Northern Prairie Wildlife Research Center
Fish were collected with seines in headwaters and with seines and trap nets on the mainstem. Seines were 1.2 m deep bag seines (4.7 mm mesh) that were stretched from bank to bank or to a maximum of 15 m. Seining was done between block nets (6.3 mm mesh) spaced 50 m apart, except when water velocity precluded blocking at one mainstem reach in 1995 and four mainstem reaches in 1996. We attempted to make three continuous downstream seine hauls between block nets that enclosed riffle and run habitat in each reach. This protocol was usually accomplished in run habitats, but the rocky bottom and swift flows in riffle habitats forced shorter and more irregular hauls. A double-wing trap net (1 m wide × 1.2 m deep, 9.5 mm mesh) was set in areas too deep for seining. Trap net wings were usually stretched from bank to bank with the trap opening downstream. Trap nets were set for about 12 hours over one night at each reach on the mainstem.
Fishes were anesthetized with carbon dioxide before being sorted by species, counted, and released. Most young-of-the-year specimens were too small to be captured, but those caught were identified and counted. Channel catfish, river carpsuckers (Carpiodes carpio) and white suckers (Catostomus commersoni) collected from the main river were also measured for total length and weight. Of the three Hybognathus species found in the Moreau River, the western silvery minnow (H. argyritis) and plains minnow (H. placitus) are difficult to separate, so in 1995 they were grouped as Hybognathus spp. However, in 1996 most were identified to species in the field by methods developed by Loomis (1997). The brassy minnow (H. hankinsoni) was easily distinguishable. Voucher specimens of all species are retained at South Dakota State University.
Fish community and population data were obtained each year. Species richness (number of species) was determined for each reach on the mainstem. We tested differences between years by using Chi-square analysis (SYSTAT 1996). Catch-per-unit-effort (CPUE) data for seines were the number of fish per meter of river seined. Relative weight for channel catfish, river carpsucker, and white sucker was determined by using the relative weight index (Wr = 100 × fish weight/standard weight) of Wege and Anderson (1978) and standard weight equations for each species (Stephen 1978, Anderson 1980, Brown et al. 1995). The data were then used to construct a curvilinear equation from which predicted Wr values were plotted. Mean Wr was calculated for standard-length categories for channel catfish (Murphy et al. 1991), but length category standards had not been proposed for river carpsucker or white sucker. To assign length category values to these species, we used their world record length (NFWFHF 1995) as recommended by Gablehouse (1984). Relative weight is an index of condition, or well being, that integrates the ecological and physiological status of a population and is used to compare population health over time or among watersheds (Murphy et al. 1991).
We made quantitative and qualitative measurements at transects spaced three mean stream widths apart at each of 11 stations that were about 500 m long (Harrelson et al. 1994, Simonson et al. 1994). Numerous physical measurements were made by using surveying equipment (Loomis 1997), but we report here only those relating to fish sampling. Gradient was calculated by subtracting the water elevation (± 0.003 m) at the left edge of transect 13 from that elevation at transect 1 at each reach and dividing the result by the reach length. Along each transect the following characteristics were determined: aquatic habitat type (pool, riffle, or run), emergent and submergent vegetation type and coverage, riparian grazing (low, moderate, high, or very high), water depth, and stream width at the water surface and bankfull points. Bottom substrate was classified by size at eight wetted points along each transect (Wolman 1954, Harrelson et al. 1994). At the transect in each reach that appeared to have the most uniform cross-sectional structure and flow, we measured transparency with a secchi disk, water quality (dissolved oxygen, pH, total dissolved solids, and conductivity) with hand-held meters, and velocity with a top-setting wading rod and a Marsh-McBirney flow meter. Velocity was measured at 0.6 of stream depth where depths were less than 0.75 m. At depths greater than 0.75 m, flow was the average of velocities at 0.2 and 0.8 of stream depth. Discharge was calculated by using AQUAPAK according to Gordon et al. (1992).