A size- and age-structured model to estimate fish recruitment, growth, mortality, and gear selectivity

作者：Catalano Matthew J^*; Allen Micheal S

来源：Fisheries Research, 2010, 105(1): 38-45.

DOI：10.1016/j.fishres.2010.03.002

• Stock assessment
• Gizzard shad
• Maximum likelihood
• Growth
• Population modeling

Most stock assessment models are solely age-based and ignore variability in length-at-age, but the incorporation of length structure can improve parameter estimates. Here, we present an age- and length-structured population model to estimate fish recruitment, von Bertalanffy growth parameters, natural mortality, and gear selectivity from survey length- and age-specific catch, fishery length distributions, and total harvested biomass. We evaluated parameter bias when fitting the model to simulated data with variable fish longevity, survey sampling duration (i.e., years of data), survey gear selectivity (i.e., asymptotic or dome-shaped), instantaneous fishing mortality rate and fishing mortality trend (i.e., stable or increasing). Median proportional bias in mortality (fishing and natural), growth, and gear selectivity parameters did not exceed 0.05 across all scenarios. Median recruitment bias ranged up to 0.12 and was greatest for the short-lived species and for scenarios with increasing fishing mortality, but bias for all other scenarios was less than 0.1. The model was demonstrated using data from a whole-lake biomanipulation of gizzard shad (Dorosoma cepedianum) at a 2300-ha lake in Florida, USA. Model estimates of gizzard shad exploitation rate were similar to an independent estimate from an in-season depletion analysis. Estimates of gizzard shad natural mortality and growth parameters were within the range of literature values for the species. The model could be useful for estimating growth, mortality, and recruitment parameters that account for gear selectivity and the cumulative effects of size-selective harvest.

• 出版日期2010-6