EVALUATION OF POPULATION EFFECTS OF BOVINE TUBERCULOSIS IN FREE-RANGING AFRICAN BUFFALO (SYNCERUS CAFFER)

Journal of Mammalogy, 82(1):231 238, 2001 EVALUATION OF POPULATION EFFECTS OF BOVINE TUBERCULOSIS IN FREE-RANGING AFRICAN BUFFALO (SYNCERUS CAFFER) TIMOTHY C. RODWELL, IAN J. WHYTE, AND WALTER M. BOYCE* Department of Veterinary Pathology, Microbiology and Immunology, University of California, One Shields Avenue, Davis, CA 95616 (TCR, WMB) Kruger National Park, Private Bag X402, Skukuza 1350, South Africa (IJW) We examined data from 3,743 buffalo (Syncerus caffer) culled between 1991 and 1998 in Kruger National Park, South Africa, to evaluate effects of bovine tuberculosis on the buffalo population and examine risk factors of bovine tuberculosis. We found no evidence that bovine tuberculosis affected fertility or lactation status of female buffalo, but adult buffalo 3 years old were underrepresented in infected herds. Logistic regression analysis demonstrated that older buffalo were at greater risk for bovine tuberculosis than younger buffalo and risk of acquiring bovine tuberculosis increased for all age groups as prevalence in the herd increased. African buffalo, bovine tuberculosis, natural populations, Syncerus caffer, wildlife dis- Key words: ease Bovine tuberculosis is a chronic wasting disease caused by the pathogen Mycobacterium bovis that infects a wide range of hosts, including humans (Morris et al. 1994). Although bovine tuberculosis primarily is a disease of domestic animals and humans, it has become widespread in wildlife (O Reilly and Daborn 1995). The disease has been well studied over a period of decades in small mammals such as brushtailed possums (Trichosurus vulpecula Barlow 1994) and badgers (Meles meles White and Harris 1995), but now bovine tuberculosis also is recognized as an important disease of bison (Bison bison Tessaro et al. 1990), deer (Odocoileus virginianus Schmitt et al. 1997), and African buffalo (Syncerus caffer Bengis et al. 1996). One of the largest free-ranging buffalo populations in southern Africa is found in Kruger National Park (KNP), South Africa. Starting from an initial size of 1,000 an- * Correspondent: wmboyce@ucdavis.edu imals in 1910, the population grew to about 16,000 by 1969 (I. J. Whyte, in litt.). After 1969, the population was kept between 22,000 and 35,000 buffalo by annual culling. However, between 1991 and 1992, a massive population crash followed a severe drought, and culling was terminated. Bengis et al. (1996) speculated that bovine tuberculosis was introduced into buffalo in southern parts of KNP during the major outbreaks that occurred in cattle outside KNP in the 1960s and again in the 1980s. During both outbreaks, cattle were infected with Corridor disease (buffalo-associated theileriosis), which indicated close contact between the 2 species (Bengis et al. 1996). By the late 1980s bovine tuberculosis largely was eradicated from domestic animal populations surrounding KNP (Kloeck 1998), but it apparently persisted undetected inside KNP until 1990 when it was diagnosed opportunistically in a single buffalo (Bengis 1999). Subsequent postmortem examinations indicated that M. bovis was 231

232 JOURNAL OF MAMMALOGY Vol. 82, No. 1 likely spread via the aerosol route in bovine tuberculosis infected buffalo and the highly social nature of buffalo facilitated infection (Keet et al. 1994). By 1998, the bovine tuberculosis prevalence in KNP was estimated to be 1.5%, 16%, and 38.2% in the northern, central, and southern zones of KNP, respectively (Rodwell 2000). Prevalence had increased significantly in the southern and central zones between 1991 and 1998, at an average annual rate of about 1.7%/year, and no indication was found that prevalence was stabilizing at 1998 levels. No practical treatment exists for bovine tuberculosis in wildlife, and although vaccines are in development, none are available for use in large free-ranging wildlife populations. There is little doubt that bovine tuberculosis can kill animals and humans (Cosivi et al. 1995), but it is unclear