Point sampling by boat electrofishing: A test of the effort required to assess fish communities

Point sampling by electrofishing is often used to study fishes in large rivers and lakes whereby a specific location is electrofished without moving the anode. Short (1-5-s) samples are taken under the belief that many small samples are preferred over a few large ones for statistical analyses. However, this typically results in relatively little time spent sampling fishes compared with time spent measuring abiotic factors and traveling among sites. We evaluated the optimal sampling duration and number of replicates per site to balance sample size and number for community-level studies. In 2004, 165 point samples were taken from shallow Canadian waters of the Detroit River. Sites were continuously electrofished for 2 min (eight 15-s intervals), and a second replicate of 2 min was taken after a pause. Subsets of the data were used to compare various designs of sampling duration and number of replicates. A sampling design of two replicates of I min appeared to be ideal because it balanced a large gain of information with a small increase in effort. This design would allow 35-50 sites to be sampled per day, depending on the detail of abiotic measurements. Compared with data from the first 15-s interval only, sampling for two replicates of I min resulted in fewer null (no fishes captured) samples (19% instead of 53%). The number of common (found at > 5% of samples) species also increased from 12 to 19. By increasing the effort for point sampling by electrofishing at each site, a better understanding of the fish assemblage was obtained. This allows for more complete analyses of community composition and habitat preference.