Development and comparison of tools to support the assessment of glass eels catch opportunities

• Main Author: Benezech, Mathilde
• Other Authors: AGROCAMPUS OUEST, INRAE, Unité EABX, 50 avenue de Verdun, 33612 Cestas, Étienne Rivot, Hilaire Drouineau
• Format: Master Thesis
• Language: English
• Published: HAL CCSD 2021
• Subjects: European eel; Overexploitation; Recruitment; Bayesian assessment model; Anguille européenne; Surexploitation; Recrutement; Modèle d’évaluation bayésien; [SDV]Life Sciences [q-bio]; Conseil
• Online Access: https://dumas.ccsd.cnrs.fr/dumas-03562603; https://dumas.ccsd.cnrs.fr/dumas-03562603/document; https://dumas.ccsd.cnrs.fr/dumas-03562603/file/2021_Benezech_Mathilde_SHA.pdf

Following the decline in glass eel recruitment since the 70s and the implementation of the Eel Regulation in 2007, France implemented a Management Plan in 2008 aiming to reduce glass eel fishing mortality by 60%. Thus, a Scientific Council developed recruitment prediction models to support the establishment of catch quotas. However, since 2015, the management target is no longer achieved, raising questions about models’ performance, and about assumptions on past and future recruitment trends. In this context, this study aims to address the limitations of the two current models by (1) proposing alternative models and (2) proposing new criteria to compare models’ performance. Thus, a model based on a spike and slab prior and one on a random slope were proposed to overcome too strong assumptions on recruitment trends, and they were fitted to the WGEEL abundance index from 1980 to 2019. To compare models, recruitment prediction error based on cross-validation as well as criteria quantifying conservation and fishery risks are studied. These indicators are measured in both retrospective and prospective exercises. The random slope model proves to be a good compromise, providing a limited recruitment prediction error and not favouring a single criterion among the exploitation and conservation criteria. In the future, it would be worthwhile exploring the impact of glass eel catches on escapement, which is the Regulation Management target. Suite au déclin du recrutement de civelles depuis les années 70 et à la mise en œuvre du règlement anguille en 2007, la France a établi un plan de gestion en 2008 visant à réduire de 60% la mortalité par pêche des civelles. Un Conseil scientifique a alors développé des modèles de prédiction du recrutement pour aider à l'établissement des quotas de capture. Cependant, l'objectif de gestion n'est plus atteint depuis 2015, soulevant des questions sur la performance des modèles, et leurs hypothèses sur les tendances passées et futures du recrutement. Ainsi, cette étude vise à remédier aux ...