9 Spatial modelling

Expanding of international trading and passenger volumes from a growing number of countries increases the risk of arrival and establishment of pests and diseases. To reduce the risk to biosecurity, regulators may aim to detect incursions of pests and diseases as early as possible and effectively monitor their movements. Spatial modelling can assist in managing risk by predicting the spread of potential pests and diseases. Risk management options include, for example, surveillance strategies and control and eradication programs for mitigating the ecological, economic and social impacts of pests and diseases.

The number of methods developed for modelling of geographic species distribution is increasing as their application expands as well, requiring the need for choice of method to be matched to application. CEBRA compared the performance of different modelling methods and provided advice on development of criteria for performance evaluation. A workshop with spatial analysis experts hosted by CEBRA focussed on developing methods for predicting spatial distribution and movements of invasive species leading to the emergence of innovative ideas and collaborations relevant for risk/hazard identification.99 In a somewhat related project, CEBRA explored several approaches to developing predictive spatial models for the potential habitat of species of concern, often termed species distribution models. While the project identified that there is no consistent approach within ecology to identifying variables that are likely to be closely associated with the distribution of a species, it developed practical guidance for the process of variable selection.100 A review of spatial models for estimating the likelihood of animal pest occurrence and potential suitable habitat in marine and terrestrial environments formed the basis of further research in these areas.101

The design and implementation of effective surveillance programs is an important component of reducing biosecurity risk post-border. CEBRA developed a simple decision tool based on a spatially explicit bio-economic model for biological invasions. In combination with simple optimisation tools it can help decision makers to find the desired balance for surveillance programs in terms of cost, resources and probability of success.102 To enable regulators to make more informed decisions around surveillance resources, CEBRA recently directed research efforts into developing an approach for producing risk maps along entry pathways. The approach is based on a Bayesian Network for estimating the probability of pest entry and exposure across New Zealand. Custom built software manages models and generates spatially explicit maps of exposure rates for different pests and diseases103 (Box 2).

Box 2: Adoption Highlight

Risk maps for entry pathways in New Zealand

The New Zealand forest industry has rolled out the full system, the software developed by the CEBRA project, and is using it to plan their annual surveillance program. Furthermore, the New Zealand Ministry of Primary Industries will also be using the model to update risk mapping for its High Risk Site Surveillance program for the season 2017-18.


99. Elith, J. & Graham, C. (2008). Do they? How do they? Why do they differ? On finding reasons for differing performances of species distribution models. Australian Centre of Excellence for Risk Analysis, report 0603.

100. Barry, S. et al. (2017). Tools and approaches for invasive species distribution modelling for surveillance. Centre of Excellence for Biosecurity Risk Analysis, report 1402B.

101. Burgman, M. et al. (2014). Spatial models for marine biofouling and post-border response. Centre of Excellence for Biosecurity Risk Analysis, report 1302A.

102. Cacho, O., Hester, S. & Spring, D. (2010). Application of search theory to invasive species control programs. Australian Centre of Excellence for Risk Analysis, report 0806.

103. Mascaro, S. & Thiruvady, D. (2017). Exposure Pathway Model for Forest Surveillance: Stage 2. Centre of Excellence for Biosecurity Risk Analysis, report 1502E.