Sources of Bias in Harvest Surveys for Marine Fisheries

Estimates of commercial fisheries harvest and effort are among the most widely used monitoring strategies. Due to their tendency toward unknown accuracy and precision, "fishery statistics" are also among the most widely criticized measures of environmental health. We are in the process of estimating the sources of bias in the commercial harvest estimation systems in Maryland and Virginia. The Maryland Department of Natural Resources uses mail-out monthly questionnaires to estimate the harvest of blue crabs, using a stratified random sampling design. We are attempting to: a) characterize non-respondents; b) characterize reports of zero catch; c) characterize biases in non-zero responses; d) explore the design and estimator efficiencies. Personal interviews and mail-out questionnaires, probably using randomized response techniques will be used. The Virginia Marine Resources Commission uses monthly site visits to each licensed dealer to estimate the harvest of blue crabs. We are: a) attempting to use inferential recovery models for past records, b) studying the feasibility of harvester based surveys. This work should provide more accurate commercial harvest estimates in Maryland and Virginia and may lead to Bay-wide fisheries statistics programs.