Abstract

Quantifying the effects of watershed improvement efforts is critical to agencies responsible for protecting water resources of the semiarid western United States. A large complex water quality data set of the Muddy Creek basin was subjected to univariate and multivariate statistical techniques to investigate whether livestock grazing and channel/floodplain best-management practices (BMPs) improved the water quality of the watershed. Following BMP implementation, flow regime analyses indicated lower hydrograph peak flows and reduced percents of flow duration intervals in the lower watershed. Hierarchical cluster analysis grouped sampling sites in the upper basin into homogeneous clusters based on similarity of biological indices and isolated distinct areas of degraded water quality. Discriminant analysis provided important data reduction by rendering seven key parameters for the spatial variation and four parameters for the temporal variation in the watershed, affording over 80% correct assignations in the spatial and temporal analyses. Canonical correlation analysis identified strong relationships among water quality, biological, and physical parameters. These data demonstrated positive trends in water quality, biological, and physical variables in the upper and lower basin despite the onset of extreme drought midway through the study period. Naturally erosive conditions along with anthropogenic impacts elevate the importance of BMP implementation and a long-term monitoring program. Both are needed to evaluate the effectiveness of improvement efforts and to help resource managers develop practical watershed improvement management strategies so that optimum conditions in Muddy Creek basin can be achieved.