Bayesian shared-parameter models for analysing sardine fishing in the Mediterranean Sea

TL;DR

This paper applies Bayesian shared-parameter models to analyze sardine fishing dynamics in the Mediterranean Sea from 1970 to 2014, revealing insights into overfishing patterns across countries.

Contribution

It introduces a Bayesian joint longitudinal modeling approach with shared random effects to study fishing dynamics, enhancing understanding of overfishing impacts.

Findings

Identification of fishing trends over time

Assessment of overfishing across countries

Model selection via Bayes factors

Abstract

European sardine is experiencing an overfishing around the world. The dynamics of the industrial and artisanal fishing in the Mediterranean Sea from 1970 to 2014 by country was assessed by means of Bayesian joint longitudinal modelling that uses the random effects to generate an association structure between both longitudinal measures. Model selection was based on Bayes factors approximated through the harmonic mean.