Fecal Bacteria in Agricultural Waters of the Bluegrass Region of Kentucky

Agricultural runoff influenced by nonpoint pollution frequently exceeds the USEPA standards for bacterial contamination of primary contact water (200 fecal coliforms/100 mL). Few studies bave evaluated the effect of cattle (Bos taurus) grazing on fecal contamination of ground water in the karst topography of central Kentucky. Our objectives were to: (i) observe the extent and pattern of fecal bacteria in agricultural waters from two central Kentucky watersheds; (U) determine if monthly sampling accurately assessed the extent and variability of fecal contamination; and (iii) assess the fecal coliform/ fecal streptococci ratio (FC/FS) as an indicator of fecal bacteria source. Springs, streams, and wells in two agricultural watersheds typical of central Kentucky were monitored for fecal coliform and fecal streptococci from December 1991 to January 1993. Springs and wells exceeded primary contact water standards, between 28 and 74% of the time; streams exceeded water quality standards between 87 and 100% of the time. When fecal bacteria were present, rainfall rapidly moved them from the soil surface into spring and well water. At two springs in Fleming county, only 29% of samples exceeded primary contact standards before cattle were present; 80% exceeded standards after cattle began grazing the surrounding pasture. Monthly sampling adequately reflected the extent of fecal contamination in our study, which had relatively continuous cattle grazing. Although the FC/FS ratio identified domestic animal contamination sources, it did not distinguish between domestic animal and human sources of contamination. G POLLUTION is an understandable concern for rural Americans. Ninety-five percent of rural residents in the USA who supply their own water Department of Agronomy, Univ. of Kentucky, Lexington, KY 405460091. Contribution of the Dep. of Agron., Univ. of Kentucky. Published as paper 94-3-56 with approval of the Director of the Kentucky Agric. Exp. Stn., Lexington.. Received 23 May 1994. Corresponding author ([email protected]). Published in ). Environ. Qual. 24:411-419 (1995). rely on groundwater as their primary source (flvento et al., 1992). Agricultural practices that promote nonpointsource pollution are a growing issue in rural areas. One pollution source is fecal contamination from domestic cattle (Bos taurus). Numerous studies show that agricultural runoif from pastures contains fecal bacteria concentrations, which frequently exceed the USEPA standard for primary contact water (200 fecal coliform/100 mL) (Doran and Linn, 1979; Jawson et al., 1982; Kunkle, 1970; Niemi and Niemi, 1991; Stephenson and Street, 1978). Watershed characteristics, land use management, and the proximity of domestic animals to streams play an important role in the severity of fecal contamination (Tiedemann et al., 1988). Cattle grazing increases fecal coliform in agricultural runoif compared with background fecal coliform levels (Dixon et al., 1977; Doran and Linn, 1979; Gary et al., 1983; Stephenson and Street, 1978; Tiedemann et al., 1988). When cattle are allowed to graze directly adjacent to streams, stream banks and bottoms became significant bacterial reservoirs (Kunkle, 1970). Several studies have been done on pastoral cattle grazing as a nonpoint source of pollution in the western USA (Doran and Linn, 1979; Jawson et al., 1982; Tiedemann et al., 1988). However these studies focused on stream, not groundwater contamination, and do not adequately represent the geological conditions and cattle management systems in the southeastern USA. Subsurface transport of bacteria to shallow springs and wells is a concern in karst areas where groundwater is utilized as a drinking water source. Our first objective was to observe the extent and pattern of fecal bacteria contamination hi Abbreviations: FC/FS, fecal coliform/fecal streptococci ratio; MUG, 4-methylumbelliferyl-p-D-glucuronide; CFU, colony forming unit. 412 J. ENVIRON. QUAL., VOL. 24, MAY-JUNE 1995 springs, streams, and wells of two typical agricultural watersheds in the Bluegrass region of Kentucky and determine how often these agricultural waters exceeded USEPA standards for minimum contact or swimming