Wednesday 3 December 2003
Managing tuberculosis in wildlife
Lecture Room C1
8:00am – 12:10pm
Wednesday 3 December: morning (Lecture room C1)
Managing tuberculosis in wildlife
Chairs: Graham Nugent, Richard Delahay and Frank Griffin
0800 – 0820 CALEY, P.; Hone, J. TB host status in wildlife: theory and practical implications.
0820 – 0840 NUGENT, G.; Griffin, F.; Hickling, G.J. Bovine tuberculosis: Host status of wild and
0840 – 0900 VICENTE, J., Höfle, U.; Garrido, J.; Barral, M.; Juste, R.; Gortazar, C. Bovine TB in
the European wild boar: dead-end host or reservoir?
0900 – 0920 KALEMA-ZIKUSOKA, G.; Kennedy-Stoskopf, S.; Michel, A.L.; Bengis, R.G.;
Woodford, M.; Adatu, F.; Levine, J.F. Tuberculosis survey in buffalo, cattle and
humans at the interface of Queen Elizabeth NP, Uganda: Implications for public health
0920 – 0940 Barlow, N.D.; KEAN, J.M. Modelling the persistence of wildlife diseases: the case of
0940 – 1000 CLIFTON-HADLEY, R.S.; Delahay, R.J. The epidemiological characteristics of
Mycobacterium bovis infection in badgers.
1000 – 1030 Morning tea
1030 – 1050 WHITE, P.C.L.; Bulling, M.T.; Garland, L.; Harris, S. Bovine tuberculosis in Britain:
integrating spatial modelling, GIS and economic approaches in the sustainable
management of the disease.
1050 – 1110 GORMLEY, E.; Lesellier, S.; Costello, E.; Sleeman, P.; Corner, L.L.A. Development
of a wildlife vaccination strategy for eradication of tuberculosis in cattle in Ireland.
1110 – 1130 DELIBERTO, T.J.; Smith, H.J.; Stevenson, J.S. Evaluation of coyotes as sentinel
animals for predicting the presence of bovine tuberculosis in Michigan, United States.
1130 – 1150 DE LISLE, G.W.; Yates, G.F.; Collins, D.M. Overview of DNA fingerprinting of
Mycobacterium bovis applied to the epidemiology of tuberculosis in wildlife in New
1150 – 1210 RAMSEY, D.; Nugent, G. Optimising surveillance strategies for detecting Tb in
CALEY, Peter and Jim HONE
Landcare Research, Private Bag 11052, Palmerston North, New Zealand and CSIRO Division of Entomology, Canberra ACT 2601 Australia (PC); Applied Ecology Research Group, University of Canberra, ACT 2601 Australia (JH).
TB HOST STATUS IN WILDLIFE: THEORY AND PRACTICAL IMPLICATIONS
Determining the host status of wildlife for infectious diseases involves several approaches with important practical implications. Observational studies can detect disease and estimate prevalence and other parameters. In single host systems, model parameters can be estimated reasonably easily, such as the basic reproductive rate (Ro), which have management significance. Experimental studies can address practical issues, such as disease responses to manipulation of host density (e.g., culling) and susceptible host density (eg vaccination). Experimental studies give stronger inference when they use randomisation, replication and controls, however few studies do so. Questions regarding the host status of species in ecological systems containing multiple hosts are not adequately answered by conventional observational studies. A combination of observation, experimentation and modelling is required. We illustrate this for the New Zealand wildlife/Tb system with reference to feral ferrets. Observational studies show that Tb infection in ferrets can be highly prevalent and -1) varied across five sites from 0.09 in females to is likely acquired by ingestion. The force of infection (; yr
7.90 in males. Experimental studies demonstrate disease prevalence is reduced when culling of an alternative host (brushtail possums) occurs. Modelling estimates the basic reproductive rate (Ro) of Tb in ferrets to vary from <1 (spillover hosts) to >1 (maintenance hosts) across several sites. At the former sites, ferret control is not necessary, but at the latter sites it is necessary, to reduce TB prevalence in ferrets. It is suggested that the approach used in the TB ferret studies be explored for other host-disease topics.