if morbidity and mortality caused by bovine tuberculosis in buffalo has been, or will be, sufficient to affect population viability. We conducted a retrospective examination of data from 3,500 buffalo culled between 1991 and 1998 to determine if any evidence existed of a negative effect of bovine tuberculosis on the buffalo population. We looked for evidence of a decline in reproductive success, a significant change in population age structure, and trends in overall population growth from 1989 to 1998. To predict how bovine tuberculosis might affect future population growth, we also assessed how age, sex, geographic location, and prevalence in the herd affected risk of contracting bovine tuberculosis. MATERIALS AND METHODS Study area. Kruger National Park is the largest wildlife refuge in South Africa. It is about 20,000 km 2 (22 19 25 32 S, 30 52 32 03 E). Together with the private game farms that surround much of its western boundary, KNP is one of the largest protected environments in Africa. To increase precision of estimates of bovine tuberculosis prevalence, the buffalo population in KNP was divided previously into 3 geographic zones (Rodwell 2000). The southern zone (3,650 km 2 ) lay between the southern KNP border and the Sabie River, the central zone (5,627 km 2 ) lay between the Sabie and Olifants rivers, and the northern zone (10,039 km 2 ) lay between the Olifants River and the northern boundary of KNP. Vegetation, soil, and topology of each zone were described by Gertenbach (1983). Buffalo movement between zones was limited because of the 2 perennial rivers (Sabie and Olifants Whyte 1996). We referred to the buffalo population in each zone as a subpopulation of the total KNP population. Data collection. We conducted a cross-sectional survey of 3,743 buffalo culled between 1991 and 1998. Buffalo herds were sampled at random from breeding herds counted in the annual KNP census of large mammals. About 20 30 buffalo were culled from each selected herd without regard to age or sex, using the approved KNP culling technique (De Vos et al. 1983). Culled buffalo were aged in the field using incisor eruption and molar wear patterns (Grimsdell 1973). Buffalo were aged from 0 (calf) to 20 years in 1-year intervals. All buffalo were eviscerated and examined in the field, and the carcasses, with thoracic organs intact, were moved to a slaughterhouse for final processing. Pregnancy status was determined in the field by examination of the uterus for presence of a fetus (Whyte 1996), and cows were determined to be lactating if they had enlarged udders where milk was expressed on palpation. Bovine tuberculosis was diagnosed by necropsy of the entire buffalo, with detailed macroscopic inspection of intestinal and thoracic organs and microscopic examination of lymph node sections (Bengis et al. 1996). Reproduction and demographic effects. The effect of bovine tuberculosis on pregnancy status and lactation of adult buffalo cows was explored with logistic regression (Hosmer and Lemeshow 1989). Pregnancy and lactation were the outcome variables, and infection with bovine tuberculosis was the covariate. Zone and age group also were included as confounding variables. Zone consisted of 3 categories (northern, central, and southern), and 4 age groups represented the 4 quartiles of the total age distribution in years (1 2, 3 4, 5 7, and 7). We conducted logistic regression analysis with SPSS 8.0 (SPSS Inc. 1997). Direct examination of survivorship was not possible from our retrospective study, so we examined the age distribution of bovine tubercu-

