USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2006–5101–D

Back to Table of Contents

Summary

This study examined how urbanization affects stream ecosystems. Objectives of the study included an examination of physical, chemical, and biological responses to urbanization of small streams in the Willamette River basin and surrounding area, Oregon and Washington.

Hydrologic variable findings indicate that as urban intensity indicators increased (urban intensity index (UII), road density, percentage of impervious surface and population density), so did the stream flashiness. For example, the response of stream flashiness metrics to urbanization was moderate (0.66 > rho ≥ 0.50) to strong (rho ≥ 0.66) among the sites examined in this study. In addition, as urban indicators increased, minimum water temperature (95^th percentile) also increased by a moderate response (0.56 ≥ rho ≥ 0.44). Other moderate responses to increased urbanization included an increase in average substrate embeddedness (riffles and runs), a decrease in average width-to-depth ratio (riffles and runs), and a decrease in the percentage of riffle habitat.

Physical data collected indicated that one of the strongest relations was detected between percentage of riffle habitat and summer dissolved oxygen concentrations (rho = 0.84). It was also discovered that there was a moderate to strong negative correlation of dissolved oxygen with minimum temperature, average substrate embeddedness, and mean habitat heterogeneity (measure of instream habitat diversity). There were other connections between physical and water chemistry, such as, streamflow flashiness metrics showed strong responses to select water-chemistry parameters, including dissolved organic carbon, sum of total pesticides, toxic equivalents (measure of water column contaminants obtained from semipermeable membrane devices), and sulfate concentration. Conversely, streamflow flashiness had a low correlation to nutrients, such as total nitrogen concentration (rho = 0.30).

Most water-chemistry metrics, including sum total insecticide, sum total pesticide, pesticide toxicity index, toxic equivalents (TEQ), and Pyrene index, responded moderately to strongly to each increased urban indicator metric (0.83 ≥ rho ≥ 0.56). For example, the sum of total insecticides correlated strongly to road density, population density, and the urban intensity index (rho = 0.73, 0.70, and 0.69, respectively). However, insecticides were slightly less correlated to percentage of impervious surface and urban plus agricultural land use (rho = 0.66 and 0.63, respectively). Sum of total pesticides was strongly related to each urban metric (rho ≥ 0.66), except percentage of impervious surface where it was moderate (rho = 0.65). The semipermeable membrane device assays correlated positively to urban indictors as well, with toxic equivalents strongly associated to all urban indicator metrics (0.81 ≥ rho ≥ 0.73). The Pyrene index increased moderately to strongly with the increasing urban metrics (0.67 ≥ rho ≥ 0.51). Total nitrogen and total phosphorus both correlated strongly to the urban intensity index (rho = 0.79 and 0.71, respectively). Total nitrogen also correlated positively with the percentage of urban plus agricultural land use. Other water-chemistry metrics with strong correlations that increased with increasing urbanization were specific conductance and dissolved sulfate, whereas dissolved organic carbon, alkalinity, chloride, and summer dissolved oxygen were moderately correlated.

Algal assemblages responded to the nutrient and organic enrichment effects of urbanization (and agriculture at some sites), which included the development of nuisance green algae growths in some streams. For example, blue-green and red algae dominated the relative density in all but one stream, Deep Creek. High-nutrient indicator diatoms and other algal assemblages that tolerate moderate to high degrees of organic enrichment were positively correlated with the urban intensity index (rho = 0.52).

Benthic algal chlorophyll-a was highly variable along the urban gradient, but the ash-free dry mass (measure of algal biomass used to indicate the degree of organic enrichment) was positively correlated with the urban intensity index (rho = 0.56). The ash-free dry mass was positively correlated with dissolved organic carbon concentrations, while dissolved oxygen was negatively correlated with dissolved organic carbon concentrations. These findings suggests that nutrient enrichment caused by urbanization increases the organic status of streams by stimulating algal growth and increasing ash-free dry mass. The organic matter eventually decomposes, which leads to the development of tolerant diatoms and heterotrophic algal taxa that make use of energy in the organic compounds. The use of such compounds by bacteria and other heterotrophic organisms consumes dissolved oxygen during respiration into carbon dioxide. Signs that this may be occurring in the more highly urbanized streams include the decrease in the relative abundance of diatom taxa requiring continuously high levels of dissolved oxygen. These sensitive taxa were most abundant at sites with a urban intensity index less than 25, and less abundant at sites higher on the urban gradient. In addition to nutrients and other effects of eutrophication such as organic enrichment, other environmental factors that were determined to be important in shaping the diatom assemblages included various measures of disturbance, such as streamflow flashiness, channel scour, and grazing benthic macroinvertebrates.

Benthic macroinvertebrate assemblages also showed a strong response to urbanization. From the most urbanized sites (urban intensity index greater than 70) to the least urbanized (urban intensity index less than 10), there was an average decrease of 12 Ephemeroptera, Plecoptera and Trichoptera taxa detected. In addition, there was a large increase in percentage of noninsects (31 percent) and percentage of abundance of tolerant taxa (48 percent) between the range of urbanization. At the sites across the full urban gradient, many macroinvertebrate metrics had strong correlations to urban indicators and water-quality variables. For example, Ephemeroptera, Plecoptera and Trichoptera taxa richness (rho = -0.75) and the nMDS ordination axis (rho = -0.83) were highly correlated to the urban intensity index. Macroinvertebrate metrics also had very strong correlations with total pesticide and insecticide concentrations, the pesticide toxicity index, contaminant measures from the semipermeable membrane device samples (TEQ and Pyrene Index), and summer dissolved oxygen. Although all four hydrologic variability and flashiness measures were moderately correlated to macroinvertebrate metrics, they represent less variation than any of the above water-quality variables. Therefore, although flashiness probably was an important issue in the urban areas of the Willamette River basin and surrounding areas, water chemistry issues including contaminants still are a dominant disturbance to macroinvertebrate assemblages followed by flashiness and sedimentation-substrate disturbance or habitat quality.

Fish assemblages showed strong correlation to urbanization, as well. For example, on average, 4 percent salmonids (sensitive) and 8 percent nonnative fish (tolerant) were found at sites with an urban intensity index greater than 25. Conversely, 17 percent salmonids and less than 1 percent nonnative fish were found at sites with an urban intensity index less than 25. All fish metrics, including fish index, salmonid percentage, and ordination axis 1 scores had strong correlations to the urban indicators and water-quality variables and moderate to strong correlation to the hydrologic variability, habitat, and water-temperature measures. Percentage of salmonids and macroinvertebrate Ephemeroptera, Plecoptera and Trichoptera taxa richness indicated a possible threshold response to urbanization at an urban intensity index of 25, which was equivalent to impervious surfaces of about 5 percent. However, due to the added agricultural land use element at sites with urban intensity index values between 25 and 60, this threshold may not be due to urbanization solely, but a combination of urban and agricultural land uses. The effects of agricultural and urban land use could not be distinguished from each other, yet combined provide a good assessment of overall watershed health.

Back to Table of Contents