Abstract
Evidence-based decision making as part of marine spatial planning (MSP) requires an understanding of marine ecosystem structure and functioning, including associated spatial-temporal variability. Community spatial structure, such as that derived from larval transport dynamics, should be considered in decision making because it can govern system stability and resilience to perturbation. We utilize a novel network-theoretic approach for evaluating connectivity and identifying statistically persistent connectivity cores in community structure for use as a decision support tool for MSP. Using the Atlantic Sea Scallop as a case study, we simulated scallop larval dispersal for spring and fall spawning periods from 2018 to 2022 using the Finite‐Volume Coastal Ocean Model (FVCOM) particle‐tracking module with explicit temperature dependent development and vertical‐swimming behavior parameterizations. We adapted a novel Recursive Significance Clustering (RSC) scheme to characterize and contrast statistically significant connectivity core structure from dispersal networks at both seasonal and 5-year timescales. Statistically persistent cores emerged within the Bay of Fundy, on Browns Bank, Georges Bank, and Sable Bank at both time scales. Regions of high spatial and temporal variability in community structure included the Great South Channel, Western Gulf of Maine, and in regions bordering communities. We also characterize potential source-sink larval dynamics using degree-based connectivity metrics, which revealed the Bay of Fundy to Browns Bank and the Great South Channel as potential sources and the Northern Gulf of Maine as a potential sink. Our results characterize variable scallop community structure and provide valuable information for MSP decisions surrounding fisheries management and offshore energy development. This framework can be applied to other species and regions for local MSP objectives.