Characterization of the Coliform and Enteric Bacilli in the Environment of Calves with Colibacillosis

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1985, p. 949-954 0099-2240/85/040949-06$02.00/0 Copyright C 1985, American Society for Microbiology Vol. 49, No. 4 Characterization of the Coliform and Enteric Bacilli in the Environment of Calves with Colibacillosis PATRICIA I. Department of Veterinary Science' and Department of Bacteriology,2 North Dakota State University, Fargo, North Dakota 58105 PLEWS,'t MARY C. BROMEL,2t AND ITHEL A. SCHIPPER1* Received 10 April 1984/Accepted 13 December 1984 In the first part of the present study the coliform and enteric bacilli in the environment of calves with colibacillosis were examined. The occurrence, number, and pathogenic properties of Escherichia coli in barnyard soils were obtained from six cattle ranches. The 0 and K serogroups of E. coli isolates obtained from the feces of calves with colibacillosis born at these cattle ranches were determined, and their serotypes were compared with the E. coli 0 and K serotypes found in soils. The results showed a reservoir of potentially pathogenic E. coli in barnyard soils contaminated with bovine feces. For the second part of this study, 6 healthy calves and 51 calves with colibacillosis were studied. The numbers of total aerobic heterotrophic bacteria, total streptococci, fecal streptococci, total coliforms, and fecal coliforms in the feces of calves were determined. In addition, coliform and enteric bacilli from the feces of both healthy and diseased calves were identified, and their indole, methyl red, Voges-Proskauer, citrate (IMViC) types were described. In parallel, the IMViC types of coliform and enteric bacilli isolated from barnyard soils previously contaminated with bovine feces were compared with those isolated from unconitaminated soils. All fecal specimens were also examined for the presence of rotavirus. No significant effect on the numbers of the bacterial types was found. The results suggest that the predominant IMViC types found in the feces of calves with colibacillosis originate from the soil, From this study it is apparent that the occurrence, number, and survival of E. coli in barnyard soils is related to ranch husbandary and sanitary practices. The coliform bacteria are among those that inhabit the intestines of humans and animals. They are gram-negative, short bacilli that ferment lactose with acid and gas production in 48 h at 37 C. Members of this group include the genera Citrobacter, Escherichia, Enterobacter, Hafnia, and Klebsiella. The coliforms can be differentiated into two groups, the coliforms and the fecal coliforms, on the basis of their ability to grow at elevated temperatures. Both groups grow at 37 C, but only the fecal coliforms are able to grow at 44.5 C. The two groups can also be differentiated by four biochemical tests known as the indole, methyl red, Voges- Proskauer, citrate (IMViC) series. Escherichia coli is known to have three IMViC types: + + - -, + - - -, and - + - -. Geldreich (7), in studies of the IMViC types found in animal feces, undisturbed soils, and polluted soils, has designated these three IMViC types as Escherichia types, and he reported that they are predominant in animal feces. The coliforms that were found to be predominant in undisturbed soils, designated as the Enterobacter group, have three different IMViC types: - - + +, - - + -, and - - - +. All other possible IMViC combinations are designated as coliforms of the intermediate type. The IMViC series and the ability of fecal coliforms to grow at elevated temperatures have been used to differentiate the coliform bacteria found in polluted surface water and wastewater (1). In addition, the association of certain IMViC types with animal feces, or with undisturbed soils, has been used as an indication of the source of coliform pollution. * Corresponding author. t Present address: Department of Biological Sciences, University of Cincinnati, Cincinnati, OH 45221. t Present address: Great Plains Gasification Associates, Bulah, ND 58523. 