Tools and approaches for invasive species distribution modelling for surveillance
Project ID: 1402B
CEBRA Project Leader: Mark Burgman
DA Sponsor: Louise van Meurs, Kim Ritman and Sally Troy
DA Project Leader: Enrico Perotti, Ranjith Subasinghe and Bart Rossel
DA Division: Plant Division
MPI Project Manager: Steve Hathaway
Collaborators: CSIRO; ABARES; AgResearch; Scion
This project aims to deliver something currently missing in existing overviews of methods for predicting distributions of species of biosecurity concern. Overviews usually either describe methods, citing commonly accepted views of their performance, or promote one method, often without clear proof of its generality. This makes it difficult for biosecurity agencies to make informed decisions about which methods to use. By bringing together a team with strong statistical and modelling experience, and with in biosecurity applications, we will target a subset of issues that will help clarify the evidence for existing assertions and develop methods for missing pieces. Broadly speaking, this project will explore alternative approaches to habitat suitablility modelling and will develop structured guidelines and protocols to assist managers to identify the most appropriate tools and approaches for specific applications. It will assess the need to map habitat suitability and the distribution of hosts. The project will also consider the implications of dealing with many different possible pests and diseases versus focussing on specific pests. Defined protocols for selecting and routinely using existing habitat suitability mapping tools will be provided.
Finally, the project report will discuss how the outputs of appropriate spatial models may be combined with other relevant data on the costs of surveillance, exposure pathway information, the potential costs of incursions (which will also be informed by habitat suitability mapping), and the potential for effective intervention and control, to design a framework for an effective and cost-efficient surveillance system applicable to all these four contexts. It will evaluate the role of existing and ongoing suveillance data, for mapping likelihood of arrival and establishment. Importantly, it will not deliver these outcomes. Rather, it will discuss the role of spatial models in these systems.