Mapping aflatoxin contamination risk in West Africa
Figure: Distribution of aflatoxins samples in Mali, Burkina Faso, Togo and Niger.
Aflatoxin contamination in grains is a significant health risk to humans and livestock as it causes human cancer, liver cirrhosis, stunting in children, and death. Moreover, it hinders international trade of cereal grains causing significant economic losses. Contamination is influenced by various factors, including temperature, humidity, precipitation, soil moisture, agronomic management practices, and storage conditions. Evidence-based information on the factors that propagate the contamination risk and maps identifying hotspots beyond the sampled farms is lacking in Africa. The risk of contamination is accentuated by conditions associated with climate change and variability. Reliable maps are necessary for identifying hotspots in current and future climatic conditions. The proposed study will utilize geotagged samples of Aflatoxin contamination levels to train machine learning models to identify the environmental risk factors and predict aflatoxin risk in major cereal grains in the West Africa region. A prototype early warning system will be developed to forecast aflatoxin risk before the start of the season, enabling farmers to prepare to apply appropriate mitigation measures.
The primary objective of the proposed study is to develop a prototype of an early warning system for forecasting aflatoxin contamination risk in major grains in West Africa. The specific objectives are to: (1) train a machine learning model to identify the environmental and agronomic management practices driving aflatoxin contamination levels in major grain crops, (2) apply the model to predict/forecast the aflatoxin contamination risk for current and future climatic conditions in West Africa, (3) identify the hotspots and the population at risk of aflatoxin contamination.
Battilani, P., Toscano, P., Van der Fels-Klerx, H. J., Moretti, A., Camardo Leggieri, M., Brera, C., . . . Robinson, T. (2016). Aflatoxin B1 contamination in maize in Europe increases due to climate change. Scientific Reports, 6(1), 24328. doi:10.1038/srep24328
Castano-Duque, L., Vaughan, M., Lindsay, J., Barnett, K., & Rajasekaran, K. (2022). Gradient boosting and bayesian network machine learning models predict aflatoxin and fumonisin contamination of maize in Illinois – First USA case study. Frontiers in Microbiology, 13. doi:10.3389/fmicb.2022.1039947
Gachoki, S., Muthoni, F., Mahuku, G., Atehnkeng, J., Njeru, N., Kamau, J., & Tripathi, L. (2025). Predicting aflatoxin risk in maize using machine learning and satellite data in East and Southern Africa. World Mycotoxin Journal, 1-20. doi:https://doi.org/10.1163/18750796-bja10018
Keller, B., Russo, T., Rembold, F., Chauhan, Y., Battilani, P., Wenndt, A., & Connett, M. (2021). The potential for aflatoxin predictive risk modelling in sub-Saharan Africa: a review. World Mycotoxin Journal, 15, 1-18. doi:10.3920/WMJ2021.2683
Kerry, R., Ingram, B., Garcia-Cela, E., Magan, N., Ortiz, B. V., & Scully, B. (2021). Determining future aflatoxin contamination risk scenarios for corn in Southern Georgia, USA using spatio-temporal modelling and future climate simulations. Scientific Reports, 11(1), 13522. doi:10.1038/s41598-021-92557-6