The Impact of Multiple Drivers on Marine Systems : Novel approaches for studying structural changes

Sammanfattning: Human action is transforming the species composition, biogeochemistry and habitats of the world’s oceans at unprecedented rates. The cumulative effect of natural and anthropogenic drivers is challenging to measure, in part due to indirect effects and the complexity of marine systems. Building on the theory of complex adaptive systems, this thesis aims to increase our understanding of how complex, heterogeneous marine social-ecological systems (SES) may respond to changing conditions. This thesis integrates resilience research with network science and describes change and structural patterns at several SES scales in order to advance our knowledge on the effects of multiple drivers.Paper I proposes a new, quantitative fish stock collapse definition, that accounts for fish stock dynamics and enables standardization and thus comparability across a large number of commercial fish stocks. Recognizing that substantial ecosystem changes are part of SES dynamics, in Paper II we review marine regime shifts worldwide to specify how co-occurring bundles of drivers are related to degraded ecosystem services for management purposes. A more detailed ecological study on regime shifts was performed in Papers III and IV. Paper III describes the late-1980s central Baltic Sea regime shift based on a food-web model. Paper IV uses a novel structural network analysis approach to detect functional shifts in complex food webs. The results of Paper IV imply that the Baltic Sea regime shift may not be a systemwide shift. Paper V uses a network approach to analyze fishing strategy diversification and social-ecological connectivity among Swedish Baltic Sea fishers, indicating that natural resource management evaluations should not be limited only to ecosystem conditions but also take account of social conditions.Overall, this thesis provides empirical evidence for the emerging perspective that marine resource science and management must account for the complexity of system elements in order to ensure the provision of ecosystem services in the future. The first application of Exponential Random Graph Modeling in ecology and an improved fish stock collapse definition provide new advanced tools for studying oceans from an SES perspective in the future.