NUGENT, Graham, Frank GRIFFIN and Graham J. HICKLING
Landcare Research, P.O. Box 69, Lincoln 8152, New Zealand (GN); Dept. of Microbiology, University of Otago, Otago University, P.O. Box 56, Dunedin (FG); Department of Fisheries and Wildlife, Michigan State University, East Lansing, Michigan 48824-1222 (GJH).
BOVINE TUBERCULOSIS: HOST STATUS OF WILD AND FARMED DEER
There are increasing indications that wild deer in New Zealand are spillover hosts for bovine tuberculosis (Tb). There is little evidence of deer-to-deer transmission, and experimental data showing that most infection in wild deer derives from some interaction with infected brush-tailed possums (Trichosurus vulpecula). In
contrast, there is ample evidence of deer-to-deer transmission on farms in New Zealand, and this disease is also self-sustaining in wild deer in Michigan, USA. In this paper we synthesize data on age-specific rates of infection in wild red deer (Cervus elaphus scoticus) in New Zealand, both a high and low possum densities,
and compare that with similar data from Tb-infected red deer farms in New Zealand, and from wild white-tailed deer (Odocoileus virginianus) in Michigan. Few wild deer in New Zealand become infected before
about 10 months age, whereas farmed deer can become infected much earlier. The prevalence of Tb generally increases with age in wild deer in both New Zealand and Michigan, indicating more-or-less constant force of infection and low rates of Tb-induced mortality. In farmed deer regular testing and removal of infected deer prevents this accumulation of infection. We compare densities of deer in these three situations, and attempt to deduce the density or range of densities (in association with other factors) below which deer cease to be true maintenance hosts for Tb.
VICENTE, Joaquin, ursula HÖFLE, Joseba GARRIDO, Marta BARRAL, Ramón JUSTE and
Instituto de Investigación en Recursos Cinegéticos IREC (CSIC-UCLM). P.O. Box 535 E13080 Ciudad Real, Spain. (JV, UH, CG); Department of Animal Health, Instituto Vasco de Investigación y Desarrollo Agrario NEIKER, Berreaga 1, 48160 Derio, Bizkaia, Spain (JG, MB, RJ).
BOVINE TB IN THE EUROPEAN WILD BOAR: DEAD-END HOST OR RESERVOIR?
The causative agent of bovine tuberculosis (bTB) is Mycobacterium bovis. Worldwide, wildlife plays an
important role in the epidemiology of bTB. In Europe, the most well known case is that of the badger in the British islands, but bTB cases have repeatedly been reported in wild and semidomestic cervids, and even in wild boars. Due to the low frequency of detection of generalized tuberculosis in M. bovis infected European
wild boars, and to the fact that these were only detected in areas with infected cattle, authors hypothesized that wild boars are the end host for M. bovis infection, with little transmision possibilities to lifestock and little relevance as a reservoir. Here we present data that suggest a more relevant role of the wild boar in bTB epidemiology. The prevalence of wild boars with bTB-compatible lesions in Spain is high 43,4 ；1,6% (n=
414 of 954 hunter harvested wild boars). Most cases are found in the south-western part of the country, where local prevalences of up to 87% have been recorded. Some wild boar populations are found to maintain bTB infection without contact with deer nor domestic lifestock. Lesions were consistently found in mandibular lymph nodes (98,8% of animals with gross lesions), but 32 (7.76 %) wild boars had generalized lesions, suggesting clinical disease and increased excretion. Risk factors identified to date include management tools such as fencing, translocations and feeding.
KALEMA-ZIKUSOKA, Gladys, Suzanne KENNEDY-STOSKOPF, Anita L. MICHEL, Roy G.
BENGIS, Michael WOODFORD, Francis ADATU and Jay F. LEVINE
Department of Farm Animal Health and Resource Management, North Carolina State University, 4700 Hillsborough Street, Raleigh, NC 27606 (GKZ, SKS, JL); Hanes Veterinary Medical Center, 4401 Zoo Parkway, Asheboro, North Carolina 27203 (GKZ); Onderstepoort Veterinary Institute, P. Bag X5, Onderstepoort 0110, South Africa (GKZ, ALM); Directorate of Animal Health, P.O. Box 12, Skukuza 1350, South Africa (GKZ, RGB); World Organization for Animal Health, Portugal (GKZ, MW); Uganda National Tuberculosis/Leprosy Programme, P.O. Box 16041, Wandegeya, Kampala, Uganda (GKZ, FA).