949 The coliform bacillus E. coli usually does not cause severe disease in normal host organisms. However, it is an opportunistic pathogen that is able to cause disease when the bacilli invade extraintestinal tissue. Strains of E. coli that are able to cause disease without invasion of extraintestinal tissue are termed enteropathogenic E. coli. These strains possess specific pathogenic properties that allow them to colonize in animal intestines, multiply in high numbers, and induce diarrhea or other severe forms of enteric disease. One such property is the presence of pili or fimbriae that are known to aid in the attachment of enteropathogenic strains to intestinal epithelial cells. These pili are proteinaceous and are classified as noncapsular K antigens. The pili of enteropathogenic E. coli associated with diarrheal disease in piglets are termed K88 pili. The pili of strains associated with diarrhea in calves and lambs are antigenically distinct from the K88 pili and have been designated K99 pili. The association between E. coli and enterotoxic colibacillosis has been well established. However, few investigations into the occurrence, number, and pathogenic properties of the E. coli strains in the environment of calves with colibacillosis have been done. Smith and Crabb (18), employing bacteriophage typing of E. coli strains isolated from the feces of cows and their calves, concluded that the dam did not appear to be a frequent source of the strains responsible for disease and that the pen in which the calves were kept was a more probable source. In a study of the husbandary factors influencing the occurrence of colibacillosis in calves, Wray and Thomlinson (19) have concluded that the use of calf-houses free of fecal contamination could break the cycle of infection. This paper describes the occurrence, number, and pathogenic properties of enteropathogenic E. coli in barnyard soil samples taken at six cattle ranches. Four of these ranches were known to have high incidences of colibacillosis, whereas

950 PLEWS, BROMEL, AND SCHIPPER the remaining two were known to have a lower incidence of colibacillosis. In addition, we describe the IMViC types of the coliform and enteric bacilli isolated from uncontaminated soils where no cattle had been kept, from barnyard soils contaminated with bovine feces, and from the feces of calves with colibacillosis diarrhea. MATERIALS AND METHODS Collection of soil samples. Soil samples from six cooperating cattle ranches designated as herds A through F were taken in the fall of 1979 before the 1980 calving season. Soil samples were taken at various sites at each ranch. At each ranch soil samples were also taken at clean uncontaminated sites where no cattle had been kept. A second set of samples was taken at the beginning of the 1981 calving season at a seventh ranch designated herd G. A third group of samples was taken in early September 1981 at the feedlots of herd G. Samples were placed in sterile Whirl-Pak polyethylene bags and transported to North Dakota State University (NDSU) where they were stored at -20 C until tested. Collection of fecal specimens. Fecal specimens were obtained from the six ranches described above during the 1980 and 1981 calving seasons. Specimens were collected from healthy calves as well as from calves with diarrhea. In the second year of this study, additional fecal specimens were collected at herd G. The samples were placed in sterile plastic containers, frozen, and then transported to NDSU and stored at 4 C until tested. Source of control cultures. Control cultures were obtained from the American Type Culture Collection, Rockville, Md. (E. coli ATCC 12795), and the NDSU Department of Veterinary Science (E. coli 117E). Culture media. Primary plating media for the enumeration of total aerobic heterotrophs and the enumeration of enteric bacilli were plate count agar (BBL Microbiology Systems, Cockeysville, Md.), Levine eosin methylene blue agar (EMB; BBL); MacConkey agar (Difco Laboratories, Detroit, Mich.), and Hoektoen enteric agar (Hoektoen; BBL). Fecal streptococci were counted with M-enterococcus medium (BBL) containing 3% starch. Lactose broth (BBL), with bromcresol purple solution (2 ml/liter) added as the indicator, lauryl sulfate broth (BBL), and E.C. medium (BBL) were used in the most probable number (MPN) determination of coliforms and fecal coliforms. Simmons citrate agar (Difco), methyl red-voges-proskauer broth (BBL), tryptone broth (BBL), Kleigler iron agar (KIA; BBL), lysine iron agar (BBL), and urease test medium (BBL) were used in the further identification of enteric isolates. Trypticase soy agar (BBL) and Trypticase soy broth (BBL) were used for subculture and storage of isolates. Isolates of E. coli that were serotyped for the K99 antigen were cultivated with liquid minca medium as described by Guinee et al. (11). Enumeration of total aerobic heterotrophs. A 10-g soil sample was blended with 90 ml of sterile distilled water at high speed in a Waring blender for 3 min. This mixture was treated as a 1:10 dilution and further dilutions were made with 90- and 99-ml sterile water blanks. Dilutions were plated in triplicate on plate count agar and distributed by the spread-plate technique (1). The plates were incubated at 30 C for 24 h. Plates containing 30 to 300 colonies were counted. Total aerobic heterotrophs in the fecal specimens were enumerated by a similar method with modifications adapted to specimen size and consistency. Whenever possible, 1.0 g of fecal material was used. If specimens contained less than 1.0 g, 10 ml of sterile saline (0.85% NaCI) was added to the APPL. ENVIRON. MICROBIOL. entire sample, and the sample was blended with a Vortex mixer to give an even suspension. One milliliter of this suspension was then used. A total of 1 g or 1.0 ml of specimen was blended with 99 ml of sterile distilled water at high speed for 3 min in a Waring blender. Serial dilutions were made and plated in triplicate on plate count agar by the spread-plate method (1). Plates were incubated and counted as described above. Enumeration of enteric bacilli. Serial dilutions of soil samples and fecal specimens prepared as described above were plated in triplicate by the spread-plate method (1) on EMB, Hoektoen, and MacConkey agar. Plates were incubated at 37 C for 24 and 48 h. The lactose-fermenting and non-lactose-fermenting colonies were differentiated and counted. Enumeration of fecal streptococci. Fecal streptococci in soil samples and fecal specimens were counted by inoculating M-enterococcus medium with 1.0 ml of an appropriately diluted sample. A pour-plate method was employed (1). After solidification, the plates were incubated at 35 C for 48 h. Plates were counted and flooded with iodine and inspected for starch hydrolysis. All samples were plated in triplicate. Enumeration of total coliform bacteria. Total coliforms were determined by the multiple tube fermentation method (1) by using a five-tube test. Lactose broth and lauryl sulfate broth were inoculated with 1.0 ml of diluted sample. All tubes were incubated at 37 C for 24 and 48 h. The MPN was determined from tables published in previously established standards (1). Broths with gas in the Durham tube and a change in the indicator to yellow were considered positive. A loopful from positive tubes was streaked on EMB and MacConkey agars. The presence of typical coliform colonies was considered positive for total coliforms. Enumeration of fecal coliform bacteria. The presence of fecal coliform bacteria was determined by inoculating E.C. medium with a 0.01-ml calibrated metal loop from each positive lactose and lauryl sulfate broth tube. Tubes were incubated in a 44.5 C water bath for 24 h. A tube was considered positive if growth and gas production occurred. A loopful from positive tubes was streaked on EMB and MacConkey agars. The MPN was determined as described above. Identification of enteric bacilli. Representative colonies of enteric bacilli were selected from EMB, Hoektoen, and MacConkey plates, and the colonies were transferred to KIA tube medium. These isolates were identified to the genus or species level by biochemical and serological methods (4). Identification of coliform