Discard Estimates from Self-Reported Catch Data: an Example from the U.S. Northeast Shelf

Abstract

Discarded catch can account for a significant amount of fishing mortality and can be a major source of uncertainty when calculating total removals for stock assessments and management. Observers provide estimates on the types and quantity of fish that are caught, but observer coverage of a fleet can often vary directly with funding. On the Northeast U.S. shelf, the Northeast Fisheries Observer Program (NEFOP) has collected data on catch and discards for assessment and management purposes since 1989. The Cooperative Research Study Fleet (hereafter “Study Fleet”), working with the Northeast Fisheries Science Center, collects self-reported catch and discards from commercial vessels; however, the Study Fleet program was not developed as an observer program, and these data are not used for management. The number of vessels reporting data in the Study Fleet program is much smaller than the total number of vessels recorded through the observer program (NEFOP plus the At-Sea Monitoring Program), and unlike the observer program, the Study Fleet program is not conducted within a statistically designed sampling program. We compared observer and Study Fleet data for six species collected with otter trawls from 2007 to 2014 to determine whether the reported kept catch and discard values and the total estimated discards from the two programs were similar. In general, the estimates of catch and discards were similar, but due to the large number of samples, significant differences were discernible. Of the 20 comparisons of total estimated discards between the two programs, 13 had similar magnitudes and trends, 4 had similar trends with different magnitudes, and 3 were not similar. Our results indicate that for the subset of species and gear tested, the fishing industry has the capacity to accurately self-report information. With a rigorous quality control program, including appropriate audit and compliance checks to ensure confidence in the data, programs collecting self-reported information for assessment and management may be possible in the future.

Received October 14, 2016; accepted May 26, 2017Published online September 5, 2017

ACKNOWLEDGMENTS

We thank the Study Fleet members for their participation and hard work. We are especially grateful to the members of the Cooperative Research Branch for making this program possible and to the observer program and the fleet for all their days spent collecting and processing the data. Mike Palmer, Dave Richardson, and Kelly Denit graciously reviewed drafts of the report, and Jon Hare provided valuable ideas and suggestions. We appreciate two anonymous reviewers whose comments improved the manuscript.