Accounting for variable effort among passes, fish size, and spatial heterogeneity in electrofishing depletion estimates

Objective The first objective of this study was to develop electrofishing depletion models that allowed effort to vary among passes while accounting for capture probability that varied with fish size. The second objective of this study was to extend the models from estimating abundance at a single site to multiple sites using a hierarchical model. Methods Simulations were used to evaluate four potential models under four different scenarios of electrofishing effort and capture probability. Further, depletion models were fit to electrofishing data from five sites in the Henrys Fork Snake River basin and two sites in the Big Lost River basin, Idaho. Results The model that accounted for heterogeneity in effort among passes and capture probability among fish size (i.e., size–effort model) resulted in the lowest average bias and root-mean-squared error in the simulations. In addition, the size–effort model resulted in greater estimates of abundance in the simulations and the empirical analysis than the other models evaluated. Conclusions Results of this study suggest that bias that is common in electrofishing depletion estimates can be reduced by using models that explicitly account for assumptions that are often violated in practice, such as nonconstant effort and heterogeneity in capture probability among size-classes of fish and passes.