Reproducing within-reef variability in coral dynamics with a metacommunity modelling framework

Reef systems span spatial scales from 10s to 100s and even 1000s of kilometres, with substantial spatial variability across these scales. Managing and predicting the future of coral reefs requires insights into reef functioning at all spatial scales. However, investigations of reef functioning often consider individual reefs as the smallest unit (10s of kilometres), despite substantial spatiotemporal variability occurring within-reefs (100s of meters). We developed C~scape, a coral metacommunity modelling framework that integrates the demography of corals with population-level responses to physical and environmental spatial layers, to simulate a mosaic of interacting coral communities across a heterogenous seascape. Coral communities are linked using biophysical connectivity modelling. Coral community growth is modelled with a logistic growth model, with the intrinsic growth parameter determined from taxa-specific Integral Projection Models to incorporate demographic mechanisms. Site-specific coral habitat parameters, derived from satellite-based geomorphic and benthic habitat maps, define the maximum coral cover and are used to modulate community growth spatially and temporally as a function of the available space suitable for corals. These parameters are a proxy for the many interacting physical and environmental factors - e.g., depth, light, wave exposure, temperature, and substrate type - that drive within-reef variability in coral demography. Using a case study from the Great Barrier Reef, we show that modulating community growth using site-specific habitat parameters enables more accurate hindcasts of coral cover dynamics, while overlooking within-reef variability may lead to misleading conclusions about metacommunity dynamics. More generally, C~scape provides a valuable framework for predicting spatiotemporal dynamics of coral communities within and between reefs, offering a mechanistic approach to test a range of management and restoration options. ### Competing Interest Statement The authors have declared no competing interest.