Surveillance of Coyotes to Detect Bovine Tuberculosis, Michigan

Kurt C. VerCauteren; Todd C. Atwood; Thomas J. DeLiberto; Holly J. Smith; Justin S. Stevenson; Bruce V. Thomsen; Thomas Gidlewski; Janet Payeur


Emerging Infectious Diseases. 2008;14(12):1862-1869. 

In This Article

Abstract and Introduction


Bovine tuberculosis (TB) is endemic in white-tailed deer (Odocoileus virginianus) in the northeastern portion of Michigan's Lower Peninsula. Bovine TB in deer and cattle has created immense financial consequences for the livestock industry and hunting public. Surveillance identified coyotes (Canis latrans) as potential bio-accumulators of Mycobacterium bovis, a finding that generated interest in their potential to serve as sentinels for monitoring disease risk. We sampled 175 coyotes in the bovine TB–endemic area. Fifty-eight tested positive, and infection prevalence by county ranged from 19% to 52% (statistical mean 33%, SE 0.07). By contrast, prevalence in deer (n = 3,817) was lower (i.e., 1.49%; Mann-Whitney U4,4 = 14, p<0.001). By focusing on coyotes rather than deer, we sampled 97% fewer individuals and increased the likelihood of detecting M. bovis by 40%. As a result of reduced sampling intensity, sentinel coyote surveys have the potential to be practical indicators of M. bovis presence in wildlife and livestock.


The emergence and reemergence of zoonotic diseases are becoming increasingly important issues for numerous reasons, including deforestation and habitat fragmentation, increased globalization of travel and trade, urbanization, and bioterrorism concerns. Diseases such as severe acute respiratory syndrome (SARS), avian influenza, transmissible spongiform encephalopathies, Rift Valley fever, West Nile disease, anthrax, and Escherichia coli O157 infections recently have resulted in major public health and economic concerns, as well as public anxiety. Over 60% of the 1,415 known human pathogens and 75% of the 175 emerging pathogens are zoonotic.[1] Many emerging diseases have spilled over from wildlife directly (e.g., West Nile virus infection, hantavirus infection, and Lyme disease) or indirectly through domestic or peridomestic species (e.g., avian influenza, SARS, and Nipah virus infections, plague).[2] Early detection of new disease outbreaks in domestic and wild animals is an essential prerequisite of disease control and eradication. Development of methods for early detection of diseases in free-ranging wildlife is problematic.

Development of practical strategies for conducting surveillance in free-ranging wildlife to detect and monitor disease and evaluate control efforts is a necessary component of predicting and managing emerging zoonoses. A case in point is bovine tuberculosis (TB). Mycobacterium bovis, the bacterial pathogen that causes bovine TB, has been identified in wildlife, domestic animals, and humans.[3–6] Transmission of M. bovis may occur through ingestion of infected tissues or, less likely, through inhalation of aerosolized bacilli;[7] typically, granulomatous lesions develop in the thoracic lymph nodes and lung after aerosol exposure, and granulomatous lesions develop in the abdominal lymph nodes after oral exposure. Bovine TB often progresses slowly, and clinical symptoms may not appear until advanced stages are reached.[8,9] In 1995, M. bovis was found in free-ranging white-tailed deer (Odocoileus virginianus) in a localized area in the northeastern Lower Peninsula of Michigan.[10] In subsequent years, a reemergence of M. bovis in Michigan cattle was detected; deer were postulated to be the source of infection. Because the socioeconomic impact of this discovery has been immense,[11] a strategy was developed and implemented to monitor and eradicate M. bovis from wildlife and cattle. Although the strategy successfully reduced the apparent prevalence of M. bovis in deer, the disease still persists at low levels (e.g., 2001–2006 statistical mean 2.3%) because of high deer densities (statistical mean 13/km2) and spatiotemporal crowding resulting from supplemental feeding.[12] As prevalence of M. bovis in deer decreases, the sample size required to detect positive deer increases, making monitoring of the disease in deer more difficult and costly. Eventually, prevalence in deer may become too low to accurately estimate through current methods because of the difficulty and expense of obtaining a sufficient sample size, and consequent difficulty of verifying disease eradication. We hypothesized that the presence of M. bovis in wild deer at low prevalence could be more accurately determined through an indirect estimator (i.e., a sentinel species).

Use of sentinel animals has been suggested as a cost-effective way to infer prevalence in host populations when direct estimation in such populations is difficult.[13] As facultative scavengers, coyotes (Canis latrans) may act as biological sensors and bio-accumulators of M. bovis. by consuming infected host material, resulting in high rates of infection. Furthermore, social foraging by coyote populations[14,15] should increase the likelihood of multiple coyotes ingesting infected tissue from the same M. bovis–positive deer. As a logical corollary, the increased numeric exposure of coyotes to M. bovis should mediate an increased detection probability relative to sampling effort. Support for this hypothesis was provided by research,[5] which reported an apparent prevalence of M. bovis. in opportunistically sampled coyotes as 4% in the general area where apparent prevalence in deer averaged 2.3% from 1995 through 2001.[16] Finally, coyote home-range sizes (statistical mean 14.25 km2, 95% confidence interval [CI] 9.54–18.96 km2) in Michigan allow for reasonable estimates of where infection was acquired.[17]

We report on a sentinel-based surveillance program designed to detect M. bovis in coyotes. Specifically, we sought to determine whether 1) M. bovis occurrence in coyotes was detectable, given reduced sampling intensity relative to white-tailed deer, and 2) prevalence of M. bovis was greater in coyotes than deer for a given area. If so, coyotes should be effectual sentinels of M. bovis occurrence in free-ranging white-tailed deer.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.