<p>Environmental non-governmental organizations are now major players in environmental science and conservation. The largest now produce applied conservation science and work on local, national, and international scales and across scales to conserve marine and terrestrial ecosystems and connect local level environmental issues to international economic and political processes. However, despite the growing role of these organizations, there is still a lack of comprehensive examinations of their programs with a full analysis of programmatic design, structure, processes, and outcomes. </p><p>To fill this gap in both conservation practice and academic theory, I conducted a multi-scalar examination of Conservation International's Marine Management Area Science initiative. This $12.5 million initiative, lasting from 2005 until 2010 and funded by the Gordon and Betty Moore Foundation, had four main nodes of research and conservation work: Fiji, Belize, Eastern Tropical Pacific Seascape, and Brazil. Using non-governmental organization and science and technology studies literature as a theoretical framework, I endeavored to determine what factors affect how environmental non-governmental organizations manage the boundaries across multiple scales and between science and policy in international marine management area science initiatives. Drawing upon methodological approaches in multi-sited ethnography and participant action research, I conducted qualitative field research at each of the initiative's four main node sites and at Conservation International's headquarters, while simultaneously engaging with Conservation International so to return results back to the organization for adaptive management and learning.</p><p>My results are consistent with theoretical predictions and lend lessons learned to conservation practice. My research shows that managing the boundaries across scales and between science and policy in international marine management area science initiatives depends on how the program was initiated, the use of networks, partnerships, and coalitions, the level of programmatic participation, the degree of accountability and the ability to learn, the translation of scientific knowledge, and the assessment context. Future research on other environmental non-governmental organization programs has the potential to extend these findings.</p> / Dissertation