isolates. Bacterial colonies with a typical E. coli morphology and appearance on primary plating medium were counted; representative colonies were selected and transferred to KIA tube medium. In addition, selective plate media were streaked with inocula from positive lactose broth cultures, and typical coliform colonies were transferred to KIA medium. These primary isolates were tested for the ability to grow at 44.5 C in E.C. medium. Isolates capable of growth at elevated temperatures and any atypical isolates that grew poorly or produced very little or no gas in E.C. medium were biochemically characterized by their IMViC reactions. The identity of atypical strains giving a + + - - IMViC series was confirmed with API 20E strips (Analytab Products, Plainview, N.Y.). Typing sera. The following poly OK typing sera were used: (i) E. coli OK poly A containing antibodies to antigens 026:K60, 055:K59, 0111:K58, and 0127:K68 (Difco), and

VOL. 49, 1985 COLIFORMS AND BACILLI IN CALVES WITH COLIBACILLOSIS 951 (ii) E. coli OK poly B containing antibodies to antigens 086a:K61, 0119:K69, 0124:K72, 0125:K70, 0126:K71, and 0128:K67 (Difco). All isolates showing agglutination with poly A and poly B antisera were then serotyped with the individual sera. The K typing sera used were purchased from the Department of Veterinary Science, Pennsylvania State University, University Park, Pa. Additional K typing sera used were the gifts of H. W. Moon of the Animal Health Center, Ames, Iowa, and Steven Clegg of the University of Iowa, Iowa City, Iowa. Slide agglutination. Isolates of E. coli were subcultured in Trypticase soy broth before serotyping for 0 group antigens and the K88 antigen. Strains tested for K99 antigen were first subcultured in Trypticase soy broth with vigorous aeration at 37 C. Logarithmic-phase cells were then transferred to liquid minca medium and grown with vigorous aeration at 37 C for 8 to 10 h. The cells were sedimented by centrifugation, and the supernatant was drawn off. The pellet was dispersed with a Pasteur pipette, and the suspended cells were typed with K99 antisera. RESULTS All soil and fecal samples in the present study were analyzed for total aerobic heterotrophic bacteria, total streptococci, fecal streptococci, total coliforms, and fecal coliforms. Additionally, E. coli plate counts were done on all fecal samples. All counts represent the logarithm of viable count per gram of soil or gram of feces. The mean value and the standard deviation are given followed by the number of positive samples. The seven soil samples taken from areas where no cattle had been kept contained only aerobic heterotrophic bacteria, (8.24 + 0.78, 7) and total coliforms (3.87 ± 0.92, 7). However, streptococci and fecal coliforms we're not found in uncontaminated soil samples. In addition, these organisms were not found in uncontaminated soil samples taken at herds E, F, and G (data not shown). Twenty-two soil samples taken in feedlots, barns, and pastures contained all the bacterial types expected, including both fecal streptococci and fecal coliforms. Their values were as follows: total aerobic heterotrophic bacteria, 8.59 ± 0.69, 22; total streptococci, 4.77 ± 1.14, 21; fecal streptococci, 4.17 ± 1.46, 6; total coliforms, 4.02 ± 1.57, 22; and fecal coliforms, 3.75 ± 1.52, 17. In contrast to the results obtained from the ranches described above, eight soil samples from herds E and F did not contain fecal streptococci or fecal coliforms. However, similar counts were obtained for the following: total aerobic heterotrophic bacteria, 8.79 ± 1.07, 8; total streptococci, 4.42 1.26, 4; and total coliforms, 4.70 + 1.70, 5. Bacterial numbers in the feces of 51 calves with diarrhea were as follows: total aerobic heterotrophic bacteria, 9.39 ± 0.99, 45; fecal streptococci, 6.95 ± 1.13, 12; total coliforms, 7.89 1.75, 50; fecal coliforms, 9.97 ± 1.60, 38; and E. coli, 8.44 1.09, 38. Bacterial numbers in the feces of six healthy calves were as follows: total aerobic heterotrophic bacteria, 9.96 ± 0.60, 6; total streptococci, 8.42 + 1.98, 5; fecal streptococci, 9.99, 1; total coliforms, 