Bovine TB infection at Karamea – why is TB persisting there?
Young dairy cows on a farm/forest margin at Karamea – Caroline Thomson
The decline of TB-infected cattle and deer herds across New Zealand from 1700 in the mid-1990s to less than 40 in 2015 is a disease management success story. However, in the Karamea district in Westland, reduction in infected herd numbers has proven harder to achieve and progress has been considerably slower than elsewhere.
Despite possum control on the farmland and adjacent forest, combined with a higher than average level of herd TB testing, herd infection has persisted since the early 1970s. Landcare Research, in collaboration with Mark Neill (TBfree New Zealand) looked at both livestock- and possum-related aspects prevalent at Karamea to identify the likely causes of the continuing infection.
Herd characteristics and testing
The number of infected dairy herds at Karamea declined between 1993 and 2011, with the largest declines following aerial control operations against possums in the adjacent forest (suggesting possums were driving the infection). However, these declines were relatively short-lived. Although possum control, along with the test and slaughter programme for livestock, reduced the proportion of infected herds, it was difficult to drive the prevalence below 0.2 (i.e. 20% of herds infected). Most Karamea dairy herds are considered to be 'closed' and do not trade in livestock (i.e. cull animals are sent directly to slaughter). Purchasing stock is rare and where it does occur, most farmers buy from disease-free areas. A survey of farmers found a significant correlation between herd infection status and the use of run-off blocks generally close to forest. However, there was little evidence of a relationship between persistent infection and bush-pasture margins, offal pits, home kill, stock water systems, retention of carry-over cows, or animals sent to slaughter.
Possum distribution, abundance and infection
Possum control has been carried out periodically in the Karamea district since the 1970s and annually since 1996. In the winter of 2008, all forest adjacent to Karamea farmland from Blue Duck Creek in the south to Kohaihai in the north and extending 5–10 km into the forest interior was aerial sown with 1080 bait. In 2009 and 2010, Bruce Warburton and Jackie Whitford distributed chewcards (baited coreflute plastic that possums chew) throughout traditionally infected farms to assess the relative abundance and distribution of possums. Additionally, all possum carcasses retrieved from TBfree New Zealand-funded control operations and from the trapping that followed detection surveys were necropsied and their lymph nodes pooled for culture. In 2011, the possum interference rate on chewcards placed along the forest margin was 2.8% (192/6789), indicating very low possum numbers across all local habitats. Comparing the mean distance between captured possums and between the same number of possums when distributed randomly between all trap sites showed they were significantly aggregated, particularly on forest margins.
Between 2006 and 2011 only four infected possums were identified at Karamea, and all were from a 2006 survey of 249 possums actively screened for TB. Subsequently, a further 386 possums were sampled, but no TB was detected. However, in the winter of 2012 a control operation detected four infected possums (two with visual lesions and two identified by culture) from a sample of 50 taken in the Arapito Valley (Karamea River).
Factors predicted to be driving the continuing infection
Farmers, ecologists, veterinarians, possum trappers, and anti-1080 people all have their preferred explanation of why TB persists at Karamea, ranging from infected immigrant possums, within-herd infection, infected stoats, infected seals, and infected possums in swampy areas not being effectively controlled. However, the major contenders are possum related infection and in-herd infection. To objectively determine the factors driving the infection, Dean Anderson used livestock testing data from 2003–2011 to develop a hierarchical Bayesian statistical model to make inferences on wildlife and livestock factors influencing the probability that an individual cow on a farm is infected at a specific time. Using his model, Dean showed the probability that an individual cow is infected is best predicted by models including the proximity of the farm to forest, and whether the farm had previously been infected. Herd exposure to forest represented the risk that an infected possum would come onto a given farm, and is proportional to the farm area and its proximity to forest.
So what did the team learn about TB infection at Karamea? Herds continue to become infected, some for many years. Although some of this infection might be caused by continuing infection within herds (i.e. in cows not expressing a detectable immune response to skin or blood tests), the trapping of infected possums and model predictions suggest that the perpetuation of infection is mostly influenced by wildlife factors and by the proximity of farms to forest. As long as infected possums remain in the adjacent forest (Kahurangi National Park) there will always be the potential for new TB cases in Karamea livestock. Although possum numbers on Karamea farms are low as a result of control, eradication of the disease will require ongoing effective possum control over large areas of the adjacent forest, maintaining possum control on the farmland, and continuing herd testing.
This work was funded by TBfree New Zealand