Relative Efficiency of a Pump for Sampling Larval Gizzard and Threadfin Shad

Efficiency of a 1.18-m3/min-capacity pump for sampling larval fishes quantitatively was compared with that of a 0.25-m2 plankton net during monthly nighttime collections in a reservoir. Net samples contained seven fish taxa whereas only two were collected by pumps; however, gizzard shad Dorosoma cepedianum and threadfin shad D. petenense (which usually could not be distinguished from each other as small larvae) accounted for more than 97% of the specimens caught in both samplers. Regression analyses indicated that pump catch rates (expressed as number per 1,000 m3) averaged 39.5% of the net catches for 3–6-mm shad larvae, 8.5% for 7–12-mm larvae, and 1.1% for 13–20-mm larvae. Coefficients of variation among triplicate pump samples averaged about 2.6 times higher than they did for net samples for 3–6-mm larvae, 4.5 times higher for 7–12-mm larvae, and 9.0 times higher for 13–20-mm shad. Although pumps usually provided density estimates that were lower and less precise than those obtained from nets, linear relationships between pump and net catch rates indicated that pumps can provide a measure of relative abundance useful for comparing densities of larval fish in habitats where sampling with nets is not feasible.