8.81 2.25, 6; fecal coliforms, 8.64 ± 2.51, 6; and E. coli, 8.86 1.80, 6. A comparison of the number of total aerobic heterotrophic bacteria in the feces of healthy calves with the numbers of total aerobic heterotrophic bacteria in the feces of diseased calves by the F test for sample variance indicated that there was no detectable significant difference in the variance observed (P > 0.05). Similarly, the F test showed that there was no detectable significant difference in the variance in the number of total streptococci or the number of E. coli in the feces of these two groups of animals. Parr (15) has reported the decline and eventual disappearance of E. coli from fecal samples stored in an ice box. Evidence of a similar decline in viable E. coli numbers was seen in the present study. The data presented in Table 1 show the decline in total coliform, fecal coliform, and fermentative enteric bacilli in nine fecal specimens with increasing storage time. The number of E. coli found in diarrheal specimens in the present study was lower than expected. The possibility of the loss of viable count because of prolonged storage is a reasonable explanation for the lower counts. Starch-hydrolyzing Streptococcus species were detected in the feces of one of six healthy calves and in the feces of 12 of 51 diseased calves. The streptococci found were not of the enterococcal group but were group D streptococci, as determined by the method of Facklam (6). Because these strains were of bovine origin and rapidly hydrolyzed starch (12), they were designated as fecal streptococci. This species occurs in fresh feces (12) but is known to perish rapidly outside of the animal body. It is likely that this organism was lost in the majority of the fecal specimens during storage before analysis could be performed. In addition to the bacteriological examination of fecal specimcens, the specimens were examined for the presence of rotavirus by four methods: agar-gel immunodiffusion (E. R. Gion, M.S. thesis, NDSU, Fargo, 1980), fluorescent antibody tissue culture (Gion, M.S. thesis), counterimmunoelectrophoresis (5), and enzyme-linked immunosorbent assay (5). Rotavirus was detected in two of the healthy animals and in eight of the diseased calves. To determine whether the presence of rotavirus had any effect on the numbers of total aerobic heterotrophic bacteria, total streptococci, and total coliforms, a multivariate analysis of variance was performed by a statistical analysis system Manova procedure (10). The presence of rotavirus was found to have no significant effect on the number of total heterotrophic bacteria, total streptococci, or total coliforms (P > 0.5), nor was there evidence of an overall effect by rotavirus when the three groups of organisms were considered together by the Manova procedure. Table 2 shows the identification, the IMViC type, and the frequency of occurrence of the enteric and coliform bacilli isolated from uncontaminated soils, barnyard soil previously contaminated with bovine feces, and the feces of calves with colibacillosis. Isolates of E. coli were tested for the presence of 10 0-type antigens (026, 055, 086a, 0111, 0119, 0124, 0125, TABLE 1. Effect of the length of storage time of fecal samples on bacterial counts Log of counts (MPN) of: Log of plate Fecal counts of Days of sample Total Fecal fermentative storage coliforms coliforms enteric bacilli 32N2 10.1 10.1 10.0 17 2N6 9.6 9.6 9.8 34 31N2 9.7 9.7 9.6 48 3N3 8.2 8.2 7.6 48 28N2 7.4 7.4 7.4 54 7N1 8.9 7.5 9.5 83 9N1 7.7 7.7 7.6 83 501 4.5 0.0 0.0 97 832 3.8 0.0 0.0 97

