An evaluation of the efficiency of minnow traps for estimating the abundance of minnows in desert spring systems

Authors: Paul* Scheerer, James Peterson, Shaun Clements

Year: 2015 (xlvii)

Abstract

It is important to accurately quantify population parameters for native desert fishes of special conservation concern. We evaluated the efficiency of baited minnow traps for estimating the abundance of two ESA listed species, Foskett Speckled Dace Rhinichthys osculus ssp. and Borax Lake Chub, Gila boraxobius, in desert spring systems in southeastern Oregon. We estimated capture and recapture probabilities using the Huggins closed-capture estimator and compared the abundance estimates with the commonly used Lincoln-Petersen estimator. We evaluated alternate sample designs using simulation. Trap capture probabilities averaged 23% and 26% for Foskett Speckled Dace and Borax Lake Chub, respectively, but differed substantially among sample locations, through time, and nonlinearly with fish length. Recapture probabilities for Foskett speckled dace were, on average, 1.6 times greater than first capture probabilities, suggesting trap happy behavior. We found the Lincoln-Petersen estimator underestimated Foskett Speckled Dace and Borax Lake Chub abundance by 48% and 20%, respectively comparing to the Huggins estimator. These biases were due to variability in capture and recapture probabilities. Simulation of fish monitoring that included the range of capture and recapture probabilities observed indicated that bias and error decreased with increasing capture occasions and that variability in these probabilities over time of +/-10% greatly reduced the ability to detect annual decreases in abundance using raw catch data. Failure to account for this variability in capture and recapture probabilities can lead to poor data quality and study inferences. Thus, we recommend employing sampling designs and estimators that can account for this variability.