Northern Prairie Wildlife Research Center
Results similar to those obtained for the Red River basin were obtained by Livingstone et al. (1982) for African rivers. Discharge explained 74% of the variation in fish species richness, while stream length and watershed area were not significant in multiple regression models. These results were explained by the fact that over the study area, mean annual precipitation varied greatly (Livingstone et al. 1982). Streams of equal watershed area and/or length were receiving different amounts of annual precipitation. Since the discharge of a river is a function of rainfall per unit area, it was the best predictor of species richness in the African rivers. This rationale may also be applicable to the Red River basin. Because of an east-west gradient in annual rainfall, drainage basins of comparable size and streams of comparable length do not receive equal precipitation and therefore do not have comparable rates of discharge. An example can be noticed when comparing the Buffalo River (eastern basin) with the Goose River (western basin). The watershed area of the Buffalo River was 1189 mi2 and that of the Goose River was 1280 mi2 (Table 7). The stream length of the Buffalo River was 150 mi and that of the Goose River was 186 mi. However, the discharge of the Buffalo River was 132 cfs whereas the discharge of the Goose River was only 71 cfs. Similar comparisons can be made between eastern and western streams in the basin, such as the Tamarac or Middle Rivers compared to the Rush River, or as an extreme case, the Red Lake River compared to the Sheyenne River.