Landscape Influences on Stream Habitats and Biological Assemblages

Island Biogeography of Native Fish Faunas among Great Plains Drainage Basins: Basin Scale Features Influence Composition

Christopher W. Hoagstrom and Charles R. Berry, Jr.

doi: https://doi.org/10.47886/9781888569766.ch11

Abstract.—We reviewed native fish zoogeography in 22 major tributary basins of the Missouri River basin in the Great Plains geomorphic province and used island biogeographical approaches to study the influence of basin area and isolation on faunal composition. Basin area was correlated with elevation range and basin isolation was negatively correlated with annual freeze-free days. Ninety-six species were native to the tributary basins. Ninety-one were of southern (Gulf of Mexico drainage) origin. Fifty were found in four or fewer tributary basins and, except for three mountain species, were only found from the Cheyenne basin downstream. Twenty-five widespread species were either present among tributary basins during glaciation or colonized the region during recession of the continental glaciers. Sixty-six more restricted species presumably colonized more recently. Five species colonized from Pacific Ocean drainages via interdrainage connections in the Rocky Mountains. The hypothesis that variation between some closely related Great Plains fishes reflects the former presence of a prehistoric “Ancient Plains Drainage” is no longer tenable given recent geological findings, but a series of stream captures between the ancient Arkansas and Kansas basins could account for such variation. All analyses indicated that native fish faunal composition among tributary basins was heavily influenced by factors related to basin area and isolation. A presence–absence matrix of fishes by tributary basin had very high nestedness, whether ordered by basin area or basin isolation. Cluster analysis of Wilcoxon two-sample tests of individual species distributions revealed seven species groups with distinct distribution patterns. The three largest groups were most prevalent in less isolated (southern) tributary basins. A nonmetric multidimensional scaling analysis (NMDS) based on Sørensen’s index of similarity indicated that two axes (both correlated with tributary basin isolation, one correlated with tributary basin area) accounted for 95% of variance between distance in the ordination space and distance in the original n-dimensional space. A cluster analysis of NMDS scores identified five tributary basin groups. The five southernmost basins (Kansas to White) composed one group, and the eight basins to the north (Bad to Little Missouri) composed another. The nine northernmost basins were split into three groups, one including small basins isolated from the Rocky Mountains, another including large basins with Rocky Mountain headwaters, and the last including two basins that were mostly within the Rocky Mountains. The influence of tributary basin area on faunal composition was presumably due to increased chance of colonization, higher habitat stability, and higher habitat diversity in larger tributary basins. The influence of tributary basin isolation was presumably due to higher colonization rates and more equitable climate in southern tributary basins. Fish faunas of the Missouri River basin in the Great Plains have experienced cyclical geomorphic and climatic instability for roughly 2.8 million years and were assembled like island faunas by variable colonization and extinction rates mediated by tributary basin area and isolation. This contrasts with the highly diverse freshwater fish faunas of the Central Highlands that have differentiated through speciation within regions that have been relatively stable geomorphically and climatically for more than 38 million years.