A hierarchical Bayesian model for embedding larval drift and habitat models in integrated life cycles for exploited fish
|Copyright||2013 by the Ecological Society of America|
|Author(s)||Rochette Sebastien1, 2, Le Pape Olivier2, Vigneau Joel3, Rivot Etienne2|
|Affiliation(s)||1 : IFREMER, Dept Dynam Environm Cotier, Lab Applicat Geomat, F-29280 Plouzane, France.
2 : Agrocampus Ouest, UMR Ecol & Sante Ecosyst 985, Lab Ecol Halieut, F-35042 Rennes, France.
3 : IFREMER, Lab Ressources Halieut, F-14520 Port En Bessin, France.
|Source||Ecological Applications (1051-0761) (Ecological Soc Amer), 2013-10 , Vol. 23 , N. 7 , P. 1659-1676|
|WOS© Times Cited||12|
|Keyword(s)||Eastern Channel (Western Europe), habitat suitability model, hierarchical Bayesian model, hydrodynamic models, integrated model, larval drift, life cycle, nursery, recruitment, sole, Solea solea, spatially structured|
|Abstract||This paper proposes a hierarchical Bayesian framework for modeling the life cycle of marine exploited fish with a spatial perspective. The application was developed for a nursery-dependent fish species, the common sole (Solea solea), on the Eastern Channel population (Western Europe). The approach combined processes of different natures and various sources of observations within an integrated framework for life-cycle modeling: (1) outputs of an individual-based model for larval drift and survival that provided yearly estimates of the dispersion and mortality of eggs and larvae, from spawning grounds to settlement in several coastal nurseries; (2) a habitat suitability model, based on juvenile trawl surveys coupled with a geographic information system, to estimate juvenile densities and surface areas of suitable juvenile habitat in each nursery sector; (3) a statistical catch-at-age model for the estimation of the numbers-at-age and the fishing mortality on subadults and adults. The approach provided estimates of hidden variables and parameters of key biological significance. A simulation approach provided insight to the robustness of the approach when only weak data are available. Estimates of spawning biomass, fishing mortality, and recruitment were close to the estimations derived from stock-assessment working groups. In addition, the model quantified mortality along the life cycle, and estimated site-specific density-dependent mortalities between settled larvae and age-0 juveniles in each nursery ground. This provided a better understanding of the productivity and the specific contribution of each nursery ground toward recruitment and population renewal. Perspectives include further development of the modeling framework on the common sole and applications to other fish species to disentangle the effects of multiple interacting stress factors (e.g., estuarine and coastal nursery habitat degradation, fishing pressure) on population renewal and to develop risk analysis in the context of marine spatial planning for sustainable management of fish resources.|