Influence of Environmental Factors on Biotic Responses to Nutrient Enrichment in Agricultural Streams

Abstract

The influence of environmental factors on biotic responses to nutrients was examined in three diverse agricultural regions of the United States. Seventy wadeable sites were selected along an agricultural land use gradient while minimizing natural variation within each region. Nutrients, habitat, algae, macroinvertebrates, and macrophyte cover were sampled during a single summer low-flow period in 2006 or 2007. Continuous stream stage and water temperature were collected at each site for 30 days prior to sampling. Wide ranges of concentrations were found for total nitrogen (TN) (0.07-9.61 mg/l) and total phosphorus (TP) (<0.004-0.361 mg/l), but biotic responses including periphytic and sestonic chlorophyll a (RCHL and SCHL, respectively), and percent of stream bed with aquatic macrophyte (AQM) growth were not strongly related to concentrations of TN or TP. Pearson's coefficient of determination (R(2)) for nutrients and biotic measures across all sites ranged from 0.08 to 0.32 and generally were not higher within each region. The biotic measures (RCHL, SCHL, and AQM) were combined in an index to evaluate eutrophic status across sites that could have different biotic responses to nutrient enrichment. Stepwise multiple regression identified TN, percent canopy, median riffle depth, and daily percent change in stage as significant factors for the eutrophic index (R(2) = 0.50, p < 0.001). A TN threshold of 0.48 mg/l was identified where eutrophic index scores became less responsive to increasing TN concentrations, for all sites. Multiple plant growth indicators should be used when evaluating eutrophication, especially when streams contain an abundance of macrophytes.

FIGURE 1

Location of the Streams Sampled Within the Ozark Highlands (OZRK), Upper Mississippi (UMIS), and Upper Snake (USNK) Regions (gray-shaded areas) Sampled by U.S. Geological Survey, 2006-2007. See Appendix 1 for names and locations of sample sites.

FIGURE 2

Cumulative Distributions of (a) Total Nitrogen (TN) and (b) Total Phosphorus (TP) Across All Study Sites, n = 70. Dashed lines indicate nutrient (TN and TP) trophic boundaries suggested by Dodds et al. (1998). L, low enrichment; M, moderate enrichment; H, high enrichment.

FIGURE 3

The Relative Percentage of All Sites and Individual Regions Classified as Eutrophic (highly enriched) on the Basis of Total Nitrogen (TN > 1.5 mg/l), Total Phosphorus (TP > 0.75), Periphytic Chlorophyll a (RCHL > 70 mg/m²), Seston Chlorophyll a (SCHL > 30 ug/l) (Dodds et al., 1998), and Aquatic Macrophyte (AQM) Coverage of Stream Bottom >40% (Chambers et al., 1999; Suplee et al., 2009).

FIGURE 4

Relation of Periphytic Chlorophyll a (RCHL) to Macrophyte (AQM) Percent Coverage of Stream Bottom for All Study Sites (n = 70).

FIGURE 5

Bivariate Plots of the Biotic Response Variables Periphytic (a and b) Chlorophyll a (RCHL), (c and d) Sestonic Chlorophyll a (SCHL), and (e and f) Aquatic Macrophyte (AQM) Percent Coverage and Total Nitrogen (TN) and Total Phosphorus (TP) Concentrations. The lines indicate the best fit linear regression for streams in the OZRK (solid line), UMIS (dashed line), and USNK (dotted line). Regression equations of all sites combined and individual regions can be found in Table 4.

FIGURE 7

Eutrophic Index Scores as a Function of Total Nitrogen (TN) Concentrations for All Study Sites (n = 70). Piecewise regression line with identified breakpoint threshold for TN of 0.48 mg/l. Dashed lines are the 95% confidence interval for the regression line. The gray line represents a nonparemetric LOWESS (Locally Weighted Scatterplot Smoother) fit of the data (Helsel and Hirsch, 2002).

FIGURE 6

Bivariate Plots of Eutrophic Index Scores and (a) Total Nitrogen (TN) and (b) Total Phosphorus (TP) for All Study Sites. The lines indicate the best fit linear regression for all study sites. Dashed lines are the 95% confidence interval for the regression line.