Catchability model selection for estimating the composition of fishes and invertebrates within dynamic aquatic ecosystems

We evaluated alternative catchability models for estimating the composition of fishes and decapods in two coastal rivers sampled via capture-recapture boat electrofishing, seine removal sampling, and throw trap removal sampling. Information criteria were used to select between linear and nonlinear catchability models and assess heterogeneity in catchability coefficients across populations, sampling reaches, and sampling events. The selected electrofishing catchability model assumed linearity between catch per unit effort and population density, with heterogeneity in catchability coefficients among taxa and between rivers. Linear models were selected for seine and throw trap samples, with heterogeneity in catchability coefficients among taxa, sampling reaches, and sampling events. We emphasize the importance of estimating population density, rather than abundance exclusively, when accurate community composition estimates are needed. The analytical techniques employed have broad-scale application in quantifying key components of ecosystems by integrating information from multiple sampling gears targeting different taxa or trophic guilds and in accounting for variable sampling intensity and catchability heterogeneity in the estimation of community composition.