Stream health shows indirect and threshold responses to implementation of agricultural conservation practices

Many agricultural conservation practices (hereafter conservation practices) are intended to protect streams from the deleterious effects of agriculture, but desired effects of their implementation are not always realized. Understanding the pathways through which conservation practices influence stream health can guide future implementation. Our goal was to quantify how conservation practices influence stream health in pastureland watersheds. We asked 4 research questions: R1) What are the effects of land use and conservation practices on instream conditions? R2) What are the effects of instream conditions on biota? R3) What are the relative effects of land use, conservation practices, and instream conditions on biota? R4) How do conservation practice density (no./ha) and agricultural land use interact to influence stream health? We collected data on water quality (total suspended solids, total N, total P, and Escherichia coli bacteria), instream habitat (substrate embeddedness and bank stability), and macroinvertebrate communities from 31 sites in the upper Tennessee River basin of southwest Virginia, USA. The sites collectively represented spatial analytical gradients of agricultural land use and density of conservation practices, with the dominant practices including prescribed grazing, brush management, and fencing. We found that conservation practice implementation was indirectly related to biotic assemblages through changes in water quality and habitat, but practice density in the study area was not associated with instream conditions that supported biotic communities representative of reference conditions. Further improving instream habitat conditions may require innovative conservation practices, targeted placement of practices in critical source areas of pollutants, or higher densities of practices.