952 PLEWS, BROMEL, AND SCHIPPER APPL. PLENVIRON. MIcR MICROBIOL. TABLE 2. Coliforms and enteric bacilli identified in uncontaminated control soil, barnyard soil, and calf feces and IMViC types Frequency of occurrence (%) in the following samples: Organism Type no. IMViC type Control soil Barnyard soil Calf feces (n = 50) (n = 50) (n = 60) Escherichia types Citrobacter species 1 + + - - 0 16 0 Enterobacter agglomerans 1 + + - - 0 2 0 Escherichia coli 1 + + - - 0 14 29 Klebsiella oxytoca 1 + + - - 0 0 2 Citrobacter freundii 4 - + - - 19 10 13 Enterobacter agglomerans 4 - + - - 4 2 6 Escherichia coli 4 - + - - 0 0 2 Hafnia als'ei 4 -+-- 8 4 0 Klebsiella ozaenae 4 - + - - 0 6 9 Escherichia coli 10 + - - - 0 0 2 Enterobacter types Enterobacter spp. 2 - - + + 13 0 2 Enterobacter cloacae 2 - - + + 0 10 0 Klebsiella pneumoniae 2 - -+ + 0 2 2 Enterobacter spp. 14 - - + - 2 0 0 Hafnia alvei 14 --+- 0 0 2 Intermediate types Citrobacter spp. 5 - + - + 0 4 0 Citrobacter freundii 5 - + - + 0 0 6 Hafnia alvei 5 -+-+ 21 0 0 Klebsiella ozaenae 5 -+-+ 21 0 0 Klebsiella pneumoniae 5 - + - + 0 0 4 Yersina enterocolytica 5 - + - + 0 0 2 Citrobacter spp. 6 + + - + 0 8 2 Citrobacter diversus 6 + + - + 0 8 0 Enterobacter cloacae 7 - +++ 0 2 0 Klebsiella oxytoca 7 - + + + 0 0 2 Klebsiella pneuinoniae 7 - +++ 0 0 9 Serratia spp. 7 -+++ 0 4 0 Klebsiella oxytoca 8 + + + + 0 0 2 Enterobacter cloacae 9 + - + + 0 2 0 Klebsiella oxytoca 9 + - + + 13 6 0 Hafnia alvei 12 -++- 0 0 4 Yersina pseudotuberculosis 13 ---- 0 0 2 0126, and 0128) and for two K-type antigens (K88 and K99) as well. Isolates from two soil samples and three fecal specimens obtained from herd A were found to be serologically identical. These strains agglutinated in all but one of the 0-typing sera (0126). These strains were K88- and K99+. Similarly, four soil isolates and two fecal isolates from herd C were found to be identical. The strains were positive for all 0-type antigens except 0124, 0126, and 0128. As with the isolates from herd A, these strains were K88- and K99+. DISCUSSION The absence of streptococci and fecal coliforms in the uncontaminated soil samples from all six cattle ranches taken at sites where no cattle had been kept demonstrates that these organisms are not normal soil inhabitants. Their presence in barnyard soils subject to fecal contamination indicates that bovine feces are the source of these organisms. This is in agreement with the findings of Geldreich (7). The repeated isolation of K99-positive enteropathogenic E. coli from barnyard soils and the recovery of serologically identical enteropathogenic E. coli from soil samples collected in the fall and from the feces of calves with colibacillosis born at the same locations indicate a reservoir of enteropathogenic E. coli in barnyard soil contaminated with bovine feces. Furthermore, the results of this study suggest that the occurrence and survival of fecal coliforms in barnyard soils is related to ranch husbandary and sanitary practices. The absence of fecal coliforms and fecal streptococci at herds E and F should be noted. These two ranches were known to have had an incidence of neonatal calf diarrhea that was significantly lower than that at herds A through D where fecal coliforms and fecal streptococci were consistently isolated from barnyard soils. Husbandary practices at these two ranches included the clearing of manure from the lots and the barnyards where calves were born and reared. This was done as soon as spring weather conditions permitted the moving of cattle to pasture locations. After the cattle were removed, these sites were allowed to stand empty until the fall. The lower incidence of diarrhea in calves born at these two ranches provides circumstantial evidence that calves may acquire disease-causing enteropathogenic E. coli strains from contaminated barnyard soil. The pattern of coliform survival in soils observed in this study was similar to previously reported survival patterns (2, 3, 13). Although E. coli is known to die off rapidly in soils, the survival of enteric bacteria in soil is increased, and regrowth is possible when sufficient organic matter and moisture are present (9). Removal of the manure from the barnyards at herds E and F eliminated moisture-retaining

