Analyzing Correlations between Stream Benthic Macroinvertebrate Community Structure, Water Quality, and Watershed Land Use in the Continental United States

摘要

The objective of this project was to explore national-scale data relationships between land use, primarily variables pertaining to agricultural production, and in-stream biotic conditions in the mainland U.S. Using the Causal Analysis/Diagnosis Decision Information System (CADDIS) and data from the USGS National Water Quality Assessment Program (NAWQA), benthic macroinvertebrate observed versus predicted (0/E) index values were analyzed against a number of land use and habitat variables in 115 sitesacross the U.S. Constituent loads, or the mass amounts of a chemical constituent entering the waterway per year, were estimated for each site using the USGS LOAD-EST program and were evaluated with respect to macroinvertebrate 0/E and land use values. Relationships between variables, defined causally through CADDIS, were analyzed using simple linear regression to avoid analytical bias. Regression analyses indicated that forest and urban land cover were significantly correlated to macroinvertebrate 0/E values (p = 0.0012 and p < 0.001). Consistent and relatively strong positive correlations were found between land use and nutrient loading (p < 0.001 for all constituents). However, contrary to expectations, no correlation was observed between agricultural land use and O/E values, while negative correlations were observed between all nutrient loading variables and O/E values. These national-scale data support the complex process relationship between agricultural land use and benthic macroinvertebrate condition.