2007 GSA Denver Annual Meeting (28–31 October 2007)

Paper No. 8
Presentation Time: 3:20 PM


ZELT, Ronald B., N/a, U.S. Geological Survey, 5231 South 19th Street, Lincoln, NE 68512, rbzelt@usgs.gov

Riparian characteristics were sampled once in 2003-04 for an ecological and nutrient survey of 141 small perennial, agricultural streams in 5 study areas across the U.S. Riparian variables were selected to include stream-habitat and land-cover indicators from 22 categories. Principal component (PC) analysis of 40 variables indicated 4 nontrivial components to retain, including 60 percent of the variance. Cluster analysis identified 6 groups of distinctly different sampling sites. Rank correlations of habitat variables and response variables were analyzed within site clusters, study areas, and overall.

PCs 1 and 2 contained 43.1 percent of the variance in the habitat data. PC1 had intense loadings related to channel shading, whereas PC2 had intense loadings related to riparian woodland extent. Channel canopy closure was highly correlated with PC1 scores and was considered the best measure of channel shading.

All variables related to riparian cropland extent were significantly correlated (r > 0.35) with total phosphorus concentration (TP), suggesting agricultural use of phosphorus on cultivated fields as a likely near-stream source. However, a weak relation between TP and cropland extent at the basin scale may reflect the combined effects of delivery ratio to channels, channel storage, and phosphorus-management practices on cropland. Stream segment-level riparian cropland extent was correlated more strongly with TP (r > 0.54) than was any individual stream-habitat variable. At all spatial scales examined, riparian wetland extent was negatively correlated with TP for western sites.

Riparian shading of streams was consistently an inverse correlate of aquatic macrophyte or macroalgae extent (AM). For all eastern sites combined, AM was correlated with open-canopy angle (CA) and with total nitrogen (TN) concentrations in stream samples; AM exhibited a positive relation to TN when CA was less than 40 degrees, but the relation was negative when CA was greater than 40 degrees.