Development of a Rapid Bioassessment Protocol for Sampling Fish in Large Prairie Rivers

We used seining and hoop netting to collect fish at 15 sites in five large prairie rivers in Oklahoma to (1) determine the amount of effort needed to detect the maximum number of species at a sample site and (2) examine the selectivity of fish species detected by the two gear types. Analysis of the similarities of the fish collected in six different habitat types identified two distinct habitat types based on fish species composition: shallow–backwater (SBW) habitat (depth ≤ 0.75 m) and deep–nonwadeable (DNW) habitat (depth > 0.75 m). We estimated that between 6 and 10 (mean = 8) SBW habitats and between 1 and 6 (mean = 4) DNW habitats at each sample site were needed to obtain maximum species richness during a sampling event. The sampling distance needed to encounter the minimum number of habitats ranged from 400 to 1,600 m and averaged 887 m. Gear evaluation showed that seining captured more species per unit effort than hoop netting (3.6 and 1.4, respectively); however, hoop netting captured significantly larger fish (527 mm; P < 0.001) than seining (42 mm). Based on these collections, we present recommendations for sampling fish assemblages in large prairie rivers in the southern Great Plains to aid in the rapid bioassessment and monitoring of fish assemblages in large prairie rivers.