Sampling Effort and Estimates of Species Richness Based on Prepositioned Area Electrofisher Samples

Estimates of species richness based on electrofishing data are commonly used to describe the structure of fish communities. One electrofishing method for sampling riverine fishes that has become popular in the last decade is the prepositioned area electrofisher (PAE). We investigated the relationship between sampling effort and fish species richness at seven sites in the Tallapoosa River system, Alabama, based on 1,400 PAE samples collected during 1994 and 1995. First, we estimated species richness at each site using the first-order jackknife and compared observed values and jackknife estimates of species richness to estimates based on historical collection data. Second, we used a permutation procedure and nonlinear regression to examine rates of species accumulation. Third, we used regression to predict the number of PAE samples required to collect the jackknife estimate of species richness at each site during 1994 and 1995. We found that jackknife estimates of species richness generally were less than or equal to estimates based on historical collection data. The relationship between PAE electrofishing effort and species richness in the Tallapoosa River was described by a positive asymptotic curve as found in other studies that incorporated different electrofishing gears in wadable streams. Results from nonlinear regression analyses indicated that rates of species accumulation were variable among sites and between years. Across sites and years, predictions of sampling effort required to collect jackknife estimates of species richness suggested that doubling sampling effort (to 200 PAEs) would typically increase observed species richness by not more than six species. However, sampling effort beyond about 60 PAE samples typically increased observed species richness by less than 10%. We recommend using historical collection data in conjunction with a preliminary sample size of at least 70 PAE samples to evaluate estimates of species richness in medium-sized rivers. Seventy PAE samples should provide enough information to describe the relationship between sampling effort and species richness and thus facilitate evaluation of a sampling effort.