water. Economical and reliable water quality assessments are essential for effective water quality management (Cotter, 1985). Monthly sampling, while cost effective, may not reflect water quality variation due to water flow dynamics (Thomas et al., 1992). Fecal contamination of agricultural waters reflects complex interactions affecting the survival, infiltration, and movement of enteric organisms in soil, water, and sediment. Since these interactions are dynamic, fecal bacteria concentrations can vary dramatically with time at any given site. Diffuse loading of fecal contamination causes wide variations in fecal bacteria concentrations (Niemi and Niemi, 1990). Davis et al. (1977) suggested that one sample taken on a monthly basis may not represent water conditions throughout the period before subsequent samples are taken. Our second objective was to determine if monthly sampling of agricultural waters accurately assessed the extent and variability of fecal contamination. Rural agricultural waters receive three principle sources of fecal bacteria: human, domestic animal, and wild animal. To properly assess fecal contamination of a site, it is necessary to identify the contamination source. Geldreich (1976) suggested that the fecal coliform/fecal streptococci ratio (FC/FS) could be used to differentiate between contamination from human (FC/FS > 4), domestic animal (FC/FS between 0.1 and 0.6), and wild animal (FC/FS < 0.1) sources. Mean FC/FS ratio has been used to characterize some sites (Doran and Linn, 1979; Jawson et al., 1982). The frequency of FC/FS ratios representative of each contamination source has also been used (Tiedemann et al., 1988). Doran and Linn (1979) indicated that the FC/FS ratio is useful in distinguishing between domestic animal and wild animal sources of contamination, but its usefulness in differentiating between human and nonhuman sources of contamination is questionable. Our third objective was to determine whether the FC/FS ratio could be used to identify sources of fecal contamination in agricultural springs, streams, and wells. DESCRIPTION OF STUDY SITES This research was done on two watersheds located in central Kentucky, one in Bourbon County and one in Fleming County. These sites reflect agricultural systems common to central Kentucky. Land use is intermixed grain, hay, and tobacco (Nicotiana tabacum L.) croprazed pastures, feedlots, dairy lots, and forested Fhe soils of both counties are common to central Kentucky. They are deep, well drained soils on undulating broad ridgetops and moderately deep or shallow, well-drained soils on hilly uplands, formed in residuum from limestone or interbedded shale and limestone (G.W. Thomas and J. Haszler, 1994, personal communication). Bourbon County The geology of Bourbon County, interbedded limestone and shale or limestone with karst topography, is typical of the inner Bluegrass region of Kentucky. The Bourbon County watershed is dominated by .Lowell (fine, mixed, mesic Typic Hapludalf), Maury (fine, mixed, mesic Typic Paleudalf), and McAfee (fine, mixed, mesic Mollic Hapludalf) silt loam soils. Soil depth to bedrock ranges from 51 to 102 cm for the McAfee soil and 152 to 305 cm for the Maury soil (Richardson et al., 1982). Water samples were taken from four springs and two streams on a beef farm (Fig. I). Throughout the sampling period, Bourbon County water temperatures fluctuated between 8.3 and 14.2°C for springs and between 5.4 and 25.1 °C for streams. In 1992, the Bourbon County watershed was approximately 583 ha with 2% of the land in tobacco, 15% in corn (Zea mays L.), 4% in soybean [Glycine max (L.) Merr.], and 79% in hay-pasture. Approximately 400 beef cattle were managed at a stocking rate of one head per hectare. Table 1 shows the land use for each site in Bourbon and Fleming County (G.W. Thomas and J. Haszler, 1994, personal communication). The Bourbon County farm is representative of typical beef cattle management in the Bluegrass region of Kentucky. Pastures are intensively grazed, then left to fallow for varying periods. In the winter, hay is fed to cattle in some pastures, which concentrates cattle in small areas. Cattle were present at least once on all sites during the study, except at Spring B2, which was located below a house.