Assessment of Southeast Alaska Pink Salmon Abundance Based on Commercial Catch and Effort and Sex Ratio Data
J. Zheng and O.A. Mathisen
- Price: $0.75 Sale: $0.00
This is part of Fishery Stock Assessment Models| Format | Price | |
|---|---|---|
| PDF download [486.5 KB] Bypass cart and download |
Free | Add to Cart |
Description
Pink salmon (Oncorhynchus gorbuscha) is the most abundant salmon species in Southeast Alaska and supports an important commercial fishery. Like most salmon fisheries in Alaska, pink salmon fisheries in Southeast Alaska are managed by a fixed escapement policy. To achieve a targeted escapement, managers must know the abundance of the incoming spawning run. The accuracy of abundance information acquired inseason substantially affects the manager’s ability to achieve management objectives. To improve accuracy of inseason forecasts of southern Southeast Alaska pink salmon runs, we incorporated sex ratio information into inseason forecast models to annually adjust timing and shape of the run timing curves. First, we developed a sex ratio index and subsequently evaluated three inseason forecast models—linear, nonlinear, and combined—using this index and cumulative catch of all gears or cumulative catch per unit effort of the seine fishery from 1983 to 1997. Based on a cross-validation evaluation of forecast accuracy, the nonlinear model outperformed the linear and combined models. Cumulative catch per unit effort was a better predictor than cumulative catch in the first three weeks (weeks 28-30) of a fishing season, and vice versa in the remaining five weeks. Inseason abundance estimations greatly improved the preseason forecasts. Incorporating sex ratios into inseason forecast models correctly adjusted the run timings during a large majority of years and thus improved overall forecasts starting in the second week. In weeks 29-32, the best performing model using sex ratios improved forecasts more than 30% over the best model without using sex ratios; improvements included averages of relative forecast errors, absolute deviations, or squared residuals. Averages of relative forecast errors in weeks 29-34 were less than 24% for the best performing model using sex ratios and less than 38% for the best model without using sex ratios, compared to 51% for preseason forecasts. Average relative forecast errors from the best model were less than 20% before the run midpoint and less than 14% after the run midpoint.
Item details
- Item number: AK-SG-98-01z
- Year: 1998
- DOI: https://doi.org/10.4027/fsam.1998.26