TUBERCULOSIS SURVEY IN BUFFALO, CATTLE AND HUMANS AT THE INTERFACE OF
QUEEN ELIZABETH NP, UGANDA: IMPLICATIONS FOR PUBLIC HEALTH AND
A tuberculosis (Tb) survey was conducted from 2001 to 2002 at the interface of Queen Elizabeth National Park (QENP), Uganda, to assess the risks of disease transmission between buffalo (Syncerus caffer),
domestic cattle (Bos sp.) and people. Using the gamma interferon test, Mycobacterium bovis was found in
5% of buffalo (n = 42) and 5% of cattle (n = 75). This prevalence in buffalo was lower than that in the interior of the park, where 21.4% had been recorded with Tb in 1997. Thus buffalo may have become maintenance hosts and can potentially transmit M. bovis to cattle. Based on interviews, 29 people (n = 63)
previously had Tb. Using sputum smears, acid-fast bacteria were found in 28% of people (n = 40). The sputum smears identified an additional 8 people with Tb. Almost 50% of people consumed unboiled milk, putting them at risk of getting M. bovis from cattle. Tb control in both buffalo and cattle may be needed to reduce the risks of M. bovis transmission to people. Tb control in buffalo, especially in the interior, is also needed to prevent spillover to previously recorded species, such as warthogs (Phacochoerus aethiopicus).
BARLOW, Nigel D. and John M. KEAN
AgResearch, P.O. Box 60, Lincoln, New Zealand.
MODELLING THE PERSISTENCE OF WILDLIFE DISEASES: THE CASE OF BOVINE TB
An outstanding question in the ecology of wildlife diseases is the reconciliation of observations on disease prevalence, impact on host abundance, and persistence of the disease in the face of control. The usual approach to modelling diseases uses simple Anderson/May-type deterministic models, and this approach has proven value in terms of simplicity and clarity. However, in the case of bovine Tb in possums such simple models could not reproduce the above features of the disease. A simple modification, making transmission non-linear, seemed to overcome the problem but left two questions unanswered. Firstly, what is the mechanistic basis for such a transmission term, and how does this mechanism account for the ability of a disease with low prevalence and little impact on the host, to display quite rapid recovery after a cull of the host and persistence in the face of sustained culling? Secondly, are there any alternative Anderson/May model variants that give similarly realistic results? To address the first question we developed spatially explicit cellular automaton model equivalents to the simple Anderson/May model, and found that they did give the same effect as non-linear transmission in the simple model and accounted for observed disease behaviour, given heterogeneity in possum densities such that Tb ‘hotspots’ were associated with patches of high possum density. In answer to the second question, we found that heterogeneity in individual susceptibility between possums also reconciled observations and reproduced realistic disease behaviour. We have therefore identified two alternative, and not mutually exclusive mechanisms for persistence of bovine Tb in possums subject to control. This system may provide relevant lessons for modelling bovine Tb in other hosts.
CLIFTON-HADLEY, Richard S. and Richard J. DELAHAY
Veterinary Laboratories Agency, Weybridge, Surrey, UK (RSC-H); Central Science Laboratory, Sand (RJD). Hutton, York YO41 1LZ, UK
THE EPIDEMIOLOGICAL CHARACTERISTICS OF MYCOBACTERIUM BOVIS INFECTION IN
In the UK the European badger (Meles meles) is implicated in the transmission of Mycobacterium bovis to
cattle. Understanding the underlying causes of the distribution of bovine TB in badgers is important in the context of effective management of the disease in cattle. The epidemiological characteristics of M. bovis
infection in badgers has been studied using data from a long term epidemiological and ecological study of a wild population of badgers in South West England.
WHITE, Piran C.L., Mark T. BULLING, Lincoln GARLAND and Stephen HARRIS
Environment Department, University of York, Heslington, York YO10 5DD (PCLW, MTB); School of Biological Sciences, University of Bristol, Woodland Road, Bristol BS8 1UG (LG, SH).
