Prediction of aflatoxin contamination outbreaks in Texas corn using mechanistic and machine learning models

Authors: Castano-Duque, Lina; AVILA, ANGELA - University Of Texas At Arlington; Mack, Brian; WINZELER, H - University Of Texas At Arlington; Blackstock, Joshua; Lebar, Matthew; Moore, Geromy; Owens, Phillip; Mehl, Hillary; SU, JIANZHONG - University Of Texas At Arlington; Lindsay, James; Rajasekaran, Kanniah

Submitted to: Frontiers in Microbiology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/5/2025
Publication Date: 3/5/2025
Citation: Castano-Duque, L.M., Avila, A., Mack, B.M., Winzeler, H.E., Blackstock, J.M., Lebar, M.D., Moore, G.G., Owens, P.R., Mehl, H.L., Su, J., Lindsay, J.A., Rajasekaran, K. 2025. Prediction of aflatoxin contamination outbreaks in Texas corn using mechanistic and machine learning models. Frontiers in Microbiology. 16. Article 1528997. https://doi.org/10.3389/fmicb.2025.1528997.
DOI: https://doi.org/10.3389/fmicb.2025.1528997

Interpretive Summary: Aflatoxins are carcinogenic and mutagenic mycotoxins produced by fungi that contaminate the food and feed supply under field or storage conditions. To predict mycotoxin outbreaks, we employed an ensemble of models to estimate the probability of aflatoxin contamination in corn (maize), at county level, across Texas. Our models utilized high-throughput dynamic geospatial data from remote sensing satellites, soil property maps, and meteorological data. We developed three model ensemble analysis pipelines: two mechanistic models that used weekly aflatoxin risk indexes (ARIs) as inputs, and one weather-dependent model. Our study concluded that Texas had significant geographical variability in ARI and ARI-hotspot responses due to its diverse regions across the state (hot-dry, hot-humid, mixed-dry and mixed-humid). This diversity leads to high temporal and spatial variability in weather and subsequent planting times, resulting in a wide variation of maize temporal development. Our neural network model identified a positive correlation between AFL outbreaks and ARI hot-spots prevalence in the hot-humid areas of Texas, where high temperature, precipitation and relative humidity in March and October led to increased aflatoxin contamination events. We found that depending on the region in Texas, there is a positive correlation between aflatoxin outbreaks and soil pH in hot-dry and hot-humid regions and minimum saturated hydraulic conductivity in mixed-dry regions. Conversely, there was a negative relationship between aflatoxin outbreaks and soil rock fragments (hot-dry region), maximum soil organic matter (hot-dry region), calcium carbonate (hot-dry, and mixed-dry), and soil depth (mixed-dry, hot-humid and mixed-humid region). Our results demonstrate intricate relationships between soil hydrological parameters, fungal communities and plant health. It is possible that soil fungal communities are more diverse, and the plants are healthier in soils with high organic matter content, leading to lower aflatoxin outbreaks. These findings suggest that prior to any implementation of prediction, prevention, or mitigation strategies of mycotoxin outbreaks the complex interactions should be considered by Texas corn growers.

Technical Abstract: Aflatoxins are carcinogenic mycotoxins that contaminate food and feed, to predict mycotoxin outbreaks in maize, we developed an ensemble of models to estimate probability of aflatoxin contamination, at county level, across Texas. Our models utilized dynamic geospatial data from remote-sensing satellites, soil properties, and meteorological data. We developed three model ensemble pipelines: two mechanistic models that used aflatoxin risk indexes (ARIs) as inputs, and one weather-centric model. The ARIs were weighted based on a maize phenological model that estimated the maize planting times based on satellite-acquired data. The best performing model was Ratkowsky-ARI nnet with an accuracy of 73%, sensitivity of 71% and specificity of 74%. We concluded that Texas had significant geographical variability in ARI and ARI-hotspots due to its diverse regions (hot-dry, hot-humid, mixed-dry and mixed-humid). We found a negative corre-lation between aflatoxin outbreaks and maximum soil organic matter (hot-dry region), and cal-cium carbonate (hot-dry, and mixed-dry). Perhaps, soil fungal communities are more diverse, and plants are healthier in soils with high organic matter and calcium content, leading to lower af-latoxin outbreaks. Our results demonstrate intricate relationships between soil parameters, fungal communities and plant health that should be considered by Texas corn growers as part of miti-gation strategies.