Aspects of eradicating disease from wildlife vectors

Bovine tuberculosis (Tb) occurs in cattle and in wildlife, with various species infected in a range of countries. The control or eradication of Tb can be empirically based or based on understanding of disease processes such as described in mathematical models. The eradication of Tb from Australia was largely empirically based, with water buffalo (Bubalus bubalis) apparently being the only significant wildlife disease host. Buffalo populations were apparently greatly reduced to achieve Tb eradication. This paper examines the alternative approach of basing control or eradication on mathematical models, and compares some predictions of one-host (wildlife) and two-host (cattle and wildlife) disease models with empirical data on Tb in both cattle and in wildlife. The one-host and two-host models were based on those published and also derived. For example, reported positive regressions between prevalence’s of Tb in cattle and in brushtail possums in New Zealand and also between prevalence’s of Tb in red deer and in brushtail possums in New Zealand, are consistent with two-host disease models assuming frequency dependent transmission. However most models of Tb in wildlife, e.g. the Barlow models, assume density dependent transmission. The implications for disease eradication are that models assuming density dependent transmission have a threshold host (wildlife) density about zero. Hence, if frequency dependent, not density dependent, transmission actually occurs, then host (wildlife) density would have to be reduced much lower, in order for eradication of Tb from cattle.