Making Invasion models useful for decision makers; incorporating uncertainty, knowledge gaps, and decision-making preferences
Uncertainty is inherent in model-based forecasts of ecological invasions. In this chapter, we explore how the perceptions of that uncertainty can be incorporated into the pest risk assessment process. Uncertainty changes a decision maker’s perceptions of risk; therefore, the direct incorporation of uncertainty may provide a more appropriate depiction of risk. Our methodology borrows basic concepts from portfolio valuation theory that were originally developed for the allocation of financial investments under uncertainty. In our case, we treat the model-based estimates of a pest invasion at individual geographical locations as analogous to a set of individual investment asset types that constitute a ‘portfolio.’ We then estimate the highest levels of pest invasion risk by finding the subset of geographical locations with the ‘worst’ combinations of a high likelihood of invasion and/or high uncertainty in the likelihood estimate. We illustrate the technique using a case study that applies a spatial pest transmission model to assess the likelihood that Canadian municipalities will receive invasive forest insects with commercial freight transported via trucks. The approach provides a viable strategy for dealing with the typical lack of knowledge about the behaviour of new invasive species and generally high uncertainty in model based forecasts of ecological invasions. The technique is especially useful for undertaking comparative risk assessments such as identification of geographical hot spots of pest invasion risk in large landscapes, or assessments for multiple species and alternative pest management options.