Pathogen and pest communities in food security crops across climate gradients: Anticipating future challenges in the Great Lakes region of Africa

Natural and human-driven disasters are a significant challenge to the sustainable production of food security crops in the Great Lakes region of Africa. A relevant research gap is the effect of climate change on the distribution of pathogens and pests of food security crops in this region. We evaluated the current geographic risk of pathogens and pests in the production of four food security crops - banana, cassava, potato, and sweetpotato - in the region, and the potential future risk under climate change. First, cropland connectivity analysis identified locations likely important in the spread and establishment of crop-specific pathogens and pests for each crop, with locations in Rwanda and Burundi emerging as important across crops. Second, we surveyed 27 economically important pathogens and pests in Rwanda and Burundi, mapping their distribution across landscapes and quantifying patterns of association. Cropland density, cropland connectivity, altitude, and contemporary and long-term temperature and precipitation were strongly associated with disease severity and pest infestation. Among ten machine-learning algorithms evaluated, random forests and support vector machines generally performed best for predicting severity and infestation. Third, an increase in temperature across altitudes is projected under future climate change scenarios in this region. We found evidence that 44% of the pathogen and pest species we studied in banana, potato, and sweetpotato could become more common with warmer temperatures at higher altitudes, while 17% may become less common. Ongoing development of pathogen and pest forecasts can guide strategic surveillance, mitigation and adaptation to future epidemics and pest invasions in high-risk crop locations under climate change.