Sampling Considerations in Using Stream Insects for Monitoring Water Quality

Abstract

Three transects in an apparently uniform riffle area of a second-order stream were evaluated for variability in the seasonal and spatial distributions of 23 insect species. Organism densities of species and instars from each season, transect, and location were compared, using three-way analysis of variance, and components of variation attributable to time, transect, and location were estimated. Time differences were most prominent among the early instars and pupae. Transect and location differences were most commonly found in the mature larvae. Filter feeders showed the most time and location differences, whereas collector-gatherers showed more frequent transect differences. The number of 0.09-m2 Hess samples required to detect a 100% change in population densities between two locations ranged from three to 1,560. In single-organism studies, sample sizes may be substantially reduced by restricting sampling or analysis to specific substrates, locations, seasons, or instars. In community production rate studies, necessary sample size can be determined based on a biomass weighted average of individual organism sample sizes necessary to maintain a given confidence interval.