The impact of data, modeling approaches, and control rules on the performance of management strategies: Applications to West Coast groundfish fishery

摘要

The foundation of sustainable fisheries management is based on setting catch limits that will maintain stock biomasses at or near levels that produce maximum sustainable yield. Determining appropriate harvest rates and the resulting catches depends on the ability of available data and estimation methods to accurately estimate productivity, unfished biomass, and current biomass. However, uncertainty surrounds the ability to estimate these quantities accurately. Incorporating uncertainty into management advice can improve the ability of fishery managers to achieve their objectives, decreasing the risk of overfishing and increasing the probability of sustainably managing fish stocks.;This dissertation examines the impact of data availability to accurately estimate biological parameters and hence stock biomass, alternative estimation methods to set harvest limits that prevent overfishing, and the impact of differing management actions to rebuild and maintain stocks at target biomass levels. Simulation was used to address questions concerning data, modeling, and management decisions by creating a 'true' population to measure and evaluate outcomes. Chapter 1 highlights the importance of continued data collection when fish stocks are rebuilding from low biomasses. Retaining data collection at historical levels during rebuilding allowed for improved parameter estimation, which resulted in reduced variability in estimated stock size with larger average catches during rebuilding. In contrast, when data collection intensity was reduced during rebuilding, the estimates of relative stock size become more variable between assessments, resulting in stocks being prematurely declared rebuilt to the target biomass. Estimates of stock productivity at the time of the first assessment were poor, but continued data collection during rebuilding allowed for improved estimates of steepness. However, when data were limited estimates of productivity were highly variable during rebuilding. Chapter 2 evaluated the impact of data on assessment performance by examining estimation methods designed for application when data limitations prevent the use of complex estimation approaches. Overall, the application of the catch-only (data-limited) and data-moderate estimation methods coupled with management buffers to reduce harvest had mixed results, failing to consistently protect stocks from experiencing overfishing. However, two methods were identified that performed well for rockfish stocks that prevented overfishing when the population had experienced limited historical exploitation. Specific recommendations are made for management of US west coast groundfish and when it might or might not be appropriate to apply such methods.;The final two chapters each performed a management strategy evaluation that examined alternative actions to meet the goals of managers and to explicitly identify the trade-offs among different approaches. Chapter 3 simulated fish stocks that were in an overfished state, below the management-defined minimum stock size threshold, and implemented alternative actions designed to allow rebuilding to the target biomass. Rebuilding approaches that applied precaution when determining rebuilding harvest rates, buffering against future uncertainty, were less responsive to noise in assessment estimates, had fewer changes in harvest rates during rebuilding, and successfully rebuilt stocks on time. The final chapter examined the trade-offs achieved by alternative harvest control rules, the ability of each strategy to maintain stocks at or near target biomass, and prevent stocks from becoming overfished for US west coast flatfish. All harvest control rules examined were effective at maintaining stocks at or near target biomasses and preventing stocks from declining below minimum stock size thresholds. However, trade-offs are associated with each harvest control rule. The harvest control rules that involved the highest proxy biomass targets resulted in higher probabilities of the relative spawning biomass being within 10% of the target biomass, with lower average annual variation in catch, but also resulted in the lowest average catches over the last twenty-five years of the management period. The more aggressive harvest control rules that involved lower proxy biomasses, resulted in higher average catches, but had higher average annual variation in catch and lower probabilities of the stock biomass being within 10% of the proxy biomass.