Nitrate-nitrogen, land use/land cover, and soil drainage associations at multiple spatial scales

Managing non-point-source pollution of water requires knowledge of land use/land cover (LULC) influences at altering watershed scales. To gain improved understanding of relationships among LULC, soil drainage, and dissolved nitrate-N dynamics within the Calapooia River Basin in western Oregon, we selected 44 watersheds ranging in size between 3 and 33 km(2) for monthly synoptic sampling of surface water quality between October 2003 and September 2004. Seasonal associations were examined between dissolved nitrate-N and proportion of woody vegetation or poorly drained soils at differing scales (10, 20, 30, 60, 90, 150, 300 m, and entire watershed), which we defined as influence zones (IZs), surrounding stream networks. Correlations between nitrate-N and proportion woody vegetation or poorly drained soil at each IZ were analyzed for differences using the Hotelling-Williams test. We observed negative correlations (r = -0.81 to -0.94) between nitrate-N and proportion of woody vegetation during winter and spring. Poorly drained soils had positive correlations (r = 0.63-0.87) with nitrate-N. Altering the scale of analysis significantly changed correlations between nitrate-N and woody vegetation, with IZs <150 m being stronger than the watershed scale during winter. However, absolute differences in correlation values were small, indicating minimal ecological consequence for significant differences among scales. In contrast, nitrate-N correlations with poorly drained soil were stronger at the watershed scale than the 10- through 90-m IZs during winter and spring, and absolute differences were sufficient to suggest that scale is ecologically important when determining associations between dissolved nitrate-N and poorly drained soils.