Stream water flowing from watersheds subjected to continuous and deferred rotation grazing by livestock was sampled to enumerate bacteria for detecting differences between grazing treatments and streams. Fecal coliforms, fecal streptococci, total counts at 200 C, and bacteria capable of fluorescing under long wave radiation were selected as indicators of pollution. The study was conducted two summers, 1979 and 1980, on mountain rangeland near Laramie, Wyo. Bacteria counts for different indicator groups varied in their ability to detect change between grazing treatments as well as between streams. Fluorescing bacteria and total counts were of little value in explaining nonpoint source pollution whereas fecal coliform and streptococci were. Variation in counts of fecal coliform and streptococci could not be fully accounted for by differences in grazing management but is partially explained by beaver damming of stream flow. Given that beaver impoundment of selected stream reaches is equal, variation in nonpoint pollution may be caused by differences in grazing treatments. Recently the role of livestock grazing in causing nonpoint pollution to stream flow has received attention (Stephenson and Street 1978, Doran and Linn 1979, Jawson et al. 1982). To help differentiate best management practices for grazing livestock on native rangeland, bacterial indicators of pollution must be determined. Many researchers have observed elevated fecal coliform and fecal streptococci counts in streams draining grazed rangeland (Morrison and Fair 1966, Kunkle and Meiman 1967, Stevenson and Street 1978, Doran and Linn 1979, Jawson et al. 1982). Declining counts of these bacteria were observed in the streams in pastures after the removal of livestock (Kunkle and Meiman 1967, Skinner et al. 1974b, Milne 1976, Skovlin et al. 1977, Duff 1978, Doran and Linn 1979). Variations in fecal coliforms in streams have been partially explained by rainfall runoff (Doran and Linn 1979, Jawson et al. 1982) deferred grazing management, animal density, access to streams, as well as physical, hydrologic and biological characteristics of drainage basins (Stephenson and Street 1978). In contrast Buckhouse and Gifford (1976) were unable to detect any significant changes in bacterial contamination associated with cattle grazing on watersheds using rainfall simulation to create overland flow. Milne (1976) and Stephenson and Street (1978) have found that livestock being fed on irrigated pastures during winter have a significant effect on bacterial contamination of adjacent streams. Milne (1976) reported, however, that this high contamination was short lived and quickly declined with the cessation of grazing. Bacterial concentrations have been shown by many to vary with stream flow (Morrison and Fair 1966; Kunkle and Meiman 1967; Skinner et al. 1972, 1974b; Stevenson and Street 1978; Jawson 1982). Indicator bacteria have been found to be more numerous in sediments than overlying water (Hendricks 1971, VanDonsel and Authors are associate professor, former graduate student, and assistant professor Division of Range Management; and associate professor, Division of Microbiology and Veterinary Medicine, College of Agriculture, University of Wyoming, Box 3354, Laramie 8207 I. This study was supported by Wyoming Agricultural Experiment Station funds, WYO-153-80. Published as Journal Article 1210 of WAES. 142 Geldrich 1971, Matson et al. 1978). Disruption of sediment has been found to increase fecal coliforms in the overlying water (Grimes 1975). Stephenson and Street (1978) inferred and Stephenson and Rychert (1982) confirmed settling of organisms in sediment could explain higher counts in streams at times when livestock were not present. These studies suggest that it is important to sample both sediment and overlying water to obtain an accurate evaluation of water quality in pastured lands. In addition to sedimentation, increased numbers of indicator bacteria may result from release of cells growing in algae mats (McFeters et al. 1977). They stated that stream aspect, shade, depth, and temperature were important for sustaining algae growth which allowed growth of trapped indictor bacteria. Johnson et al. (1978) reported that suspended solids in a stream typical of those draining the front range of Colorado were higher in a nongrazed pasture than in an adjoining downstream grazed pasture. Removal of beaver dams in the nongrazed pasture was thought to be the probable cause of the increased suspended solids in the stream. Foote (1937) found little effect of beaver upon the numbers of the coli-aerogens group of bacteria in water above and below beaver ponds. However, he stated that under certain circumstances beaver could excrete large numbers of these organisms. The purpose of the present study was to monitor change in bacterial counts within streams flowing through 2 different large pasture grazing systems. Beaver activity was prevalent on streams flowing through the study region. Their possible role in explaining differences found in bacteria counts between grazing systems and streams was investigated. Null hypotheses tested include (1) there were no differences in bacterial counts between grazing systems and (2) and there were no differences in bacteria counts between streams studied.