After site selection and before data analysis: sampling, sorting, and laboratory procedures used in stream benthic macroinvertebrate monitoring programs by USA state agencies

Abstract

A survey of methods used by US state agencies for collecting and processing benthic macroinvertebrate samples from streams was conducted by questionnaire; 90 responses were received and used to describe trends in methods. The responses represented an estimated 13,000-15,000 samples collected and processed per year. Kicknet devices were used in 64.5% of the methods; other sampling devices included fixed-area samplers (Surber and Hess), artificial substrates (Hester-Dendy and rock baskets), grabs, and dipnets. Regional differences existed, e.g., the 1-m kicknet was used more often in the eastern US than in the western US. Mesh sizes varied among programs but 80.2% of the methods used a mesh size between 500 and 600 (mu or u)m. Mesh size variations within US Environmental Protection Agency regions were large, with size differences ranging from 100 to 700 (mu or u)m. Most samples collected were composites; the mean area sampled was 1.7 m 2 . Samples rarely were collected using a random method (4.7%); most samples (70.6%) were collected using "expert opinion", which may make data obtained operator-specific. Only 26.3% of the methods sorted all the organisms from a sample; the remainder subsampled in the laboratory. The most common method of subsampling was to remove 100 organisms (range = 100-550). The magnification used for sorting ranged from 1 (sorting by eye) to 30x, which results in inconsistent separation of macroinvertebrates from detritus. In addition to subsampling, 53% of the methods sorted large/rare organisms from a sample. The taxonomic level used for identifying organisms varied among taxa; Ephemeroptera, Plecoptera, and Trichoptera were generally identified to a finer taxonomic resolution (genus and species) than other taxa. Because there currently exists a large range of field and laboratory methods used by state programs, calibration among all programs to increase data comparability would be exceptionally challenging. However, because many techniques are shared among methods, limited testing could be designed to evaluate whether procedural differences affect the ability to determine levels of environmental impairment using benthic macroinvertebrate communities.