February 2001 RODWELL ET AL. BOVINE TUBERCULOSIS RISKS 233 losis positive and bovine tuberculosis negative buffalo herds as an indirect measure of the cumulative effect of survivorship differences in diseased and nondiseased herds. We consolidated all bovine tuberculosis positive herds into 1 group (1,796 buffalo from 71 herds) and all bovine tuberculosis negative herds into a 2nd group (1,995 buffalo from 74 herds). Consolidated groups were divided into 3 biologically meaningful age groups. Young cows (birth to 2 years) were prereproductive, whereas prime (3 12 years), and old ( 12 years) included the reproductive adults. Differences in age distribution between the bovine tuberculosis and non bovine tuberculosis groups were analyzed with a 3-by-2 contingency table (Sokal and Rohlf 1995), using Epi Info 6.04 (Division of Surveillance and Epidemiology, Centers for Disease Control and Prevention, Atlanta, Georgia). Population growth of buffalo was analyzed using census data collected between 1989 and 1998. Census data were collected each year by helicopter in a total count of the buffalo population in KNP. All breeding herds were photographed for precise counting, and the same methods were used every year (Joubert 1983; Whyte 1996). Census data were retrospectively allocated into the 3 zones in KNP, and linear regression (Sokal and Rohlf 1995) was used to analyze changes in population size in the most recent 5-year period, 1993 1998. Bovine tuberculosis risk factors analysis. Factors potentially affecting the risk of a buffalo acquiring bovine tuberculosis were examined with logistic regression. Bovine tuberculosis was the outcome variable, and age, sex, and zone were covariates. To control for herd effects on the probability of a buffalo being bovine tuberculosis positive, bovine tuberculosis prevalence in the herd from which the sample was obtained was used as a covariate in the model (Weigel et al. 1995). There were 4 age groups corresponding to quartiles of the age distribution in years (1 2, 3 4, 5 7, and 7). Zone consisted of 3 categories, and prevalence in the herd consisted of 4 categories (2 11%, 11 25%, 25 41%, and 41%) corresponding to quartiles of the total distribution of herd prevalence values in the sample. Only herds with 1 bovine tuberculosis positive buffalo were included in the logistic regression analysis. Variables with P 0.05 were excluded from the final model. Following standard procedures (Collett 1994), the final model was converted into a probability equation for bovine tuberculosis and graphed. RESULTS Bovine tuberculosis effects. A total of 1,638 reproductive-aged female buffalo were examined for bovine tuberculosis and their pregnancy and lactation status. Of the 1,638 females, 647 were pregnant (115 bovine tuberculosis positive, 532 bovine tuberculosis negative), and 991 were not pregnant (130 bovine tuberculosis positive, 861 bovine tuberculosis negative). Bovine tuberculosis infection did not have a significant effect (P 0.59) on pregnancy status in the logistic regression model (Table 1). Age group (P 0.01) and zone (P 0.01) were significant. Of the 1,638 females, 465 were lactating (62 bovine tuberculosis positive, 403 bovine tuberculosis negative) and 1,048 were not lactating (148 bovine tuberculosis positive, 900 bovine tuberculosis negative). No lactation data were available for 125 of the 1,638 reproductiveaged buffalo. Bovine tuberculosis infection (P 0.79) was not significant in the logistic regression model with lactation as the dependent variable (Table 1). Lactation rates varied by age group (P 0.01) and zone (P 0.01). Age distribution of buffalo herds with bovine tuberculosis differed from that in buffalo herds with no detectable bovine tuberculosis (P 0.01). Proportions of prime (49%) and old (3%) buffalo in the bovine tuberculosis positive herds were lower than the proportions of prime (54%) and old (4%) buffalo in the bovine tuberculosis negative herds. In contrast, young buffalo were significantly overrepresented in the bovine tuberculosis positive herds (48%) relative to the bovine tuberculosis negative herds (42%). Buffalo subpopulations in each zone remained stable from 1989 to 1991, at about 17,000 (northern), 8,000 (central), and 4,000 (southern) buffalo (Fig. 1). From 1991 to 1993, the northern zone subpopulation decreased about 55%, whereas the

