A descriptive and epizootiologic study of Brucellosis and Tuberculosis in bison in Northern Canada

Abstract

Studies were conducted on bovine brucellosis (Brucella abortus) and tuberculosis (Mycobacterium bovis) in hybrid bison and other wildlife species in and around Wood Buffalo National Park (WBNP), and in wood bison in the Mackenzie Bison Sanctuary. Aerial surveys provided total counts of 4,661 bison in March, 1984, and 4,582 bison in March, 1985, in WBNP. Compared to previous surveys, these numbers indicate that the bison population is declining. Surveys showed that bison leave the park on at least three corridors, including the southwest corner of the park adjacent to the Fort Vermilion, Alberta, agricultural zone. During the same period, a survey of the east side of the Slave River Lowlands found 370 hybrid bison. This population also has declined.

The complete or partial remains of 164 bison were found in and around WBNP between June, 1983 and August, 1985. Of the 72 carcasses suitable for analyses, 18 (25%) had evidence of brucellosis and 15 (21%) had tuberculosis, with two of these cases having concurrent infections. B. abortus biotypes 1 and 2, including a urease-negative strain of biotype 1, were isolated from 11 bison and 7 others had antibody titres to B. abortus. Tissues of 7 bison were culture-positive for M. bovis and 8 other bison had granulomatous lesions but the bacterium was not isolated. The prevalence estimates of the two diseases in the sample were conservative because of the limitations imposed by opportunistic sampling. The 95% confidence intervals for the prevalence of the diseases in the bison population in and around WBNP were 15.54 36.60% for brucellosis and 12.16 - 32.02% for tuberculosis based on this sample. The diseases were found in bison outside of the park near the three recognized corridors of trans-boundary movement.

Infection with B. abortus was associated with severe arthritis in three bison and a subcutaneous abscess in a fourth bison. Only 9 (33%) of the 27 adult female bison in the collection were pregnant. Lesions produced by M. bovis infection in bison were similar to those reported in cattle, ranging from focal granulomatous lymphadenitis to generalized, multisystemic disease. Three of four bison killed by wolves in WBNP had extensive tuberculosis, suggesting an association between the disease and predation. Hunters killed, butchered and utilized 56 of the bison in the sample; 15 (27%) of these bison had brucellosis and 9 (16%) others had tuberculosis, indicating a risk of disease transmission from bison to people. Other pathologic conditions found in the bison were documented.

Postmortem examination, histology and bacteriology did not detect the presence of brucellosis or tuberculosis in 30 wood bison killed in the Mackenzie Bison Sanctuary. Serological tests on these and an additional 21 wood bison from the same location were all negative for antibodies to B. abortus. Statistical analyses and epidemiological information indicate that this population of wood bison is free of the two diseases. Because this population is near WBNP and because it is expanding both numerically and geographically, the wood bison of the Mackenzie Bison Sanctuary are at risk of contracting brucellosis and tuberculosis from the infected hybrid bison in and around WBNP. Fifteen of the 20 mature female wood bison were pregnant; this was a significantly (a = .01) higher pregnancy rate than that observed in the hybrid bison sample from WBNP.

Other wildlife species in and around WBNP were surveyed for the presence of brucellosis and tuberculosis. The 808 animals examined included rodents, shrews, red foxes, wolves, fishers, black bears, marten, lynx, moose, woodland caribou, coyotes, wolverines, a woodchuck, a red squirrel and a raven. B. abortus was isolated from tissues of 4 of 13 wolves, 1 of 37 red foxes, and 1 of 5 moose. All isolates were biotype 1 and one of these, from a wolf, was the urease-negative strain. No pathology was associated with the infections in the wolves or the fox, but the moose was severely emaciated, and had suppurative pleuritis and peritonitis and abscessed lymph nodes. This concurs with previous reports of brucellosis in moose which suggest that it is a severe, usually fatal disease in that host. There was no indication of tuberculosis in any of the wildlife other than bison. Other pathologic conditions seen in the various wildlife species were documented.

Two herds of cattle adjacent to WBNP and the Mackenzie Bison Sanctuary were tested for brucellosis. None of the 98 cattle at Mills Lake nor the 42 cattle on Ryan Island were positive on the buffered plate antigen test. This indicated that the cattle were not a source of B. abortus for uninfected wood bison, nor had they become infected through potential contact with diseased hybrid bison.

Arsenic, cadmium, copper, lead, manganese, mercury, selenium and zinc levels were evaluated and compared in liver and kidney samples from 16 hybrid bison in WBNP and 20 wood bison from the Mackenzie Bison Sanctuary. None of the levels of metals in bison tissues were deficient when compared to values from cattle. However, selenium levels in both wood bison and hybrid bison, and manganese levels in hybrid bison, would be considered marginal for cattle.

Arsenic, cadmium, lead and mercury levels were 50- to 100-fold below those considered to be toxic in livestock.

Strains of B. abortus from bison in WBNP had oxidative metabolic profiles typical of strains isolated from cattle. The patterns of lysis by Tblisi, Firenzi, Weybridge, R/C, R/0 and R varieties of Brucella-phages did not differentiate strains of B. abortus from livestock, bison, wolves or a fox. Susceptibility to ampicillin, cephalothin, chloramphenicol, colistin, gentamycin, nalidixic acid, polymixin-B, streptomycin and tetracycline was similar among strains of B. abortus from wildlife and livestock. However, B. abortus biotype 2 from livestock was susceptible to erythromycin whereas biotype 2 from a bison was not. No naturally occurring Brucella-phages or plasmids were found in strains of B. abortus from wildlife in and around WBNP.

An attempt was made to evaluate calf production and yearling recruitment into the bison population of WBNP by conducting herd segregation counts on foot in the park. Sample sizes were small compared to the total population, but did suggest that either calf production or early survival was poor, and that yearling recruitment was very poor. These data are appended to the thesis.