Coral connectivity: How biophysical modelling choices shape connectivity inference

Chinenye Jane Ani , Camille Grimaldi, Takuya Iwanaga, Eric Treml, Barbara Robson 

Understanding coral connectivity is essential for predicting metapopulation dynamics and guiding restoration under accelerating climate change. Connectivity facilitates the exchange of organisms among reefs and can promote recovery following disturbance. Biophysical models are widely used to predict and explore coral connectivity; however, their estimates carry substantial uncertainty arising from limitations in hydrodynamic simulations, uncertainty in larval biological processes, and interactions between these processes and natural variability. We synthesise how physical, biological, and behavioural processes are represented in biophysical models, assess key sources of natural variability and uncertainty across the literature, and provide guidance for better connectivity modelling. We show that connectivity is not a single measurable property, but a model-dependent construct emerging from interactions among hydrodynamic forcing, larval traits, and numerical implementation. These processes introduce multiple, interacting sources of uncertainty that influence the magnitude, direction, and interpretation of connectivity estimates. We identify major gaps in current understanding, including limited empirical data for larval trait parameterisation, computational constraints on representing interannual variability in ocean circulation, limited availability of fine-scale hydrodynamic models, and insufficient observational and genetic data for scale-appropriate evaluation. We also highlight untested contributors to connectivity including mesophotic reefs and eco-evolutionary dynamics, that may substantially influence connectivity. Emerging approaches—including climate projections, socio-ecological frameworks, near-real-time modelling, remote sensing, and artificial intelligence—offer opportunities to improve the representation and prediction of connectivity. Advancing coral connectivity modelling therefore requires improved empirical biological data, careful treatment of uncertainty, scale-appropriate evaluation of model outputs against independent data, and transparent reporting of modelling assumptions.