Dr Dean Anderson and M. Cecilia Latham, researchers in Manaaki Whenua’s Wildlife Ecology and Management team, in collaboration with Dr Nari Williams at Plant & Food Research, and Maori colleagues, have designed and published a method to address this issue for the microscopic soil-borne pathogen Phytophthora agathidicida (PA), the cause of kauri dieback.
They adapted the Proof of Absence framework, a Bayesian statistical model developed to guide surveillance efforts with the aim of proving freedom from TB in wildlife, to the problem of planning surveillance for a microscopic plant pathogen.
PA surveillance is largely done by taking and lab-testing soil samples, as kauri trees can develop kauri dieback-like symptoms from multiple factors. On the other hand, if PA is not detected during surveillance, the question then becomes how confident can we be that PA is not in that area given that absence of evidence is not evidence of absence?
It was important to investigate how Proof of Absence modelling can help answer two main questions that communities face in the management of kauri forests:
- If PA and associated disease symptoms are detected in a surveyed area, the immediate management priority is one of containment: to delimit the disease front where PA is present, and apply actions that limit future spread of the pathogen to uninfected neighbouring sites.
- If PA is not detected, then it is important to determine the surveillance effort required to be confident that PA is absent from a high-risk zone in the forest.
Part of the modelling exercise included developing a risk of pathogen presence map which is based on relationships between the likelihood of PA presence and five environmental variables that are known to influence pathogen ecology. Importantly, it does not preclude the inclusion of other factors considered to be of local importance.
The researchers say that the mapping method is simple and easily transferable between locations and different pathogens of cultural and ecological significance. It should prove to be an effective tool to target trees for surveillance, monitoring and protection, and for local communities and rangers to plan their on-the-ground activities, for example ensuring that monitoring is done exclusively within high risk or low risk areas, but not crossing between levels of risk without the proper hygiene measures.
Results from Proof of Absence analyses showed that many samples are needed to achieve 95% confidence in pathogen absence, but this varied between forests and sectors of each forest. However, these analyses are valuable to guide field efforts as well as to ensure realistic expectations among practitioners.
