The southeastern United States supports a high diversity of stream and riverine fishes, yet little river mileage and few species have been afforded protected. Kentucky’s Upper Green River Basin is the subject of a U.S.D.A. Conservation Reserve Enhancement Program (CREP). The Kentucky CREP was initiated in 2001. A driving emphasis of the Kentucky CREP program is to reduce inputs of nonpoint source materials (e.g., sediment, nutrients) into both surface and subsurface waters through incentive-based cooperative agreements with landowners. Typical CREP landowner implementations involve conversion of agricultural lands, namely pasture/hay and row cropping activities, into prescribed plantings of either bottomland hardwoods or upland grasses. Despite the anticipated ecological benefits of CREP, the apriori biotic health of aquatic communities of the Upper Green River Basin had yet to be assessed objectively prior to 2001. During 2002 and 2003 both macroinvertebrate and fish assemblages were sampled extensively across a wide range of stream sizes and adjacent land-use practices in order to assess apriori relationships between biotic communities across a broad range of watershed- and reach-scale environmental parameters.
Results/Conclusions
A recently-published within-subbasin analysis assessed whole fish assemblage relationships to watershed- and reach-scale environmental variables and showed that (a) species distribution patterns were due to stream-size gradients and not by land-use patterns or indicators of anthropogenic disturbance, and (b) species were associated either with small, upland segments or only within the largest, deeper segments. Two additional analyses have been performed. First, data was pooled across all subbasins yet divided sites into two distinct watershed size groups to assess if there were size-driven biotic responses to environmental variables. Even amongst the smallest stream segments an agricultural influence was not revealed. A series of canonical correspondence analyses (CCA) and indicator species analyses revealed mainly similar stream-size related relationships across the two data groups. Second, data were reduced solely to darter species (Etheostoma and Percina) to limit noise induced by the large number of species obtained. Similar to previous analyses, CCA and multiple linear regression models between abundance of significant indicator species analyses and environmental variables similarly revealed only stream size patterns. Overall, the diverse fauna across a broad suite of streams and the lack of clear relationships between species distribution patterns and agricultural indicators suggests the fish fauna within Kentucky’s Upper Green River Basin is reasonably intact and that landscape-level shifts away from row cropping and pasture/hay activities may have more a direct effect on terrestrial communities.
