Coliform densities in urban waters of West Texas.

Introduction

Urban impoundments commonly receive nonpoint-source pollution of many types and from a variety of sources (e.g., lawn and garden chemicals, pet and wildlife feces, petrochemicals from parking lots and streets). Storms and other surface water runoff events may deliver large amounts of contaminants into urban waters, significantly affecting water quality in short time intervals (1). Moreover, inhabitants of urban waters, such as fish and waterfowl, can be vectors of human pathogens such as Aeromonas sp., Salmonella sp., and Escherichia coli (2,3). For these reasons, the quality of surface waters in urban settings is of considerable interest to the general public.

In the southern high plains of west Texas, shallow ephemeral ponds, or playas, fill with seasonal rainfall and provide habitat for thousands of migratory waterfowl (4,5). Urban playas in this region often are deepened and incorporated into stormwater management plans as catchments for surface water runoff. Parks or open areas surround many of these playas and are popular sites for recreational activities, including sport fishing and wind surfing. These activities bring human beings into contact with playa water. Initial observations suggest that playas in the city of Lubbock, Texas, have coliform counts that greatly exceed U.S. Environmental Protection Agency (U.S. EPA) standards for human contact, which may pose risks to human health (6). The authors are aware of no studies of coliform densities in urban playas. For this reason, E. coli and total-coliform densities were surveyed in playas of the city of Lubbock and were compared with watershed and water chemistry characteristics in an effort to identify factors that may be correlated to bacterial populations.

Methods

The authors investigated 20 playa lakes located within the city of Lubbock, Texas (101 [degrees] 52[minutes] N, 33 [degrees] 35[minutes] W), that remain permanently inundated. Ten playas were sampled on each of two consecutive days, May 25 and May 26, 1995. Water characteristics on these dates should be representative of general playa features, although the authors are aware of no previous studies of seasonal patterns in water quality attributes of urban playas. Five-hundred-milliliter (mL) grab samples of water were collected from three locations in each playa; one from each of the prevailing upwind (southwest) and downwind (northeast) sides of the lake, and another from a randomly selected crosswind side (northwest or southeast). Samples were obtained 20 cm below the water surface within 2 meters of shore, placed in a cooler, and returned to the laboratory for analyses. This approach is consistent with standard sampling protocols suggested by the American Public Health Association (APHA) for water quality investigations (7).

Water samples were assayed for concentrations of ammonia, nitrate, orthophosphate, alkalinity, total acidity, and total chlorine using a surface water quality test kit (from the Hach Company in Loveland, Colorado). A Hach pH meter was used to measure pH. Total coliform and E. coli were isolated by filtering 1 mL of undiluted sample through a Whatman 0.45-micrometer ([[micro]meter]) cellulose nitrate filter, placing the filter onto a sterile pad containing m-ColiBlue[TM] selective media, and incubating at 30 [degrees] to 33 [degrees] C for 24 hours (8). After incubation, colony-forming units (CFU) were counted for each bacterial type. E. coli form blue colonies on this medium, and other coliforms form red colonies. Although colony densities of 30 to 300 CFU per plate are considered ideal for visual quantification, the extremely high densities of bacteria in these samples necessitated enumeration with a dissection microscope (at 40X magnification). Even so, counts were truncated at 600 CFU per plate because individual colonies were impossible to distinguish at higher densities.

Landscape features associated with each playa were evaluated according to aerial photographs: total watershed area, total surface area of playa, and total area of adjacent park or open land. Basins were classified as primary basins if they received runoff water primarily from the surrounding watershed or as secondary basins if they also received substantial overflow from other playas.

Results

Observations revealed very high coliform populations in most playas, with total-coliform densities often exceeding 600 CFU/mL and E. coli densities ranging between 10 and 60 CFU/mL [ILLUSTRATION FOR FIGURE 1 OMITTED]. Pearson Product-Moment correlations were calculated for E. coli densities, water quality parameters, and watershed characteristics (9). Although significant relationships were found to exist between some water quality parameters and watershed characteristics, none of the correlation coefficients for bacteria were significant [ILLUSTRATION FOR FIGURE 2 OMITTED]. A stepwise multiple linear regression, using E. coli densities as the dependent variable and water quality and watershed characteristics as independent variables, failed to detect any parameter that could explain a significant component of the variation in E. coli (9).

