Evaluation of Design-Based Estimators in Federal Groundfish Fisheries off Alaska
Jennifer Cahalan, Jason Gasper, and Jennifer Mondragon
- Price: $10.00 Sale: $0.00
Incremental changes to the methods used to estimate total catch in the federal groundfish and halibut fisheries in Alaska have provided opportunities to expand the types of statistics used for catch estimation. These changes coincide with new monitoring requirements associated with catch share programs, improvements in monitoring and reporting technology, improvements in sampling design used to collect at-sea and shoreside information, and improvements to database architecture and estimation routines. Taken together, these important changes allow us to take advantage of a randomized sampling design and consider using design-based estimators for estimation. At present, the expansion of observer sample data to estimate discarded catch for observed trips is based on imputation routines and the use of ratio estimators. However, recent research indicates design-based estimators are robust to highly variable sample data that are encountered for less common species, such as bycatch of nontarget species. Further, design-based estimators allow for the estimation of precision under the hierarchical sampling procedure currently employed. In this paper we investigate patterns of total discards and evaluate how the differences in the precision of trip-total discard estimates relate to species rarity. We focus on fisheries with complete observer coverage (all trips observed), recognizing the methods used in this study may have application to fisheries with random deployment of observers (partial coverage).
- Item number: AK-SG-15-01i
- Year: 2015
- Pages: 16
- DOI: https://doi.org/10.4027/fbgics.2015.09