Data from: Anthropogenic disturbance modifies spatial patterns of reef sharks

This repository contains data and R scripts used to model elasmobranch relative abundance (count data per BRUV deployment) throughout the Dutch Caribbean. The workflow focuses on testing how habitat and depth relate to observed counts for focal taxa, while accounting for broad spatial structure using a categorical “Location” term (treated as a random effect). Models are fitted with Bayesian inference using the INLA framework, followed by diagnostic checks (overdispersion, residual plots) and exploratory assessment of spatial autocorrelation in residuals (mapping + variograms). Posterior predictive simulations are used to evaluate dispersion and whether the fitted Poisson models can reproduce the observed proportion of zeros.