Inferring the potential spread of Xylella fastidiosa in Great Britain

Xylella fastidiosa (Xf) is a significant threat affecting the agricultural and horticultural industries worldwide. Once restricted to Americas, severe European outbreaks have been discovered, the most infamously known in Apulia (Italy), where the bacterium is still spreading killing millions of olive trees. It is a great concern due to the high levels of plant trade, but in order to prevent and control its emergence in novel locations, it is vital to understand drivers of entry, establishment, and spatiotemporal spread, which are not fully understood. Great Britain is currently Xylella-free, but due to the extensive plant trade network, is considered potentially at risk. Although suitability distribution models (SDMs) suggest unfavourable climatic conditions for Xf emergence there, this approach neglects crucial epidemiological dynamics. Hence, to improve Xf pest risk assessment in Britain, we adapted a spatially explicit mechanistic spread model originally designed to describe the Apulian outbreak, incorporating British specific eco-epidemiological factors. We simulated deterministic short-distance and stochastic long-distance dispersal at country-level, accounting for British environmental gradients and taking a “worst-case scenario” approach to define susceptible host population. Infection was randomly seeded and different scenarios were simulated to evaluate different parameter values, including different dispersal processes, as estimation was uncertain due to lack of epidemiological data. Results showed that temperature was a limiting factor to disease transmission, and host distribution significantly affected local prevalence and extent of spread, suggesting that investigation of spatial dynamics is crucial to reveal areas at risk in case of Xf introduction in new locations. Despite the high degree of parameter uncertainty, our results displayed consistent qualitative patterns, highlighting differences with SDMs and other forms of risk assessment. The model is flexible to be updated whether new data become available, and eventually adapted to other areas, with the ultimate aim of providing guidance to develop preparedness measures.