Bovine tuberculosis in brushtail possums: models, dogma and data

Three different models of bovine tuberculosis (Tb) in brushtail possums were evaluated against their stated purpose, and testable assumptions and predictions evaluated against available data where possible. Not surprisingly, two of the models may be falsified based on currently available data with respect to either important model assumptions or predictions, and the third may suffer from being right for the wrong reason. This does not mean that these models are not useful. To the contrary, I argue that all models, especially those published in the scientific literature have largely addressed their stated purpose, and have contributed to our understanding of and ability to manage bovine tuberculosis infection in brushtail possum populations. No model, however, satisfactorily explains the pronounced spatial clustering of possum Tb, and the models critiqued have provided little strong inference as to the routes of transmission of Tb among possums. This situation is not helped by the scarcity of datasets on Tb in uncontrolled possum populations that are readily available to confront competing possum/Tb models with. As time passes, there is a very real risk that these data sets will be lost. This is of particular concern, as the expansion in the area of New Zealand under active possum management means the future opportunity to collect further data on Tb in uncontrolled possum populations is severely limited.