Understanding connectivity of marine species is crucial for their management. This connectivity, however, is difficult to quantify; propagules of marine species are typically small, numerous, and can travel large distances. Consequently, oceanographic models are often used to simulate larval dispersal. To avoid high computational costs, these models typically use, pooled or cross-sectional data and fixed biological parameters. Here we explore how temporal and spatial variation in current velocities, and pelagic larval duration and buoyancy can alter patterns of marine connectivity of Crown-of-Thorns starfish on Australia’s Great Barrier Reef. Our results reveal highly variable reef connectivity in space and time mediated by all three factors. No individual reef acted consistently as a population source or sink. It is, therefore, important to consider many factors concurrently when estimating connectivity for understanding these population dynamics, especially where such estimates are relied on for evidence-based decision making.

The structure of coral reef communities results from interacting evolutionary, ecological and environmental forces. How these factors interact in structuring these communities at a global scale, and how such effects might vary among biogeographical regions is unclear. We partitioned sources of reef community assemblage patterns by environmental, latent (i.e. unobserved), and random factors on 291 coral reefs distributed across five biogeographical regions. We then estimated how these factors were related to variations in abundance and co-occurrence among 16 functional groups. Latent factors better explained the distributions of opportunistic functional groups like algae, whereas environmental factors better explained abundance and co-occurrence of hard corals. Co-occurrence patterns revealed complex interactions between coral and algae groups that were not related to environmental factors but influenced by regional biogeography. Our results show that environmental factors are not the sole drivers of coral reef structure highlighting the importance of assemblage-level interactions and unobserved variables.