Human activities affecting the environment are intensifying. In marine ecosystems, direct stressors of overfishing and habitat destruction have led to biodiversity declines, prompting calls for conservation action and more sustainable resource management. Managing stressors becomes more challenging when stressors interact, or are indirect, whereby the source of the threatening activity is displaced from impacts, such as the localised effects of global warming, or runoff from land-use change degrading coral reefs. Effective management requires improved understanding of ecosystem responses to multiple stressors and consideration of interactions between stressors and species.Opposing objectives and outcomes of marine conservation (i.e. reducing stressors) versus resource management (often increasing stressors) compromise effective decision-making. Approaches are needed that include ecological and socio-economic information to promote long-term sustainability of extractive uses, protect functioning ecosystems, and conserve biodiversity. The choice of tool used to inform ecosystem management is typically constrained by management goals and available resources -spatial planning is typically used for conservation, whereas resource management relies more on population modelling.The goal of this thesis is to harness spatial and population modelling techniques to combine resource management with biodiversity conservation and link direct and indirect stressors to marine biodiversity persistence. Chapter 1 provides the context for the thesis, where I explore these themes and outline differences in decision-making for conservation versus extraction. In Chapter 2, I explore how agencies have managed stressors to date, and show that a decision-theoretic approach known as "structured decision-making" (SDM) is a more logical and effective way of managing stressors than traditional approaches that map threats. SDM requires clear objectives, canvassing of alternative management options, and linking outcomes for biodiversity to the effectiveness of each management option. SDM ensures that conservation actions occur where they are the most cost-efficient or effective. I apply recommendations from Chapter 2 in two case studies for managing multiple stressors to marine ecosystems: 1) spatial planning for oil palm agriculture and coral reef fisheries in Papua New Guinea (Chapters 3 and 4), and 2) managing whale species affected by harvesting, climate change and krill population dynamics in the Southern Hemisphere (Chapters 5 and 6).In Chapter 3 I combine ecosystem models and threat maps with spatial conservation planning to predict responses of coral and seagrass ecosystems to changing land uses. My novel framework identifies high priority areas for marine reserves whilst accounting for indirect impacts of land use change such as oil palm development and uncertainty in future stressor intensities. I demonstrate ii how we can reduce adverse impacts of oil palm expansion and make more effective whole-ofsystem decisions for coasta...