Estimating walleye (Sander vitreus) density, gear catchability, and mortality using three fishery-independent data sets for Oneida Lake, New York

We used three long-term data sets (gill nets, trawls, and adult population estimates) for walleye (Sander vitreus) to simultaneously estimate density, gear catchabilities, and mortality using an age-structured, nonlinear model. Model con- straints included a fixed natural mortality rate and age- and gear-specific but time-invariant catchabilities. Trawl catchabil- ity decreased with age, whereas gillnet catchability increased towards a maximum by age-4. A sensitivity analysis was conducted to investigate how the information content of the different data sets influenced parameter estimates. Estimated catchability values were relatively robust to changes in data weighting. Estimated gillnet catchability values were fairly consistent with those derived from more conventional methods. An additional mortality term was added to reflect double- crested cormorant (Phalacrocorax auritus) predation, and vulnerabilities associated with angling and cormorants were cal- culated using independent length frequency information. Estimated subadult mortality increased and the influence of fish- ing mortality slightly decreased during recent years when double-crested cormorants were abundant and more restrictive size limits were in place. Walleye density in Oneida Lake (New York, USA) in the last decade is estimated to be approxi- mately half of that before 1990. Resume´ : Nous utilisons trois banques de donnees along terme (recoltes au filet maillant et au chalut, estimations de la population adulte) pour determiner simultanement, a l'aide d'un modele non lineaire structureen fonction de l'age, la den-