Total-coliform data often were right-truncated because of extremely high counts (see above), so Spearman Rank Correlation was used to evaluate relationships between the ranked total coliform densities and the ranks of the watershed and physicochemical parameters [ILLUSTRATION FOR FIGURE 2 OMITTED] (9). The relationship between total coliform and watershed size as significant (p = .021), and negative, indicating that coliform concentrations were lower in the playas that have larger watersheds. This result suggests a dilution effect for basins that receive greater surface water inputs from larger drainage areas.

TABLE 1

Average Water Quality Characteristics (mg/L) from
Stormwater Runoff and Selected Playas in Residential and
Commercial Areas of Lubbock, Texas

Characteristic                 Stormwater     Playas

Chemical oxygen demand          268.97       67.10
Biological oxygen demand         51.98        8.06
Total Kjeldahl nitrogen           3.01        3.11
Nitrate-nitrite                   0.69        0.22
Orthophosphate                    0.18        0.27
Total phosphorus                  0.33        0.76

Source: T.E. Ennis, "City of Lubbock Stormwater NPDES Permit Data
Analysis" (11).

High densities of total coliforms appeared to be associated with primary basins [ILLUSTRATION FOR FIGURE 1 OMITTED]. For this reason, a Mann-Whitney U test was used to evaluate the relationship between ranked total coliform densities and basin rank (10). The test found that the distributions of total-coliform populations differed significantly (p = .019) between primary and secondary basins. This result suggests that total-coliform populations may be diluted by overflow from adjacent basins.

Discussion

Communities located on the Texas high plains face a dilemma in managing surface water runoff because of drainage restrictions imposed by the lack of topographic relief. No rivers or streams exist to receive runoff. Although total annual rainfall in this region is relatively low, much of that total occurs during brief, localized, summer convection storms. Thus, local flooding from stormwater is a potential problem. In the city of Lubbock, a network of surface drainages along streets and subsurface drainages via culverts link urban playa basins that have been enlarged to receive and retain surface and stormwater runoff. This approach to runoff control can, however, lead to other water management problems, such as nonpoint-source inputs of biological and chemical contaminants from surrounding watersheds. Current research in Lubbock is focusing on the quality of stormwater runoff entering playas (Table 1). The runoff has been reported to contain total and suspended solids and to show chemical oxygen demand comparable to that of raw sewage (11-13). The authors are aware, however, of no other integrated study of water chemistry and bacterial densities for recipient playas.

U.S. EPA recommends that E. coli or enterococci be used as indicators of health hazards for recreational waters, largely because densities of E. coli and enterococci have been correlated to gastrointestinal illnesses for swimmers in both fresh and marine waters (14-16). The Texas Environmental Protection Agency recommends a fecal-coliform mean of no more than 2 CFU/mL for primary- or secondary-contact waters (6). Peak densities should not exceed 4 CFU/mL in more than 10 percent of the samples collected in a 30-day period. Similarly, mean E. coli densities should not exceed 1.26 CFU/mL for primary contact, nor exceed peak values of 5.76 CFU/mL in 10 percent of the samples collected over a 30-day period for secondary contact. In the present study of Lubbock playas, both total-coliform and E. coli densities were 10 to 100 times greater than recommended or peak levels. Measurements were not repeated over time, so the temporal component of these standards was not addressed by this study. Bacterial inputs to playas probably occur via runoff, however, and because no runoff had occurred within several months prior to this study, it is unlikely that observed densities had been elevated by recent inputs. Therefore, densities observed in this study probably represent conservative estimates of E. coli and coliform densities persisting in Lubbock playas over time.