234 JOURNAL OF MAMMALOGY Vol. 82, No. 1 TABLE 1. Effect of bovine tuberculosis (BTB) on pregnancy and lactation status of female African buffalo in Kruger National Park, as determined by logistic regression. Zone and age group were included as potential confounding variables. Variable Statistical significance Pregnancy status Odds ratio Odds ratio 95% CI Statistical significance Lactation status Odds ratio Odds ratio 95% CI BTB 0.59 1.10 0.77 1.59 0.79 0.94 0.62 1.43 Zone Northern Central Southern Age group (years) b 3 4 5 7 7 1.95 2.33 13.25 25.48 1.48 2.56 1.68 3.24 8.65 20.29 17.01 38.16 0.14 a Reference category for dummy variable. b Age group 1 2 years excluded from analysis because cows 2 years old do not reproduce. 1.88 1.31 38.39 87.07 1.40 2.52 0.91 1.89 16.60 88.75 38.31 197.87 central and southern subpopulations declined about 43% and 24%, respectively. Subpopulations reached their minimum levels in 1993, 1994, and 1995 in the southern, central, and northern zones, respectively (Fig. 1). After reaching their respective minimum levels, buffalo subpopulations grew in the southern (P 0.01), central (P 0.01) and northern zones (P 0.05). Bovine tuberculosis risk analysis. Data from 1,755 culled buffalo (70 herds) were analyzed with logistic regression to determine if age group, sex, zone, and prevalence of bovine tuberculosis in the herd had a significant effect on the risk of acquiring bovine tuberculosis. Sex (P 0.34) and zone (P 0.26) were not significant variables and were excluded from the model. The final model included age group, prevalence in the herd, and a constant (Table 2). Probability of bovine tuberculosis infection increased with age up until the 5- to 7-year age group (Fig. 2), but buffalo 7 years of age had a lower probability of bovine tuberculosis than those in the 5- to 7-year age group. The relationship between risk of bovine tuberculosis and age remained unchanged as prevalence in the herd increased, and all age groups had a higher risk of bovine tuberculosis as prevalence in the herd increased. DISCUSSION FIG. 1. Trends in the size of buffalo subpopulations in the northern, central, and southern zones of Kruger National Park, 1989 1998. Disease may regulate natural populations, but our understanding of how populations are impacted is still limited (Gulland 1997). The only well-documented studies showing regulation of a natural population by microparasites have been those where

February 2001 RODWELL ET AL. BOVINE TUBERCULOSIS RISKS 235 TABLE 2. Final logistic regression model of risk of bovine tuberculosis in buffalo of Kruger National Park; only variables with P 0.05 were included in final model. Variable Coefficient (B) Statistical significance Odds ratio Odds ratio 95% CI Age group (years) 1 2 3 4 5 7 7 Herd prevalence (%) 2 11 11 25 25 41 41 Constant 0.42 0.84 0.71 1.52 2.32 2.04 1.11 2.08 1.66 3.25 1.47 2.83 1.27 2.11 3.05 3.18 3.56 8.23 21.21 2.26 5.60 5.35 12.65 13.74 32.72 a Reference category for dummy variable. the pathogen was highly virulent and the effect was dramatic and evident in a single host generation; for example, rinderpest (Plowright 1982) and myxomatosis (Brothers et al. 1982). Conversely, bovine tuberculosis is a chronic disease, and clinical symptoms may take years to manifest, making it extremely difficult to identify population effects. We examined indicators of reproductive success and population demography to identify disease effects that may have implications for the future health of the buffalo population in KNP. The proportion of pregnant and lactating buffalo cows infected with bovine tuberculosis in KNP was not significantly different from buffalo cows that were not infected (Table 1). Thus, we concluded that bovine tuberculosis did not affect fertility of female buffalo or lactation status from 1991 to 1998. However, age distribution of buffalo varied significantly in bovine tuberculosis and non bovine tuberculosis infected herds. Young buffalo were overrepresented and adult buffalo were underrepresented in the bovine tuberculosis positive herds. One possible interpretation is that mortality of adult buffalo was higher in bovine tuberculosis positive herds. We found that risk of bovine tuberculosis was correlated positively with age (Fig. 2). Therefore, if bovine