Premium
Evaluation of Sample Design and Estimation Methods for Great Lakes Angler Surveys
Author(s) -
Su Zhenming,
Clapp David
Publication year - 2013
Publication title -
transactions of the american fisheries society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.696
H-Index - 86
eISSN - 1548-8659
pISSN - 0002-8487
DOI - 10.1080/00028487.2012.728167
Subject(s) - fishing , estimator , recreational fishing , statistics , recreation , sampling (signal processing) , estimation , environmental science , fishery , mean squared error , geography , mathematics , ecology , computer science , biology , engineering , systems engineering , filter (signal processing) , computer vision
The waters of the Great Lakes support outstanding recreational fishing opportunities. Total catch and effort estimates obtained from on‐site angler surveys are essential for the management of the recreational fisheries. However, quality of angler survey estimates can be greatly affected by the survey design and estimation approaches used. Using Monte Carlo simulation techniques, we evaluated the effects of two potential sources of bias (disproportional sampling of angler trips and subsampling of the fishing day) on two catch estimators: (1) a multiple‐day estimator that ignores day effects and pools the angler trip data over a multiple‐day period, and (2) a daily estimator that treats the trip data in each day separately. When catch rates are constant among different time periods of the fishing day, the daily estimator produces total catch estimates with little bias, whereas the multiple‐day estimator is prone to bias caused by disproportional sampling of angler trips. When catch rates vary among different periods of a fishing day, the daily estimator produces biased estimates of total catch when the fishing day is subsampled, whereas the multiple‐day estimator is less affected by the variation in daily time‐period catch rates and subsampling of fishing days. Quality of total catch and effort estimates, in terms of root mean square error and coverage probability of confidence intervals, is poor when the number of days sampled each month is low and fishing days are subsampled.