Fall armyworm infestation and future likelihood of spread due to environmental change in maize-cropping regions of Kenya

Abstract

Abstract Background : Fall armyworm (FAW), Spodoptera frugiperda , is a detrimental pest in maize production, and its infestation is driven by manifold climate factors. Mapping FAW infestation occurrence for climate scenarios is useful for risk management planning. Methods : We utilised public and private FAW observations and extracted various remotely sensed climate data sets as candidate predictors. A maximum extropy algorithm was used to model the probability of FAW presence in Kenya with varied input choices, and the best performance model was selected for future prediction based on performance matrics and existing biological knowledge of the importance of climate factors on FAW prevalence. Results : Our results show that the choice of null hypothesis of FAW distribution (under no climate influence), depicted as background samples, strongly influences the model output. Precipitation, vapour pressure, and temperature were all important for model predictive ability, though to varying degrees depending on the specific climatic data sources. Using combined predictions from the three best models (with an average area under receiver-operating curve of 0.86 and sensitivity of 0.77), we found that only 10.6%of areas in Kenya had current FAW presence probability more than 0.4 (i.e. moderate to high FAW suitability). Most of these areas are in high-maize yielding agroecological zones for Kenya (moist mid-altitude zone e.g. Busia, Bungoma, Siaya), and some are in top maize producer (highland and moist transitional areas e.g. Trans Nzoia). A total of 2.7–3.5% of Kenya will experience an increased probability of FAW attack by 2040, including the areas with the highest maize yield (e.g. Uasin Gishu, Nakuru), in both mild and extreme projected socio-economic pathways. Conclusion : Therefore, intensive surveillance for early detection or adopting more FAW-resistant maize cultivars in these high FAW probability areas is likely to greatly benefit and de-risk the national maize economy and food security in Kenya.