A potential source of error in estimates of coliform densities was the incubator temperature of 30 [degrees] to 33 [degrees] C, used instead of the recommended temperature of 35 [degrees] C, which may allow other bacteria to grow on m-ColiBlue[TM] selective media (8). Subsequent tests of playa water samples incubated at the recommended temperature revealed even higher densities than reported in this study, indicating that bacterial counts presented in this paper may be conservative. With such high densities of coliform bacteria, playas may present hazards to human health. In fact, opportunistic skin infections have occurred among people coming in contact with playa waters in Lubbock, although relevant medical records have not been compiled (17).

While densities of coliforms in Lubbock playas may pose a threat to public health, their origin is uncertain. Feachem reported that ratios of E. coli to total coliforms of less than 4.0 probably derive from nonhuman sources (18). Values for nontruncated data in the present study ranged from 0.04 to 0.89 and averaged 0.33 (for 12 playas), which implied a source other than human. Both chemical and biological contaminants in urban environments can be traced to many sources, such as waterfowl and pet feces, chemical applications to lawns and gardens, vehicle emissions, and erosion. The sources vary with surrounding land use and the extent of impervious surface areas (19,20). Playas in the region provide critical habitat for more than 115 species of birds, including more than two million waterfowl that use the central migratory flyway (4,21,22). Because urban playas have been modified to retain water throughout the year, waterfowl that feed in nearby grain fields and rural playas gather at night on playas within the city of Lubbock. Levesque et al. reported high densities of coliforms in waters immediately following the presence of feeding seagulls (3). Densities declined rapidly (within hours) after dispersal of the birds, however, probably as a result of dilution and loss of viability. It seems likely that migratory birds, pets, and resident wildlife provide inputs of bacteria to playas in Lubbock throughout the year, although reasons for such high densities of coliforms are not apparent. The lack of significant correlation between E. coli densities and any water quality parameter or landscape feature allows little insight into factors that may be controlling levels of these bacteria [ILLUSTRATION FOR FIGURE 2 OMITTED]. Similarly, although analyses suggested that total-coliform densities were diluted by overflow from other playas and greater runoff in larger watersheds, no positive relationships were detected between total coliforms and any water quality parameter or landscape feature.

This study shares a number of features with a recent investigation of urban runoff in southern California (23). Kebabjian found that densities of bacteria indicative of potential threats to human health (total coliforms, fecal coliforms, and enterococci) often exceeded standards for human contact in recreational waters. Existing standards for assessing risks to human health are based on bacteria originating from human wastes, however, whereas bacteria in urban runoff appear to originate primarily from pets and wildlife. Kebabjian also found that correlations between densities of indicator bacteria in urban runoff were very weak, suggesting that relationships between levels of such indicators and human pathogens maybe inconsistent (23). For these reasons, the impact of urban runoff on quality of recreational waters is extremely difficult to assess.

In conclusion, the present study revealed levels of E. coli and total-coliform bacteria in Lubbock playas that have implications for public health and should be examined in more detail despite the lack of current medical corroboration (14-16). Future research should include

1. a systematic compilation of reported incidents of gastrointestinal and skin disorders for people who have contact with playa water;

2. comparisons between densities of indicator organisms, such as total coliforms, and human pathogens in playa water;

3. continuous monitoring of bacterial densities in playas, especially with regard to runoff events;

4. an evaluation of bacterial densities in runoff water prior to entry in playas; and

5. an estimate of fecal inputs from pets and wildlife, including both fecal deposition within watersheds and bacterial characteristics of animal feces.

REFERENCES

1. Daniel, T.C., J.G. Konrad, and R.C. Wendt (1978), Nonpoint Pollution: Runoff in Urban Areas, Madison, Wisconsin: Wisconsin Department of Natural Resources, University of Wisconsin.

2. Boulanger, Y., G. Cousineau, and R. Lallier (1977), "Isolation of Enterotoxigenic Aeromonas from Fish," Canadian Journal of Microbiology, 23:1161-1164.

