Trait-mediated filtering of Phytophthora pathogen invasions through global horticultural trade networks.

Abstract

Estimates of invasion risk can support prioritisation of future threats from non-native species. Greater risk of invasion is expected when species occur in connected source regions and possess traits promoting successful transport, introduction or establishment. We compile a global database of first reports of Phytophthora de Bary species, a diverse oomycete genus attacking a broad range of plant hosts across multiple regions, sectors and ecosystem types with increasing frequency. Using Bayesian hierarchical zero-inflated models, we model global patterns of new detections since 2005 among 109 Phytophthora pathogens across 56 countries with at least two known Phytophthora species reported before 2005. We estimate the effects of trade connectivity, climate matching, national surveillance and pathogen traits on the probability of a new detection. We find that 69 (38%) Phytophthora species were either unknown or had no known source regions before 2005 and were therefore excluded from our analysis. Our study shows that invasion risk is increased for pathogens with broader thermal tolerance and the ability to produce survival structures linked to stress tolerance and asymptomatic infections. This knowledge can be used to enhance national horizon scanning and risk-based surveillance activities to better manage risks to plant health from emerging pathogens.