tuberculosis caused significant mortality, the effect would have been most noticeable in the older age groups. Evidence of age effects due to disease is rare in the literature. A study that demonstrated a similar effect to that that we observed, but on young rather than adult animals, was an examination of effects of feline panleukopenia, which caused mostly subadult mortalities in populations of feral cats (Felis catus Van Rensburg et al. 1987). After introducing the pathogen into a population of feral cats, young animals were significantly underrepresented in the population relative to older animals. An alternative explanation for the observed difference in age distributions is that a density-dependent response occurred after the earlier declines in subpopulations. Because the most severe decline was in the north (Fig. 1), and most bovine tuberculosis free herds came from the north, one would expect the young animals in the bovine tuberculosis free herds to be overrepresented if a density-dependent effect occurred. However, it was the young buffalo from the bovine tuberculosis positive herds (mostly southern herds) that were overrepresented. Prime and old buffalo in KNP produce 99% of the offspring in a herd (data not shown) and their underrepresentation in bovine tuberculosis positive herds may have

236 JOURNAL OF MAMMALOGY Vol. 82, No. 1 FIG. 2. Bovine tuberculosis (BTB) risk by age group, controlled for prevalence in the herd. Data generated from output of the logistic regression model of risk of bovine tuberculosis, which included bovine tuberculosis as the outcome variable and age group and prevalence in the herd as explanatory variables. a long-term negative effect on reproductive success of a herd. However, without additional supporting data on bovine tuberculosis-induced mortality and age-specific survivorship, we cannot definitively conclude that differences in age structure are a result of bovine tuberculosis. It is unlikely that bovine tuberculosis played a major role in the population declines of buffalo in KNP from 1991 to 1993 (Fig. 1). The decline was most severe in the northern zone where prevalence of bovine tuberculosis was below detectable levels in 1991, and least severe in the southern zone where prevalence was highest (Rodwell 2000). The decline probably was precipitated by a drought that occurred in 1991 1992. Buffalo need daily access to water, and population fluctuations of buffalo have been related directly to variation in rainfall (Mills et al. 1995). After the northern, central, and southern subpopulations reached their minimum in 1995, 1994, and 1993, respectively, all subpopulations increased (Fig. 1). The northern subpopulation grew faster than subpopulations in the central and southern zones, but samples were small (n 4 years in the northern zone), and we could not conclude with confidence that the observed trend was significant. Logistic regression analysis provided strong evidence of a positive relationship between risk of acquiring bovine tuberculosis and increasing age (Fig. 2; Table 2). That relationship was not changed by prevalence in the herd (no significant interaction was found between age group and prevalence), but risk of bovine tuberculosis increased for all ages as prevalence in the herd increased. No explanation is obvious

February 2001 RODWELL ET AL. BOVINE TUBERCULOSIS RISKS 237 for the slightly lower risk of bovine tuberculosis observed in the oldest quartile of buffalo relative to the 5- to 7-year-old buffalo (odds ratio 2.04 versus 2.32; Table 2). It may have been due to bovine tuberculosis mortality in the oldest age group, because higher bovine tuberculosis mortality would have resulted in a lower prevalence and apparent lower risk in buffalo 7 years old. Alternatively, buffalo 7 years old might have had reduced social contact with other herd members, decreasing their risk of infection. Sex was not a significant variable in the risk model for bovine tuberculosis, indicating that male and female buffalo were at equal risk for bovine tuberculosis infection. The