The objective of semi-quantitative targeted-habitat sampling is to obtain representative samples of benthic invertebrate communities from two instream habitat types: (1) a habitat supporting the faunistically richest community of benthic invertebrates (r ichest-targeted habitat, RTH), usually a fast-flowing, coarse-grained riffle; and (2) a fine-grained, organically rich depositional habitat (depositional-targeted habitat, DTH), usually a pool. Semi-quantitative sampling methods characterize the structure of invertebrate communities in terms of the relative abundances of each taxon rather than absolute density. Information on community structure is useful in constructing a variety of biological indexes, such as diversity, community similarity, and funct ional and trophic groupings. These groupings are used to (1) interpret how the community is functioning (For example, what is the source of its energy?); (2) compare sites (For example, do sites with the same physical and chemical characteristics have the same community structure?); and (3) relate community structure to physical, chemical, and land-use characteristics affecting water quality (For example, do community characteristics correlate with land-use and chemical characteristics?).
The type of sampler used to collect a semi-quantitative sample depends upon the depth, velocity, and substrate within the instream habitat type sampled. The choice of sampler is made with the advice of the regional biologist, North Carolina Ecology Group, and local biologists. Artificial substrates, such as multiplate samplers (fig. 3L), substrate basket samplers (fig. 3M), artificial leaf packs (fig. 3N), and artificial snags are used in situations where natural substrates cannot be sampled because of inaccessibility of the habitat, cost of sample collection, or safety issues associated with collecting the samples (for example, the use of divers in large, fast-flowing rivers). However, artificial substrates have a number of limitations that should be factored into their use: (1) they require two trips to the sampling site (installation and removal) separated by an interval of time to allow for colonization of the substrates, which depends on season, discharge, temperature, and other environmental variables; (2) they are susceptible to loss and vandalism; (3) they are biased toward species that are actively colonizing at the time of placement; (4) they often do not accurately depict the types or relative abundances of the benthic invertebrates at a site; and (5) they may not be sensitive to changes in water quality associated with changes in land use.
Under certain circumstances, such as in large, deep rivers with cobble, boulder, or bedrock substrate, artificial substrates may offer the only viable means of obtaining community samples. The preference in these circumstances is to use artificial substrates that mimic natural substrates such as "barbecue baskets" (fig. 3M) filled with indigenous rocks (Britton and Greeson, 1988) or artificial snags floating at the stream margins. All artificial substrates are allowed to colonize for a minimum of 6 weeks unless locally available data suggest that a longer or shorter colonization period is more appropriate.
Generally, habitats associated with deep, sandy-bottomed, fast-flowing rivers do not yield sufficient numbers of taxa to warrant the effort required to obtain representative samples. In these situations, habitats associated with the margins of the main channel and islands, such as snags and macrophyte beds, generally support accessible and faunistically rich assemblages of invertebrates (Benke and others, 1984; Thorp, 1992). The use of these