Biological Communities

Data Collection

Biological data were collected from the 150- or 250 m long stream reaches established for characterizing instream habitat. Standard NAWQA sampling protocols were used to collect benthic macroinvertebrates and algae (Moulton and others, 2002) during late May to mid-June 2003. For both organism types, a quantitative sample was composed of five subsamples from riffles with cobble/gravel substrates. This sample was called the richest targeted habitat (RTH) sample because, in these streams, riffles were presumed to contain the richest assemblage of algae and macroinvertebrates. RTH macroinvertebrate subsamples consisted of Slack samples (each 0.25 square meters, 500-micron mesh net) from five separate riffle areas in the sampling reach and combined to form a single composite sample of 1.25 square meter area. RTH algal subsamples were collected by scraping the upper surface of cobbles collected from five riffle areas (3 to 5 cobbles from each area) in conjunction with the macroinvertebrate RTH samples. A foil template of each rock surface was collected and later digitized to determine the area sampled.

In addition, for macroinvertebrates, a qualitative multihabitat (QMH) sample was collected, which consisted of macroinvertebrates collected from as many habitats in the stream reach as were accessible. The QMH sample was collected using a 500-micron mesh dip net supplemented with hand picking of substrates. Sampling effort (measured as time) was apportioned as equally as possible among accessible habitats in the sampling reach. For algae, an additional quantitative sample was collected from a composite of five samples from depositional targeted habitats (DTH). DTH samples were collected by inverting a 47-mm diameter plastic petri dish and gently pressing it into the sediment surface, and then sliding a spatula under the petri dish to trap the sediment and removing the petri dish full of sediment.

Fish communities at each site were sampled during early summer low flows (mid- to late June 2003) using published protocols (Moulton and others, 2002). Two-pass electrofishing was used at each site. A backpack electrofisher was used to sample all sites except the Tar River (02081190) and Deep Creek (02085430) where a tote barge, equipped with a 2.5 generator powered pulsator (GPP) and two anode poles, was used. Fish were identified, examined for abnormalities, weighed, and measured in the field by Dr. Wayne Starnes and Gabriella Hogue of the North Carolina State Museum of Natural Science (NCSMNS). Voucher specimens and specimens that could not be definitely identified in the field were returned to the NCSMNS for processing and incorporation into the NCSMNS permanent collections.

Laboratory Analysis

Macroinvertebrate samples were preserved in 10-percent buffered formalin and sent to the USGS National Water Quality Laboratory (NWQL) in Denver, Colorado, for taxa identification and enumeration. Invertebrate samples were processed using standard NAWQA protocols (Moulton and others, 2000) for RTH samples using the randomized 300-organism count and for QMH samples using the method of fixed processing time designed to maximize the number of taxa enumerated samples.

Aliquots of the algal RTH samples were taken to assess assemblage composition and biomass as chlorophyll a (Chl a) and ash-free dry mass (AFDM). For DTH samples, only assemblage data were assessed. The assemblage aliquots were preserved in 5-percent buffered formalin and sent to the Philadelphia Academy of Natural Sciences for identification and enumeration (Charles and others, 2002). The biomass aliquots were filtered on 45-micron glass-fiber filters, packed in dry ice, and sent to the NWQL for analysis.

Data Processing

Prior to analysis, biological datasets were examined for errors and corrected for taxonomic ambiguities. Taxonomic ambiguities arise when organisms from a particular sample or group of samples are not identified to the same taxonomic level. For example, an ambiguity exists in a sample if some organisms are identified to Genus (for example, Hydropsyche sp.) and some organisms are identified to species within that Genus (for example, H. sparna, H. betteni). In this case, sparna and betteni are children of the ambiguous parent Hydropsyche. The presence of taxonomic ambiguities is a problem in determining taxa richness (for example, is taxa richness in the above example 1, 2, or 3?) or when comparing the taxonomic composition of one or more samples by using techniques such as ordinations, cluster analysis, similarity indices, or diversity indices.

Ambiguities in the invertebrate and algal data were resolved using software specifically developed for use in the NAWQA Program—Invertebrate Data Analysis System (IDAS, version 3.9.5; Cuffney, 2003) and the Algal Data Analysis System (ADAS, version 2.4.5). The ADAS program is a modification of IDAS for use with algae. Ambiguities in the RTH invertebrate and algal samples, and in the DTH algal sample data were resolved by applying an option in these programs that processes samples separately by site and then distributes the abundance of an ambiguous parent among children in proportion to the relative abundance of each child. This procedure maximizes taxa richness without affecting taxa abundance.

To create a comprehensive list of taxa present at each site, a qualitative richness dataset (QQ) was created for invertebrates. This dataset consisted of a combination of all taxa found in the RTH and QMH samples. Ambiguities in the QMH and QQ samples were handled by deleting the ambiguous parents, since the taxonomic information carried by ambiguous parents already resides in the children. Fish data were almost entirely identified to species level. Consequently, there was very little ambiguity in those data. A small number of individual fish were identified to a higher taxonomic level. These fish were eliminated from the analysis.

Biological metrics were calculated for fish, invertebrates, and algae. Metrics are individual variables or combinations of variables that emphasize specific data characteristics. They commonly are used in bioassessments to reduce the complex site-by-species matrix to a few variables that are thought to have significance ecologically and(or) are indicative of water-quality changes (Barbour and others, 1999). Metrics calculated from the invertebrate and algal data were based on measures of abundance, richness, functional groups (for invertebrates), biomass (for algae), tolerance, and indices of diversity. Invertebrate traits used in the calculation of metrics are from Barbour and others (1999) and Cummins (1973). Algal classifications were compiled from Lowe (1974), Bahls (1993), and Van Dam and others (1994) (Attributes file version 8, S.D. Porter, USGS Colorado Water Science Center, written commun., 2006). Metrics calculated from the fish data were based on richness, abundance, biomass, total length, tolerance, trophic guild (for example, herbivore, insectivore), and traits (for example, reproductive strategy, substrate preference). Fish tolerance classifications and trophic guilds were from the North Carolina Department of Environment and Natural Resources (2001), and trait classifications were from Goldstein and Meador (2004).