VOL. 49, 1985 COLIFORMS AND BACILLI IN CALVES WITH COLIBACILLOSIS 953 organic material. It appears that this practice is sufficient to eliminate fecal coliforms from contaminated soil. Although the ecology of coliforms and fecal coliforms was the primary concern of this study, the occurrence of another bacterial species used as an indicator of fecal pollution was studied. Streptococci were found in slightly higher numbers than the coliforms in contaminated soil at herds A through D. A similar pattern has been reported in studies of the survival of indicator bacteria in aquatic environments (8, 14), where enterococci have been found to survive longer than coliforms. These findings suggest that the survival patterns of streptococci in contaminated soils are similar to those in polluted surface waters. The results of the present study are in agreement with those of Smith (16, 17) who did not observe any difference in the principal bacterial flora of healthy and diseased animals. He reported an increased number of E. coli in a group of severely ill animals. However, in the present study, in which the severity of disease was not determined, the number of E. coli in diarrhea specimens was lower than expected. The possibility of a loss of viable count because of prolonged storage has been discussed previously. In addition, had the severely ill animals been treated as a separate group, a higher count of E. coli might have been observed. The number of E. coli in feces of both healthy and moderately ill calves do not appear to be significantly different, thus suggesting that E. coli do not need to be present in high numbers to cause mild or moderate disease. In studies by Geldreich (7) on the occurrence of different IMViC types of coliform bacteria in soil and animal feces, 12 IMViC types were reported. In the present study two additional types are reported, 13 (----) and 14 (--+-) (Table 2). Isolates belonging to three genera other than Escherichia produced a classical E. coli IMViC type 1 (Table 2). A total of 10 coliforms isolated that gave an IMViC type 1 not belonging to the genus Escherichia were identified. Furthermore, the Enterobacter agglomerans identified was cultured from E.C. broth incubated at 44.5 C. The Durham tube contained gas, but no E. coli was found when a loopful of this broth culture was streaked onto EMB medium. These findings underscore the limitations of the use of the IMViC test as a sole criterion in the identification of fecal coliforms. They also emphasize the importance of the completed test for fecal coliforms by the MPN method. No coliform bacteria producing the classical E. coli IMViC type 1 were detected in uncontaminated soils. However, three genera producing the Escherichia IMViC type 4 were isolated. However, this IMViC type was reported by Geldreich (7) to constitute 3.3% of the coliforms present in undisturbed soils. The intermediate type S was isolated frequently from uncontaminated soils in this study. This is in agreement with Geldreich's conclusion that this is a predominant type in undisturbed soils (7). According to Geldreich (7), the IMViC type 9 constituted 2.9% of the coliforms isolated from undisturbed soils, but was not found in polluted soils, and constituted less than 1.0% of coliforms found in livestock feces. In the present study, this type was found in uncontaminated and barnyard soils but was not found in the feces of calves with colibacillosis. These findings suggest that the IMViC type 9 is a soil type. Geldreich reported three predominant IMViC types in livestock feces: 1, 3, and 6. They represented 99% of the coliforms isolated from livestock feces. However, in the present study, two of these types, 1 and 6, constituted only 33% of the coliforms isolated from calf feces. Furthermore, type 3 was not found in any of the fecal samples. The data of Geldreich (7) for livestock feces were obtained from the examination of 31 fecal samples: 10 from sheep, 11 from cows, and 10 from hogs. There is no information in this reference about the age or the state of health of these animals; however, the fecal specimens examinated in this study were obtained from newborn calves and this may account for the noted discrepancy. Unexpectedly, in the present study, IMViC types 2, 4, 5, 6, 7, and 8 constituted 61% of the coliforms isolated from calf feces. These six types have been reported by Geldreich (7) to occur at the highest frequency in undisturbed soils. It is interesting that these IMViC types were not