A Hierarchical Bayesian Model to Monitor Pelagic Larvae in Response to Environmental Changes

Abstract

European anchovies and round Sardinella play a crucial role, both ecological and commercial in the Mediterranean Sea.In this paper, we investigate the distribution of their larval stages by analyzing a dataset collected over time (1998 - 2016) and spaced along the area of the Strait of Sicily. Environmental factors are also integrated. We employ a hierarchical spatio-temporal Bayesian model and approximate the spatial field by a Gaussian Markov Random Field (GMRF) to reduce the computation effort using the Stochastic Partial Differential Equation (SPDE) method. Furthermore, the Integrated Nested Laplace Approximation (INLA) approximates the posterior distributions of model parameters. The results suggest a strong relationship between sea currents' directions and the distribution of larval European anchovies. For round sardinella, the analysis indicates increased sensitivity to warmer ocean conditions.