Investigation of water quality and aquatic-community structure in Village and Valley Creeks, City of Birmingham, Jefferson County, Alabama, 2000-01

The U.S. Geological Survey conducted a 16-month investigation of water quality, aquatic-community structure, bed sediment, and fish tissue in Village and Valley Creeks, two urban streams that drain areas of highly intensive residential, commercial, and industrial land use in Birmingham, Alabama. Water-quality data were collected between February 2000 and March 2001 at four sites on Village Creek, three sites on Valley Creek, and at two reference sites near Birmingham—Fivemile Creek and Little Cahaba River, both of which drain less-urbanized areas. Stream samples were analyzed for major ions, nutrients, fecal bacteria, trace and major elements, pesticides, and selected organic constituents. Bed-sediment and fish-tissue samples were analyzed for trace and major elements, pesticides, polychlorinated biphenyls, and additional organic compounds. Aquatic-community structure was evaluated by conducting one survey of the fish community and in-stream habitat and two surveys of the benthic-invertebrate community. Bed-sediment and fish-tissue samples, benthic-invertebrates, and habitat data were collected between June 2000 and October 2000 at six of the nine water-quality sites; fish communities were evaluated in April and May 2001 at the six sites where habitat and benthic-invertebrate data were collected. The occurrence and distribution of chemical constituents in the water column and bed sediment provided an initial assessment of water quality in the streams. The structure of the aquatic communities, the physical condition of the fish, and the chemical analyses of fish tissue provided an indication of the cumulative effects of water quality on the aquatic biota.

Water chemistry was similar at all sites, characterized by strong calcium-bicarbonate component and magnesium components. Median concentrations of total nitrogen and total phosphorus were highest at the headwaters of Valley Creek and lowest at the reference site on Fivemile Creek. In Village Creek, median concentrations of nitrite and ammonia increased in a downstream direction. In Valley Creek, median concentrations of nitrate, nitrite, ammonia, organic nitrogen, suspended phosphorus, and orthophosphate decreased in a downstream direction. Median concentrations of Escherichia coli and fecal coliform bacteria were highest at the most upstream site of Valley Creek and lowest at the reference site on Fivemile Creek. Concentrations of enterococci exceeded the U.S. Environmental Protection Agency criterion in 80 percent of the samples; concentrations of Escherichia coli exceeded the criterion in 56 percent of the samples. Concentrations of bacteria at the downstream sites on Village and Valley Creeks were elevated during high flow rather than low flow, indicating the presence of nonpoint sources. Surface-water samples were analyzed for chemical compounds that are commonly found in wastewater and urban runoff. The median number of wastewater indicators was highest at the most upstream site on Valley Creek and lowest at the reference site on Fivemile Creek. Concentrations of total recoverable cadmium, copper, lead, and zinc in surface water exceeded acute and chronic aquatic life criteria in up to 24 percent of the samples that were analyzed for trace and major elements. High concentrations of trace and major elements in the water column were detected most frequently during high flow, indicating the presence of nonpoint sources. Of the 24 pesticides detected in surface water, 17 were herbicides and 7 were insecticides. Atrazine, simazine, and prometon were the most commonly detected herbicides; diazinon, chlorpyrifos, and carbaryl were the most commonly detected insecticides. Concentrations of atrazine, carbaryl, chlorpyrifos, diazinon, and malathion periodically exceeded criteria for the protection of aquatic life.

Trace-element priority pollutants, pesticides, and other organic compounds were detected in higher concentrations in bed sediment at the Village and Valley Creek sites than at the reference site on Fivemile Creek. Bed-sediment concentrations of chromium, copper, lead, mercury, and silver were highest at the most upstream site on Valley Creek; and concentrations of cadmium, nickel, selenium, and zinc were highest at the second downstream site on Village Creek. Bed-sediment concentrations of arsenic, cadmium, chromium, copper, lead, mercury, nickel, selenium, and zinc from the Village and Valley Creek sites exceeded median concentrations observed nationwide. Concentrations of cadmium, selenium, and zinc were highest in fish-liver tissue samples collected from the second downstream site on Village Creek — concentrations of copper and mercury in fish-liver samples were highest at the most downstream site on Village Creek.

The highest total concentration of organic compounds detected in bed-sediment samples occurred at the most upstream site on Valley Creek and the lowest total concentration occurred at Fivemile Creek. In Village Creek, concentrations of 75 percent of the detected organic compounds increased in a downstream direction; in Valley Creek, concentrations of about 70 percent of the detected organic compounds decreased in a downstream direction. Concentrations of 10 organic compounds in bed-sediment samples, including chlordane and p,p'-DDE, exceeded levels considered harmful to aquatic organisms at sites on Village and Valley Creeks. Concentrations of dieldrin, chlordane, and polychlorinated biphenyls in fish-tissue samples exceeded National Academy of Science/National Academy of Engineering guidelines for the protection of fish-eating wildlife.

Fish and benthic-invertebrate community structure differed between Village and Valley Creeks and the reference streams. Multiple lines of evidence, including the richness and density of benthic invertebrates as well as fish-community structure, indicate that the aquatic community in Village Creek is similar to that of Valley Creek, but that the integrity of the aquatic communities in both creeks is poor in comparison to that observed at the two reference sites.

The abundance of mayflies and the number of EPT (Ephemeroptera, Plecoptera, Trichoptera) taxa (two well-known indicators of good water quality) were negatively correlated with industrial land use. The abundance of midges (an indicator of poor water quality) was positively correlated with industrial land use — and midge density was positively correlated with commercial land use, providing additional evidence that these streams have been negatively affected by urbanization in the basins. The percentage of mosquitofishes (a tolerant species) was positively correlated with commercial land use. In contrast, the numbers of fish species, fish families, and the percentage of sunfishes (intolerant species) were positively correlated with forested land use, indicating that the more diverse fish communities were found in basins with a higher percentage of forested land. The concentrations of 12 water-quality constituents (including several nitrogen species, chloride, copper, and molybdenum, and the detection frequency of wastewater indicators) and 18 organic compounds detected in bed sediment were positively correlated with industrial land use. Mercury and molybdenum concentrations detected in fish-liver tissue also were positively correlated with industrial land use. Bed-sediment and water-quality constituents that were found to have significant correlations with land use often were found to be correlated with many biological indicators, further supporting the link between increased urbanization and changes in aquatic-community structure.

The water quality and aquatic-community structure in Village and Valley Creeks are degraded in comparison to streams flowing through less-urbanized areas. Low community richness and increased density of certain species within the fish and benthic-invertebrate communities indicate that degradation has occurred during an extended period of time. Decreased diversity in the aquatic communities and elevated concentrations of trace elements and organic contaminants in the water column, bed sediment, and fish tissues at Village and Valley Creeks are indicative of the effects of urbanization. The degree of degradation may be related to point and nonpoint sources of contamination originating within the basins. Industrial land use, in particular, was significantly correlated to elevated contaminant levels in the water column, in bed sediment, in fish tissue, and to the declining health of the benthic-invertebrate communities. The results of the 16-month study have long-range watershed management implications, demonstrating the association of urban development and stream degradation. These data can serve as a baseline from which to determine the effectiveness of stream-restoration programs.