BOVINE TUBERCULOSIS IN BRITAIN: INTEGRATING SPATIAL MODELLING, GIS AND
ECONOMIC APPROACHES IN THE SUSTAINABLE MANAGEMENT OF THE DISEASE
Bovine tuberculosis (TB) is a disease of livestock with significant wildlife reservoirs in many parts of the world. In Britain, the principal wildlife reservoir is believed to be the Eurasian badger (Meles meles). Despite
badger population control since 1975, TB in cattle is now greater than at any time since the 1950s. The badger-TB interaction in Britain is an example of a disease-host system that shows a high degree of spatial heterogeneity. Understanding the factors contributing to this heterogeneity is of key importance in understanding the persistence of the disease. We have developed a GIS-based spatial stochastic model to replicate real badger populations in specific landscapes. Comparisons of the results of this model with homogeneous spatial models highlight the importance of heterogeneity in determining disease persistence, and also allow the design and assessment of targeted management strategies for specific locations. Long-term badger control strategies need to be sustainable, cost-effective and publicly acceptable. We use our model to show how the economic efficiency of control, measured in terms of the total benefits and the benefit:cost ratio, is critically dependent on the timing of control and the level of TB infection present in the badger population in a specific area. These results suggest that, for wildlife diseases that show marked heterogeneity, control strategies should employ a flexible approach at the landscape level, with control being adapted locally to specific situations.
GORMLEY, Eamonn, Sandrine LESELLIER, Eamon COSTELLO, Paddy SLEEMAN and Leigh L.
Dept. of Large Animal Clinical Studies, Faculty of Veterinary Medicine, University College Dublin, Ireland (EG, SL, LC), Central Veterinary Research Laboratories, Abbotstown, Dublin (EC); Dept. of Zoology, NUI Cork (PS).
DEVELOPMENT OF A WILDLIFE VACCINATION STRATEGY FOR ERADICATION OF
TUBERCULOSIS IN CATTLE IN IRELAND
Wildlife species, such as badgers, act as maintenance hosts for Mycobacterium bovis and contribute to the
spread and persistence of tuberculosis in associated cattle populations in Ireland. In areas in which there is a Tb problem affecting a number of herds, the involvement of infected badgers in the introduction of M. bovis
infection into herds act as a constraint to eradication of the disease. The need to devise and implement an effective scheme for the control of Tb in badgers that may contact cattle, is a prerequisite if eradication of Tb from the cattle population is to be achieved. Targeted vaccination of badgers against Tb is an option that if successfully employed, could directly facilitate the advancement of bovine tuberculosis eradication in affected areas. The development of a vaccine against Tb for use in the badger population has been initiated. An M. bovis infection model has been developed and studies have been carried out to evaluate the immune responses of captive badgers to M. bovis infection and M. bovis BCG vaccination. As a long-term control
measure, the aim of vaccination will be to decrease the incidence of infection of susceptible hosts or reduce development of disease pathology in infected individuals to the extent that the disease will eventually be eliminated in badger populations. In conjunction with improvements in sensitivity and specificity of the diagnostic tests used in cattle, this could directly facilitate the completion of bovine tuberculosis elimination in areas so affected.
DELIBERTO, Thomas J., Holly J. SMITH and Justin S. STEVENSON
USDA/APHIS/WS/National Wildlife Research Center, 4101 LaPorte Avenue, Fort Collins, Colorado 80521 (TJD); 2957 Brodie Road, South Branch, Michigan 48761 (HJS, JSS).
EVALUATION OF COYOTES AS SENTINEL ANIMALS FOR PREDICTING THE PRESENCE
OF BOVINE TUBERCULOSIS IN MICHIGAN, UNITED STATES
In 1994, bovine tuberculosis (TB) was discovered in Michigan white-tailed deer (Odocoileus virginianus)
and epidemiological evidence suggests that deer have subsequently transmitted the disease to cattle. Recently, Michigan has been successful at reducing TB prevalence in deer by decreasing population size via increased hunter harvest. However, as deer populations and TB prevalence decrease, the ability to monitor prevalence requires an increased sampling effort, which is socially, politically, and logistically difficult. Therefore, the value of a sentinel species to supplement deer surveillance becomes increasingly important. An effectual TB sentinel species is one that has a detectable level of the disease in its population, acquires the disease primarily through contact with the reservoir of interest, and has a documented home range such that reasonable inferences on the point of infection in the sentinel species and its relationship to the disease in the reservoir can be made. The coyote (Canis latrans) is a suitable sentinel species to monitor TB prevalence in
Michigan deer. Approximately 4-6% of coyotes in northern Michigan have TB, and the primary route of exposure appears to be through the consumption of contaminated food. Data from radio collared coyotes 22 and 65 km for females and males, illustrated average home range sizes of juveniles to be 42 km22 respectively. Comparatively, home range sizes for adults averaged 56 km and 71 kmfor females and males,
respectively. Accordingly, an individual coyote with TB was likely to acquire the infection through contact 2with infected deer within a 70 km area of its location.