3. Levesque, B., P. Brousseau, E. Dewailly, J. Joly, M. Meisels, D. Ramsay, and P. Simard (1993), "Impact of the Ring-Billed Gull (Larus delawarensis) on the Microbiological Quality of Recreational Water," Applied Environmental Microbiology, 59(4): 1228-1230.

4. Bolen, E.G., H.L. Schramm, Jr., and L.M. Smith (1989), "Playa Lakes: Prairie Wetlands of the Southern High Plains," Bioscience, 39(9):615-623.

5. Haukos, D.A., and L.M. Smith (1994), "The Importance of Playa Wetlands to Biodiversity of the Southern High Plains," Landscape Urban Planning, 28:83-98.

6. U.S. EPA (1988), Bacteria Water Quality Standards Criteria Summaries: A Compilation of State/Federal Criteria, EPA 440/5-88/007, Washington, D.C.: Government Printing Office.

7. Standard Methods for the Examination of Water and Waste Water (1992), 18th edition, Washington, D.C.: APHA.

8. Hach Manual (1992), Loveland, Colorado: Hach Company.

9. Statistical Package for the Social Sciences (SPSS) (1990), Base System User's Guide, Chicago: SPSS, Inc.

10. Sokal, R.R., and F.J. Rohlf (1981), Biometry, San Francisco: WH. Freeman and Company.

11. Ennis, T.E. (1994), "City of Lubbock Stormwater NPDES Permit Data Analysis," Master's thesis, Lubbock, Texas: Texas Tech University.

12. Thompson, G.B., B.J. Claborn, R.M. Sweazy, and D.M. Wells (1974), Variation of Urban Runoff Quality and Quantity with Duration and Intensity of Storms, Interim report, Lubbock, Texas: Water Resources Center, Texas Tech University.

13. Wells, D.M., B.J. Claborn, R.H. Ramsey, and R.M. Sweazy (1975), Variation of Urban Runoff Quality and Quantity with Duration and Intensity of Storms: Phase III, Vol. 4, Project Summary, Final Report of the Office of Water Research and Technology, Lubbock, Texas: Water Resources Center, Texas Tech University.

14. Wolf, H.W. (1972), The Coliform Count as a Measure of Water Quality, In R. Mitchell, ed., Water Pollution Microbiology, New York: Wiley-Interscience, pp. 342-343.

15. Cabelli, V.J. (1981), Health Effects Criteria for Marine Recreational Waters, EPA-600/1-80-031, Cincinnati, Ohio: Government Printing Office.

16. Dufour, A.P. (1984), "Bacterial Indicators of Recreational Water Quality," Canadian Journal of Public Health, 75:49-56.

17. Kimbrough, R.C., Health Sciences Center, Texas Tech University Health Sciences Center, Personal communication.

18. Feachem, R. (1975), "An Improved Role for Faecal Coliform to Faecal Streptococci Ratios in the Differentiation Between Human and Non-Human Pollution Sources," Water Research, 9:689-690.

19. Porcella, D.B., and D.L. Sorensen (1980), Characteristics of Nonpoint Source Urban Runoff and its Effects on Stream Ecosystems, Corvallis, Oregon: Corvallis Environmental Research Laboratory, Office of Research and Development (U.S. EPA).

20. Jones, R., and C.C. Clark (1987), "Impact of Watershed Urbanization on Stream Insect Communities," Water Research Bulletin, 23:1047-1055.

21. Atrazine Hazards to Fish, Wildlife, and Invertebrates: A Synoptic Review (1989), Denver, Colorado: U.S. Fish and Wildlife Service.

22. Haukos, D.A., and L.M. Smith (1992), Ecology of Playa Lakes, Waterfowl Management Handbook Leaflet 13.3.7, Fort Collins, Colorado: U.S. Fish and Wildlife Service, Office of Information Transfer.

23. Kebabjian, R. (1994), "Monitoring the Effects of Urban Runoff on Recreational Waters," Journal of Environmental Health, 56(9):15-18.

Corresponding Author: Daryl L. Moorhead, Ecology Program, Department of Biological Sciences, Texas Tech University, Lubbock, Texas 79409-3131.