Modelling the efficacy of a threshold-triggered control strategy for the invasive tree species Prosopis

Abstract

Invasive biological species threaten biodiversity and economies worldwide, often resulting in devastating environmental and socio-economic effects. As a result, there has been a growing realisation of the importance of understanding and effectively managing the complex interactions that arise from biological invasions. A prerequisite for launching such a management strategy, however, is an effective means of predicting the expected effects of the strategy over time. To this end, a spatio-temporal model is proposed in this paper for predicting the extent to which a particularly problematic invasive plant species, Prosopis, spreads when confronted with a threshold-triggered control method. In particular, a spatial analysis is advocated during which the study region is discretised and a data set constructed. Thereafter, the power of machine learning algorithms is leveraged to predict habitat suitability for Prosopis. Finally, a cellular automaton is adopted to simulate the spatial spread and temporal growth of Prosopis within the discretised study region. The practicality of our modelling approach is illustrated by means of a real-world case study in the Northern Cape region of South Africa.