those commonly associated with livestock feces, but rather they are associated with undisturbed and polluted soils. This suggests that the predominant IMViC types found in the feces of calves with colibacillosis originate in the soil. If newborn calves acquire enteropathogenic E. coli from contaminated soils, as the 0- and K-type serotyping of soil and fecal coliform isolates in the present study suggests, it seems reasonable that the coliforms that are naturally present in soils also would be present in their feces. ACKNOWLEDGMENTS Partial funding for this work was provided by the North Dakota Beef Commission, 107 S. 5th St., Bismarck, N.D. We thank Robert B. Carlson for assistance in the statistical analysis of the data presented here. LITERATURE CITED 1. American Public Health Association. 1975. Standard methods for the examination of water and wastewater, 14th ed. American Public Health Association, Inc., New York. 2. Bell, R. G., and J. B. Bole. 1978. Elimination of fecal coliforms from soil irrigated with municipal sewage lagoon effluent. J. Environ. Qual. 7:193-196. 3. Edmonds, R. L. 1976. Survival of coliform bacteria in sewage sludge applied to a forest clearcut and potential movement to groundwater. Appl. Environ. Microbiol. 32:538-546. 4. Edwards, P. R., and W. H. Ewing. 1972. Identification of Enterobacteriaceae, 3rd ed., Burgess Publishing Co., Minneapolis, Minn. 5. Ekern, L. J., I. A. Schipper, and K. J. McMahon. 1981. Neonatal bovine enteritis: detection of rotavirus by counterimmunoelectrophoresis and enzyme-linked immunosorbent assay. Can. J. Comp. Med. 45:135-139. 6. Facklam, R. R. 1980. Streptococci and aerococci, p. 88-110. In E. H. Lennette, A. Balows, W. J. Hausler, and J. P. Truant (ed.), Manual of Clinical Microbiology, 3rd ed. American Society for Microbiology, Washington, D.C. 7. Geldreich, E. E. 1966. Sanitary significance of fecal coliforms in the environment. Water Pollution Control Research Series, Publication no. WP-20-3. Cincinnati Water Research Laboratory, Cincinnati, Ohio. 8. Geldreich, E. E., H. F. Clark, and C. B. Huff. 1964. A study of pollution indicators in a waste stabilization pond. J. Water Pollut. Cont. Fed. 36:1372-1479. 9. Gerba, C. P., C. Wallis, and J. L. Meknick. 1975. Fate of wastewater bacteria and viruses in soil. ASCE J. Irrig. Drainage Div. 101:157-174. 10. Goodnight, J. H. 1979. GLM procedure, p. 237-263. In J. T. Helwig and R. A. Council (ed.), SAS Users Guide. SAS Institute Inc., SAS Circle, Cary, N.C. 11. Guinee, P. A. M., W. H. Jansen, and C. M. Agterberg. 1976. Detection of the K99 antigen by means of agglutionation and immunoelectrophoresis in E. coli isolates from calves and its correlation with enteropathogenicity. Infect. Immun. 13: 1367-1377. 12. Hungate, R. E. 1966. The rumen bacteria, p. 8-90. In the rumen

954 PLEWS, BROMEL, AND SCHIPPER and its microbes. Academic Press, Inc., New York. 13. Klein, D. A., and L. E. Casida, Jr. 1967. Escherichia coli die-out from normal soils as related to nutrient availability and indigenous microflora. Can. J. Microbiol. 13:1461-1469. 14. McFeters, G. A., G. K. Bissionnette, J. J. Jezeski, C. A. Thomson, and D. G. Stuart. 1974. Comparative survival of indicator bacteria and enteric pathogens in well water. Appl. Microbiol. 27: 823-829. 15. Parr, L. W. 1936. Sanitary significance of the succession of coli-aerogenes organisms in fresh and stored feces. Am. J. Pub. Health. 26:39-45. APPL. ENVIRON. MICROBIOL. 16. Smith, H. W. 1962. Observation on the aetiology of neonatal diarrhea in calves. J. Pathol. Bacteriol. 84:147-168. 17. Smith, H. W. 1971. The bacteriology of the alimentary tract of domestic animals suffering from Escherichia coli infection. Ann. N.Y. Acad. Sci. 176:110-125. 18. Smith, H. W., and W. E. Crabb. 1956. The typing of Escherichia coli by bacteriophage: its application in the study of E. coli populations of the intestinal tract of healthy calves and calves suffering from white scours. J. Gen. Microbiol. 15:566-574. 19. Wray, C., and J. R. Thomlinson. 1975. Factors influencing the occurrence of colibacillosis in calves. Vet. Rec. 96:52-56.