Among- and Within-Transect Variability in Estimates of Shad Abundance Made with Hydroacoustics

Abstract

We examined variance and distribution of gizzard shad Dorosoma cepedianum and threadfin shad D. petenense using hydroacoustic techniques at three sites in Lake Texoma, Texas–Oklahoma, in August 1991. Density estimates among transects were significantly different at one site during the day and two sites at night. Fish densities were uniformly distributed among transects within sites during day and night surveys, except at one site during the day where the distribution was random. Fish were aggregated into small (≤19 m), dense clumps (up to 45 fish/m2) separated by 25 m or more during the day. In contrast, at night there were fewer, larger aggregations (≥135 m), which were closer together (≤19 m) than during the day. Coefficients of variation for density estimates within transects exceeded 100% over short subsampling intervals (7.5 m) within transects during day and night samples but decreased as subsampling interval length increased. Withintransect variation was significantly lower for night surveys than day surveys. Coefficients of variation of 20% or less can be achieved with transects about 150 m long at night. Serial correlation which can bias density estimates was low (≤13%) during the day and (≤16%) at night.