Modification and Statistical Evaluation of a Standard Sampling Method for Stream Invertebrates

Abstract

Stream health monitoring of Irondequoit Creek was begun as part of a Remedial Action Plan (RAP) for the Rochester Embayment. The New York State Department of Environmental Conservation (NYSDEC) assesses wadeable riffles by kick sampling four, 5-m transects for five min each. A 100-organism subsample of each transect is analyzed to determine stream health. Within the biomonitoring community, there is debate regarding appropriate areal (spatial) and numerical subsampling methods as they relate to biomonitoring. A pilot study indicated that reduced areal sampling provided comparable assessments to a larger area assessed in an earlier study conducted at the same location in the same month. I found that 10-20% sampling effort did not provide equivalent assessments and that 80-90% sampling effort provided virtually identical assessments to 100% sampling effort. Depending on the biotic index (Taxonomic Richness; Ephemeroptera, Plecoptera and Trichoptera; Hilsenhoff Biotic Index and Percent Model Affinity), 30-60% sampling effort gave assessment results equal to 100% sampling effort. Overall, 50% sampling effort gave stream health assessments equivalent to 100% sampling effort, and reduced sampling time.