Comparison of relative abundance indices of Brook Trout in Pennsylvania streams

Electrofishing is commonly used to assess the abundance of stream fish. Estimating the absolute population size in a sampled stream reach requires repeated sampling and use of either a mark‐recapture estimator or a removal estimator. Population estimates are then often converted to an estimate of density (number per area sampled) or the number per length of stream sampled. Alternatively, a single electrofishing event may be used to develop an index of relative abundance in some form of catch per unit effort, often expressed as catch per area sampled, catch per length of stream sampled, or catch per minute of shock time. The objective of this analysis was to determine which catch per unit effort index best described absolute estimates of Brook Trout Salvelinus fontinalis density (number/100 m 2 ) and number per length of stream (number/100 m). Electrofishing data from ten Brook Trout streams in the Sinnemahoning Creek watershed of northcentral Pennsylvania from 2010–2018 were included in the analysis. I analyzed data with linear mixed effects models, with individual streams as random effects, to determine which index of catch per unit effort best predicted Brook Trout density or number per length of stream sampled. Results indicated that if interest lies in density, catch per area was the best index of relative abundance, and if number per length of stream is of interest, catch per length was the best index. Catch per minute of shock time may be a less useful index of relative abundance because of multiple factors that could influence the unit of effort used in its calculation.