Precision of channel catfish catch estimates using hoop nets in larger Oklahoma reservoirs

Abstract

Hoop nets are rapidly becoming the preferred gear type used to sample channel catfish Ictalurus punctatus , and many managers have reported that hoop nets effectively sample channel catfish in small impoundments (<200 ha). However, the utility and precision of this approach in larger impoundments have not been tested. We sought to determine how the number of tandem hoop net series affected the catch of channel catfish and the time involved in using 16 tandem hoop net series in larger impoundments (>200 ha). Hoop net series were fished once, set for 3 d; then we used Monte Carlo bootstrapping techniques that allowed us to estimate the number of net series required to achieve two levels of precision (relative standard errors [RSEs] of 15 and 25) at two levels of confidence (80% and 95%). Sixteen hoop net series were effective at obtaining an RSE of 25 with 80% and 95% confidence in all but one reservoir. Achieving an RSE of 15 was often less effective and required 18-96 hoop net series given the desired level of confidence. We estimated that an hour was needed, on average, to deploy and retrieve three hoop net series, which meant that 16 hoop net series per reservoir could be "set" and "retrieved" within a day, respectively. The estimated number of net series to achieve an RSE of 25 or 15 was positively associated with the coefficient of variation (CV) of the sample but not with reservoir surface area or relative abundance. Our results suggest that hoop nets are capable of providing reasonably precise estimates of channel catfish relative abundance and that the relationship with the CV of the sample reported herein can be used to determine the sampling effort for a desired level of precision.