Correlations between land use and stream nitrate-nitrite concentrations in the Yahara River Watershed in south-central Wisconsin.

To better inform land management decisions, we explored relationships between land use data and stream nitrate-nitrite (NONO) concentration data in the Yahara River Watershed (YRW) in south-central Wisconsin, USA. Three metrics were used to evaluate the extent of different land uses in the watershed: (1) the area percentage of each land use in both the watershed and in a range of riparian zone widths, (2) the area factor, which refers to the ratio of the area of woodland, recreational, open and vacant subdivided land, or wetlands in the riparian zone (6.1-213.4 m widths) to agricultural areas in the rest of watershed, which indicates the buffering capacity of the riparian zone, and (3) the inverse-distance-weighted (IDW) area percentage with proximity to sub-watershed outlet and to stream, which characterizes spatial arrangement in the watershed by assigning a higher weight to patches closer to the outlet or stream and a lower weight to those farther away. We found significant, positive correlations between the extent of agricultural areas and stream NONO concentrations. NONO concentrations were highly correlated to area factor metrics for all riparian zone widths such that as area factor decreased, NONO concentrations increased. There was also a marked increase in NONO concentrations at a threshold of approximately 60% agricultural area with IDW proximity to stream. Wetland area percentage in the entire watershed and IDW wetland area percentage with proximity to stream were negatively correlated to stream NONO concentrations. Compared to the simple area percentage metric, area factor and IDW wetland area percentage with proximity to stream were better indicators of stream NONO concentrations. Results from this study indicate that, in addition to land use area percentage, spatial distributions of land uses should be considered when managing watersheds. This study also demonstrates the value of citizen-based sampling data and reveals opportunities to improve the utility of such data.