increase in prevalence of bovine tuberculosis with age is a common feature of bovine tuberculosis in bison (Fuller 1962), but absence of a predisposition based on sex is apparently unique to buffalo. For example, in red deer (Cervus elaphus), a significantly higher prevalence of bovine tuberculosis occurred in males (Lugton et al. 1998). The hypothesis was made that aggression between male red deer increased their contact and thus increased transmission potential for M. bovis. Zone was not a significant variable after age group and prevalence in the herd were accounted for in the logistic regression model of risk of bovine tuberculosis. We interpreted that to mean that risk of bovine tuberculosis infection in each zone was not different because of environmental factors intrinsic to the zones. Buffalo calves in low-prevalence herds ( 11 %) had a very low probability ( 4 %) of being infected with bovine tuberculosis, but as prevalence in the herd increased, risk of bovine tuberculosis in all age classes increased (Fig. 2). More than one-half of the buffalo in herds with prevalence 41% were likely to be infected with bovine tuberculosis before they reached reproductive maturity at about 3 years of age. Prevalence of bovine tuberculosis rose between 1991 and 1998 in the central and southern zones of KNP because of increases in average prevalence in the herd and total number of herds infected with bovine tuberculosis (Rodwell 2000). Our risk model indicates that as prevalence in the population increases, the proportion of infected calves will increase. Although bovine tuberculosis apparently did not affect fertility of female buffalo or lactation status in KNP from 1991 to 1998, risk of acquiring bovine tuberculosis increased with age, and indications were found that adult buffalo were underrepresented in the infected herds. Because the explicit mortality risk of bovine tuberculosis in buffalo is unknown, the prediction of future effects remains speculative. Longitudinal studies of age-class survivorship will be required to evaluate the potential long-term risks of bovine tuberculosis in the buffalo of KNP. ACKNOWLEDGMENTS This research was funded by Kruger National Park, Onderstepoort Veterinary Institute, University of Pretoria, and scholarships from the Austin Eugene Lyons and Floyd and Mary Schwall Foundations at University of California, Davis. We would like to express our gratitude to the Kruger National Park for allowing us to conduct this research, and Leo Braack and Petri Viljoen for coordination of the project. Sample collection and processing were made possible by the hard work of Roy Bengis, Dewald Keet, Douw Grobler, Colleen Wood, and Rene Cherry. At Decker, JJ van Altena, and section and regional rangers of KNP managed sample collection. Pathology and diagnosis of bovine tuberculosis were conducted by N. Kriek, A. Michel, and the staff at the University of Pretoria and Onderstepoort Veterinary Institute. We also are grateful to P. Kass for his statistical support, and B. Hamilton, T. Carpenter, and M. Johnson for advice on improving the paper. LITERATURE CITED BARLOW, N. 1994. Bovine tuberculosis in New Zealand: epidemiology and models. Trends in Microbiology 2:119 124. BENGIS, R. G. 1999. Tuberculosis in free-ranging mammals. Pp. 101 114 in Zoo and wild animal medicine

238 JOURNAL OF MAMMALOGY Vol. 82, No. 1 (M. E. Fowler and R. E. Miller, eds.). W. B. Saunders Company, Philadelphia, Pennsylvania. BENGIS, R. G., N. P. KRIEK, D. F. KEET, J. RAATH, V. DE VOS, AND H. HUCHZERMEYER. 1996. An outbreak of bovine tuberculosis in a free-living African buffalo (Syncerus caffer Sparrman) population in the Kruger National Park: a preliminary report. Onderstepoort Journal of Veterinary Research 63:15 18. BROTHERS, N. P., I. E. EBERHARD, G.R.COPSON, AND I. J. SKIRA. 1982. Control of rabbits on Macquirie Island by myxomatosis. Australian Wildlife Research 9:477 485. COLLETT, D. 1994. Modeling binary data. 2nd ed. Chapman & Hall, London, United Kingdom. COSIVI, O., F. MESLIN, C. DABORN, AND J. GRANGE. 