Effectiveness of multiplate invertebrate samplers, periphytometers, and electrofishing for biomonitoring in streams

Abstract

Electrofishing, multiplate invertebrate samplers, and periphytometers were used to collect biological samples from 10 sites in two southeastern coastal plain streams. The objectives were to evaluate two factors that affect the quality of biomonitoring data: inter-replicate variability and sensitivity to habitat effects that could be confounded with environmental impacts. Abundance related estimates (density, catch-per-unit-effort, biomass, chlorophyll a) exhibited high inter-replicate variability necessitating the collection of large numbers of samples (25) to detect small (40%) differences among sites. Species number estimates exhibited much lower inter-replicate variability and sample number requirements (6 samples to detect a 40% difference). Electrofishing generally required fewer samples and less time than the other methods to obtain comparable levels of precision. Species composition and abundance estimates produced by all methods were affected by longitudinal changes in habitat that caused differences between control and impact sites that could be mistaken for impacts. Biomonitoring programs can be improved by assessing the relationship between statistical power and replicate number, collecting as many replicates as is feasible, collecting pre-impact and post-impact data when possible, matching control and impact sites on habitat, comparing the ecological characteristics (e.g. pollution tolerance) of organisms at control and impact sites, and using comparisons based on species number rather than abundance.