Abstract

Large-scale patterns in fish assemblage structure and functional groups are influenced by alterations in streamflow regime. In this study, we defined an objective threshold for alteration for Oklahoma streams using a combination of the expected range of 27 flow indices and a discriminant analysis to predict flow regime group. We found that fish functional groups in reference flow conditions had species that were more intolerant to flow alterations and preferences for stream habitat and faster flowing water. In contrast, altered sites had more tolerant species that preferred lentic habitat and slower water velocity. Ordination graphs of the presence and functional groups of species revealed an underlying geographical pattern roughly conforming to ecoregions, although there was separation between reference and altered sites within the larger geographical framework.

Additionally, we found that reservoir construction and operation significantly altered fish assemblages in two different systems, Bird Creek in central Oklahoma and the Kiamichi River in southeastern Oklahoma. The Bird Creek flow regime shifted from a historically intermittent stream to one with stable perennial flows, and changes in fish assemblage structure covaried with changes in all five components of the flow regime. In contrast, the Kiamichi River flow regime did not change significantly for most flow components despite shifts in fish assemblage structure; however, most of the species associated with shifts in assemblage structure in the Kiamichi River system were characteristic of lentic environments and were likely related more to proximity of reservoirs in the drainage system than changes in flow. The spatial patterns in fish assemblage response to flow alteration, combined with different temporal responses of hydrology and fish assemblage structure at sites downstream of reservoirs, indicate that interactions between flow regime and aquatic biota vary depending on ecological setting. This supports the notion that regional variation in natural flow regimes could affect the development of flow recommendations.

First posted May 31, 2012