DE LISLE, Geoffrey W., Gary F. YATES and Desmond M. COLLINS
AgResearch, Wallaceville Animal Research Centre, P.O. Box 40-063, Upper Hutt, New Zealand.
OVERVIEW OF DNA FINGERPRINTING OF MYCOBACTERIUM BOVIS APPLIED TO THE
EPIDEMIOLOGY OF TUBERCULOSIS IN WILDLIFE IN NEW ZEALAND
Failure to eradicate bovine tuberculosis from domestic animals in New Zealand is due to continued re-infection of cattle and farmed deer herds by wildlife. While Mycobacterium bovis has been isolated from
many different species of wildlife in New Zealand, they are not equally important as a reservoir of infection or for spreading disease to domestic animals. DNA fingerprinting using restriction endonuclease analysis has been applied for over 20 years to investigate the epidemiology of M. bovis infections in New Zealand. More
than 2500 isolates of M. bovis have been examined from domestic animals and wildlife, a third of these
isolates being from wildlife. DNA fingerprinting has demonstrated the presence of geographically distinct populations of infected wildlife. Where wildlife is infected with M. bovis, the same DNA type is present in
different host species, including domestic animals. This observation highlights the spread of infection between hosts and, when combined with other information, provides strong evidence of frequent spread from wildlife to domestic animals. Wildlife is now the principal source of infection for cattle and farmed deer because spread between domestic animals has been markedly reduced by a test and cull control programme. Over the last twenty years, DNA fingerprinting has been able to follow the spread of M. bovis in wildlife
populations and the emergence of new areas of wildlife infection, providing insight into the difficulties of controlling bovine tuberculosis in New Zealand. The full value of DNA fingerprinting comes from utilising this information in conjunction with the results from wildlife surveys and the testing of cattle and farmed deer for bovine tuberculosis.
RAMSEY, Dave and Graham NUGENT
Landcare Research, Private Bag 11052, Palmerston North, New Zealand (DR); Landcare Research, P.O. Box 69, Lincoln, New Zealand (GN).
OPTIMISING SURVEILLANCE STRATEGIES FOR DETECTING TB IN WILDLIFE
Eliminating bovine tuberculosis (Tb) from domestic deer and cattle herds requires accurate knowledge of where the disease exists in both livestock and in wildlife. Although Tb-testing of livestock provides an indicator of disease presence in farmed areas, it cannot do so in unfarmed areas. Also, the occurrence of infection in livestock may lag behind its occurrence wildlife. We aimed to improve Tb surveillance strategies in New Zealand by using two wildlife species, feral ferrets (Mustela furo) and feral pigs (Sus scrofa) as
sentinels to provide cheaper and timely information on Tb distribution. The utility of any host species as a sentinel depends not only how readily it becomes infected, but also on how long and how easily that infection can be detected, and the cost of surveying that species for Tb. To optimise the cost-effectiveness of different sampling strategies for feral ferrets we used a spatially explicit model to simulate the capture of animals in traps. Systematic coverage was the most cost effective strategy as it had a consistently higher Tb -2 set for at least detection probability than other spatial trap placement designs. Trap densities of at least 4 km20 nights had a 95% probability of detecting at least 1 Tb ferret at a prevalence of 2% or greater, regardless of ferret density or ferret spatial distribution. We also developed a new surveillance strategy, the release of radio-collared Tb-free feral pigs, as a way of obtaining reliable information from areas with few other sentinels available. The use of released animals greatly increases the reliability of information about the location and timing of transmission events.