1995. Epidemiology of Mycobacterium bovis infection in animals and humans, with particular reference to Africa. Revue Scientifique et Technique 14: 733 746. DE VOS V., R. G. BENGIS, AND H. J. COETZEE. 1983. Population control of large mammals in the Kruger National Park. Pp. 213 231 in Management of large mammals in conservation areas (P. N. Owen-Smith, ed.). Haum Educational Publishers, Pretoria, South Africa. FULLER, W. A. 1962. The biology and management of the bison of Wood Buffalo National Park. Canadian Wildlife Service Wildlife Management Bulletin Series 16:1 52. GERTENBACH, W. P. D. 1983. Landscapes of the Kruger National Park. Koedoe 26:9 121. GRIMSDELL, J. J. R. 1973. Age determination of the African buffalo, Syncerus caffer. East African Wildlife Journal 11:31 54. GULLAND, F. 1997. The impact of parasites on wild animal populations. Parassitologia 39:287 291. HOSMER, D. W., AND S. LEMESHOW. 1989. Applied logistic regression. John Wiley & Sons, New York. JOUBERT, S. C. J. 1983. A monitoring program for an extensive national park. Pp. 201 212 in Management of large mammals in African conservation areas (R. N. Owen-Smith, ed.). Haum Publishers, Pretoria, South Africa. KEET, D. F., N. P. KRIEK, H. HUCHZERMEYER, AND R. G. BENGIS. 1994. Advanced tuberculosis in an African buffalo (Syncerus caffer Sparrman). Journal of the South African Veterinary Association 65:79 83. KLOECK, P. 1998. Tuberculosis in domestic animals in areas surrounding the Kruger National Park. Pp. 10 15 in The challenges of managing tuberculosis in wildlife in southern Africa (K. Zunkel, ed.). Mpumalanga Parks Board, Nelspruit, South Africa. LUGTON, I., P. WILSON, R. MORRIS, AND G. NUGENT. 1998. Epidemiology and pathogenesis of Mycobacterium bovis infection of red deer (Cervus elaphus) in New Zealand. New Zealand Veterinary Journal 46:147 156. MILLS, M. G. L., H. C. BIGGS, AND I. J. WHYTE. 1995. The relationship between rainfall, lion predation and population trends in African herbivores. Wildlife Research 22:75 88. MORRIS, R., D. PFEIFFER, AND R. JACKSON. 1994. The epidemiology of Mycobacterium bovis infections. Veterinary Microbiology 40:153 177. O REILLY, L., AND C. DABORN. 1995. The epidemiology of Mycobacterium bovis infections in animals and man: a review. Tubercle and Lung Disease 76: 1 46. PLOWRIGHT, W. 1982. The effects of rinderpest and rinderpest control on wildlife in Africa. Symposia of the Zoological Society of London 50:1 28. RODWELL, T. C. 2000. The epidemiology of bovine tuberculsosis in African buffalo (Syncerus caffer). Ph.D. dissertation, University of California, Davis. SCHMITT, S., ET AL. 1997. Bovine tuberculosis in freeranging white-tailed deer from Michigan. Journal of Wildlife Diseases 33:749 758. SOKAL, R. R., AND F. J. ROHLF. 1995. Biometry: the principles and practice of statistics in biological research. 3rd ed. W. H. Freeman, New York. SPSS INC. 1997. SPSS for Windows version 8.0: standard version. SPSS Inc., Chicago, Illinois. TESSARO, S. V., L. B. FORBES, AND C. TURCOTTE. 1990. A survey of brucellosis and tuberculosis in bison in and around Wood-Buffalo-National-Park, Canada. Canadian Veterinary Journal 31:174 180. VAN RENSBURG, P. J. J., J. D. SKINNER, AND R. J. VAN AARDE. 1987. Effects of feline panleukopenia on the population characteristics of feral cats on Marion Island [Antarctic Ocean]. The Journal of Applied Ecology 24:63 74. WEIGEL, R. M., ET AL. 1995. Prevalence of antibodies to Toxoplasma gondii in swine in Illinois in 1992. Journal of the American Veterinary Medicine Association 206:1747 1751. WHITE, P., AND S. HARRIS. 1995. Bovine tuberculosis in badger (Meles meles) populations in southwest England: the use of a spatial stochastic simulation model to understand the dynamics of the disease. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 349:391 413. WHYTE, I. J. 1996. The management of large buffalo populations. Pp. 1 12 in The African buffalo as a game ranch animal (B. L. Penzhorn, ed.). University of Pretoria, Pretoria, South Africa. Submitted 7 September 1999. Accepted 23 May 2000. Associate Editor was Renn Tumlison.