A Bayesian approach to estimating Atlantic salmon fry densities using a rapid sampling technique

Abstract

Abstract Removal sampling by electric fishing is widely used for assessing fish population size. As it is manpower consuming, less‐demanding techniques have been proposed to increase the number of sites covered with the same human resources. These techniques essentially provide relative abundance measures. To be used for absolute abundance estimation, they need to be inter‐calibrated with another method such as removal sampling. Here, hierarchical Bayesian modelling framework is used for this inter‐calibration because it allows an estimate of absolute abundance from abundance index data alone while accounting for the main sources of uncertainty. It is applied to a 0+ juvenile Atlantic salmon, Salmo salar L. data set from 21 sites on the River Faughan, Northern Ireland. A positive relationship was found between the abundance index (number of fish caught in 5 min of actual electric fishing) and the density. The estimates from the index of abundance alone are fairly imprecise, but still allow differentiation of contrasting levels of fish density.