Characterizing food web structure of a southern prairie river with conservation implications for four native stream fishes

Environmental variation is known to influence the quality, quantity, and availability of instream and terrestrial food resources for riverine consumers. We assessed the effects of spatiotemporal variation on aquatic food web structure and resource use among an understudied group of southern prairie river fishes of conservation need using stable isotope and fish gut content analysis. Because stable isotope analysis often requires lethal take, we also examined the use of fish fin as an alternative to muscle tissue for future nonlethal δ15N and δ13C analysis of plains minnow (Hybognathus placitus), prairie chub (Macrhybopsis australis), Red River shiner (Notropis bairdi), and Red River pupfish (Cyprinodon rubrofluviatilis). Food web structure, basal resource use, and fish diet composition varied across seasons and longitudinal gradients. Although δ13C values of basal instream and riparian producers frequently overlapped, the δ13C values of consumers were most similar to δ13C values of instream detritus or algae across seasons. While δ15N values varied according to species across longitudinal river gradients, δ13C values of fishes were generally negatively related to river kilometer, mimicking trends of longitudinal δ13C depletion in instream detritus. Enriched δ15N values and constricted isotopic and diet niche areas from spring to summer suggested more specialized feeding habits among prairie chub and Red River shiner, likely due to increased resource availability and prey selectivity. Fewer distinct food items in diets, downward shifts in δ15N values, and expanded diet niches during fall may suggest a shift to generalist feeding behavior for prairie chub and Red River shiner as invertebrate abundance decreased. By contrast, food resource use by plains minnow and Red River pupfish differed little between seasons. Although isotopic composition differed between tissues, fin and muscle isotope values were positively related for δ15N (ANCOVA: F1, 451 = 5312.09, P < 0.001) and δ13C (ANCOVA: F1, 451 = 7864.39, P < 0.001), suggesting that fin isotope values can be used to predict isotope values of muscle tissue. Species-specific regression models showed good linear fit, and minimum length thresholds suggest minimally invasive fin clipping is feasible for plains minnow > 70 mm, prairie chub and Red River shiner > 62 mm, and Red River pupfish > 48 mm. Our assessment of spatiotemporal changes in food web structure and fish feeding habits may provide a reference from which to compare how Red River consumers may be affected by habitat alteration and homogenization. Additionally, subsequent resource assessments may use fin tissue as a nonlethal alternative to muscle tissue to prevent future lethal take and limit impact on declining populations.