Efficacy of two-pass electrofishing employing multiple units to assess stream fish species richness

We utilized two-pass, multiple unit backpack electrofishing data to assess stream fish species richness. Twenty-six streams in the Monongahela River basin of Pennsylvania were sampled over 200 m standard reaches. Effort, as measured by the number of electrofishers employed, was increased as a function of increasing mean stream width (<30 m) to a maximum of three operated abreast across the stream. We calculated proportional fish species richness (sˆ) and the probability of detection (pˆs1) for each species. Values of sˆ ranged from 1 to 27, whereas values of pˆs1 ranged from 67 to 100 across all streams. Median pˆs1 values did not differ significantly among effort categories nor did the probability of capturing a new species on pass two (P > 0.05). Narrow 95% confidence intervals around sˆ values attest to the validity of this approach in estimating species richness. Small percids (e.g., darters) and schooling (e.g., cyprinids) fishes evaded capture on the first pass more frequently than centrarchids, ictalurids, or catastomids. Our results indicate that increasing effort (as measured by number of electrofishing units employed incrementally over a standardized length of the sampling reach) provides a practical and efficient alternative to increasing sampling reach length with increasing stream width when sampling for species richness.