish species within these ecosystems, and the relative importance of those processes when compared to human activities such as harvesting. In particular, my role in the larger project focused on three specific areas: 1) a data driven multispecies fish modeling project designed to quantify interactions (predation and competition for food) among 15 fishes in multiple areas - Gulf of Maine, Georges Bank, and southern New England; 2) a simulation modeling project based on fishes linked by a predator-prey relationship in the mid-Atlantic and structured to characterize the effects of spatial areas with low predation impacts (refugia) on the sustainability of harvesting; and 3) a simulation modeling project of a migrating predatory fish species frequently encountered in the Chesapeake Bay and nearby coastal areas that was designed to compare the relative roles of pulses of food production (higher prey availability) against harvesting on the rebuilding timeline from a depleted state. The first study, which is in the final stages of preparation for publication, showed convincingly that the strength of interactions among fish species varies considerably across spatial areas. This result brings to the forefront the idea that ecological processes involving fishes are highly variable across spatial areas, and that the boundaries of those areas generally do not match the jurisdictional borders of our current management system (e.g., state lines). The second study, which is also nearly ready for publication, demonstrated that spatial areas largely free of predation were vital to the future sustainability of prey species that experience both high levels of predation and harvest. Although not originally intended to do so, this result importantly highlights the need for thoughtful spatial management of marine ecosystems. And finally the third study, which is currently being reviewed for publication, showed that while higher food availability was beneficial to the predator, recovery time from a depleted state was far more related to harvest control than ecosystem processes. Agencies, at all levels, are seeking to develop ecosystem-approaches to management (EAM) of fisheries to ensure long-term sustainability of the exploited marine resources and ecosystems. Central to policy development are Integrated Ecosystem Assessments (IEAs) because they provide the societal, legal, and scientific basis for management. Ou...