Hierarchical model to predict common carp and bigmouth buffalo abundance from electrofishing data

Abstract

Catch per unit effort (CPUE) is used as an index of fish abundance under the premise that changes in CPUE result from changes in true density. However, catchability may also vary based on environmental conditions that affect observed CPUE. We developed a hierarchical model for estimating common carp ( Cyprinus carpio) and bigmouth buffalo ( Ictiobus cyprinellus) relative abundance with electrofishing survey data from six shallow lakes in northwest Iowa, USA, between 2018 and 2020. Common carp catchability was negatively associated with lake perimeter but unrelated to lake surface area, water depth, Secchi depth, temperature, and month of sampling. Bigmouth buffalo catchability was negatively associated with Secchi depth and water temperature and unrelated to other environmental variables. Hierarchical model posterior distributions of bigmouth buffalo density were less precise than Schnabel estimates, whereas common carp posterior distribution abundance estimates were more precise than Schnabel estimates. Our results indicate that hierarchical models can be a viable substitute for labor-intensive capture–mark–recapture methods to estimate unknown latent variables like relative abundance, and could be applied to other species, sampling gears, and management frameworks.