Fisheries Research Report 2028

Michigan Department of Natural Resources
Fisheries Research Report No. 2028, 1996.

Evaluation of Michigan Creel Survey Catch Rate Estimator

Roger N. Lockwood

Michigan Department of Natural Resources
Institute for Fisheries Research
212 Museums Annex Building
Ann Arbor, Michigan 48109-1084

Abstract.-Access-site creel surveys are conducted at numerous Great Lakes ports and on inland waters each year in Michigan to estimate angling effort and catch. Estimated catch is the product of estimated angler hours and catch rate. Catch rate has been determined by averaging angler party catch per hour by species, angling mode, and time period. This method for calculating catch rate is not weighted by number of anglers in the party or by length of fishing trip which could lead to biases in estimates. Effects of bias on accuracy and precision of catch per hour estimates were measured using Monte Carlo sampling techniques on 132 data sets from Michigan access-site creel surveys. Each data set was considered a discrete population and population catch rate parameters were compared with sample catch rate estimates. Estimated mean catch rate by angler party was significantly greater, P < 0.05, than population catch rate parameters in 82 data sets and significantly less in 49 others. Due to trip length and angler party size bias, the sample confidence limits were incorrectly represented in 123 of 132 data sets. Biases associated with averaging angling party catch rates were found to be prevalent in Michigan access-site creel surveys. I concluded that averaging angler party catch rates is inappropriate for Michigan access-site creel surveys. Calculating catch rate by dividing total catch by total hours from angler interviews eliminates the bias.