Feasibility and Evaluation of Probabilistic On‐site Surveys for Estimating Angler Use and Catch in High Mountain Lakes

Abstract

Estimating angling effort and catch at high mountain lakes can be difficult due the abundance, remoteness, and diffuse nature of angling effort that typify high mountain lake fisheries. In this study, a simulation was used to evaluate the accuracy of catch and effort estimates derived using on‐site access–access and roving–access creel surveys at a complex of 35 high mountain lakes. Five levels of angling effort and catch at the 35 lakes were simulated, and effort varied from 3,278 to 68,741 h and catch varied from 1,737 to 50,525 fish over the duration of the season. Access–access creel surveys had an average of 32% relative error and roving–access surveys had an average of 17% relative error in estimates of aggregate (i.e., at all 35 lakes) angling effort and catch when one creel surveyor was used. Estimates were relatively robust to temporal and spatial changes in patterns of effort and catch rate over the duration of the season. Relative error was inversely related to the amount of angling effort, catch, and creel surveyors. Roving–access surveys outperformed access–access surveys at all levels of sampling effort. Bias and relative error of estimates of angling effort and catch were greater at the individual‐lake scale than when estimates were derived for all 35 lakes. Results of this study suggest that on‐site surveys can provide relatively accurate estimates of angling effort and catch at as many as 35 lakes with minimal sampling effort. Received March 16, 2015; accepted August 14, 2015

Sample Size Estimation for On‐Site Creel Surveys

“…However, it can be difficult or cost-prohibitive to obtain an adequate sample size of angler interviews for each surveyed day. Consequently, estimates of variance for catch or catch rate may be unreliable at low sample sizes (Jones et al 1995;Lockwood et al 1999) or unobtainable when only one angler or no anglers are surveyed (McCormick 2015). To circumvent this problem, several authors have advocated for the use of a multi-day or stratum estimator when estimating catch rate (Rasmussen et al 1998;Su and Clapp 2013;McCormick 2015).…”

“…Development of general sample size requirements for creel surveys is not a new concept (e.g., Parkinson et al 1988;Lester et al 1991;Malvestuto and Knight 1991). However, with fisheries and wildlife agencies across North America experiencing constrained budgets (e.g., Archibald et al 2014), there is a renewed interest in refining the sample sizes that are needed to achieve acceptable levels of precision in estimates of angler effort, catch rate, and harvest (Deroba et al 2007;McCormick et al 2013;McCormick 2015). Determining sample size for creel surveys a priori is difficult because creel survey designs are often complex and commonly lack replication of sampling units within a day (e.g., only one shift or count is conducted daily; Su and Clapp 2013).…”