Temperature-Caused Fish Kill in a Flowing Great Plains River

On 9 July 2002, while sampling a flowing segment of the North Fork Ninnescah River 10 km upstream from Cheney Reservoir, Reno County, Kansas, we observed and collected dead and dying specimens of 5 fish species. We attribute the fish kill to high water temperature, which reached 38.0 C on this day. RESUMEN El 9 de julio del 2002, mientras muestreábamos un segmento con corriente del rı́o North Fork Ninnescah, 10 km rı́o arriba de la presa Cheney, en el condado de Reno en Kansas, observamos y recolectamos cinco especies de peces moribundos y muertos. Atribuimos la muerte de los peces a la alta temperatura del agua registrada ese dı́a, la cual alcanzó los 38.0 C. Effects of temperature are generally sublethal, but in extreme cases temperature and associated stressors can directly cause fish mortality (Beitinger et al., 2000). The most commonly reported instances of mortality attributable to high water temperatures are among fishes that become isolated in streambed pools for extended periods (Tramer, 1977; Matthews et al., 1982; Mundahl, 1990) or in receding lakes (Bailey, 1955). Limited volume of these isolated waters makes them susceptible to rapid changes in physical and chemical conditions. Mortality of fishes attributable to high water temperature is believed to be common, although it is infrequently observed in the field (Bailey, 1955). We are aware of only 2 published reports (Huntsman, 1942, 1946) of temperature-induced mortality among fishes in lotic habitats, both of which describe mortality in coldwater streams. The purpose of this paper is to describe observations of a fish kill in a flowing warmwater stream that we attribute to high water temperature. Our account is necessarily observational because the transience of the event, encountered in the field, precluded experimental manipulation. On 8 and 9 July 2002, we sampled the North Fork Ninnescah River (37 50 41 N, 97 56 09 W) approximately 10 km upstream from Cheney Reservoir, Reno County, Kansas. Our purpose was to document the composition of the fish assemblage in relation to habitat variables. We sampled this site for approximately one hour, beginning at 1830 h, using 2 seines (1.8 3.4 m and 1.8 9.1 m, with 5mm mesh) and making an effort to sample all available habitat types. Captured fish were preserved in 10% buffered formalin. Water quality parameters (temperature, dissolved oxygen, conductivity, and turbidity), stream width, stream depth, and current velocity were recorded in each habitat sampled. In addition to our measurements, hourly water temperature and stream discharge measurements were obtained from the U.S. Geological Survey gaging station (#07144780) located at the site. We collected 518 fish representing 9 families and 14 species on 8 July 2002 (Table 1). Red shiner (Cyprinella lutrensis) and sand shiner (Notropis stramineus) were the most common species, collectively representing 80% of the total catch. Northern plains killifish (Fundulus kansae) and white perch (Morone americana) were the third and fourth most abundant species, respectively. These 2 species combined represented an additional 12% of collected specimens. All other species individually represented fewer than 2% of collected specimens. Physical and chemical conditions (mean SD) recorded at the time of sampling were: water temperature 35.0 0.17 C, dissolved oxygen concentration 6.93 0.17 mg/L, conductivity 1,115 104 S/cm, and turbidity 69.2 9.4 NTU. Mean current velocity was 0.37 0.08 m/s, mean stream depth was 20 7.8 cm, and stream width was 34.7 m. The Southwestern Naturalist vol. 51, no. 3 (Sept. 2006), pp. 397-401. 398 vol. 51, no. 3 The Southwestern Naturalist TABLE 1—Fish collected on 8 and 9 July 2002 from the North Fork Ninnescah River upstream from Cheney Reservoir, Reno County, Kansas. Fish collected on 8 July were alive. Fish collected on 9 July were dead or moribund. Species Number collected