Habitat suitability and susceptibility modeling for strategic control of invasive Buffel grass, South Australia

Abstract

Invasive plants are a major threat to environmental conservation and are costly to control. To effectively mitigate invasion natural resource managers need to anticipate potential damage, develop policies to prevent introduction as well as mitigate spread. Weed distribution modeling provides managers with the objective information required to strategically direct control efforts. However, often the empirical species distribution data needed to model habitat susceptibility to invasion are limited. For this reason, the benefits of mechanistic models (predictions based on knowledge of species environmental tolerances) are gaining recognition and acceptance. In a recent publication by Smith et al. (2012) a framework for estimating weed invasion potential that utilized expert knowledge of dispersal, establishment and persistence was presented. Here, we construct a model for the contentious weed species, Buffel grass in accordance with the theoretical framework proposed by Smith et al. (2012). This framework distinguishes between habitat suitability and susceptibility. In our study, maps for habitat suitability and susceptibility that incorporate both expert opinion and objective empirical modeling of 2010 Buffel grass roadside survey data are created. Presented are spatially explicit models of introduction pathways, habitat suitability and landscape susceptibility for Buffel grass invasion in the arid zone of South Australia. Results show the relative susceptibility of arid South Australia to Buffel grass invasion. The inclusion of empirical data in this modeling framework presented several challenges, such as the “persistence” indicator, which requires a time component, difficult to quantify empirically. The use of this theoretical framework for spatially explicit modeling requires more thought on how to tackle scale, particularly regarding how the scale of the expert observation lines up with the scale of available environmental data layers, and this is the focus of our discussion.