Location: Mycotoxin Prevention and Applied Microbiology Research
Project Number: 5010-42000-053-006-R
Project Type: Reimbursable Cooperative Agreement
Start Date: Jun 30, 2026
End Date: Sep 14, 2027
Objective:
Objective 1) Retrain recently published Illinois mycotoxin prediction models by using daily data from Illinois and Texas to develop region-specific models that accurately forecast contamination levels quarterly and alert farmers of heightened risks on a monthly basis.
Objective 2) Validate and refine fungal growth and mycotoxin contamination models with empirical field data collected in Illinois and Texas in 2026 and 2027 (Bring biocontrol in Texas as a factor).
Approach:
Objective 1 will be primarily conducted by cooperators/ coPIs. Published gradient boosting model (GBM) and bayesian network (BN) model use daily weather data for IL and cooperators will also test a new neural network (NN) model. The models will be compared for accuracy and the ability to predict high, medium and low levels of aflatoxin (AFL) and fumonisin (FUM) contamination post-harvest. In addition to the input features used in current models, soil type characteristics of the areas in the counties that are used for growing corn will also be included. An alert system that will use model output will serve as the main parameters to monitor monthly in order to generate an alert and recommendation system for stakeholders that indicates their monthly/weekly mycotoxin risk factor due to how similar their current weather conditions are to the features that historically led to high contamination levels of AFL and FUM. A forecasting alert system will also be generated, which will give a prediction of state-specific forecast for end of the year contamination levels by county.
Objective 2 will include field sampling at two timepoints during the year: prior to planting and at corn flowering/silking stage. ARS co-PI will be involved in IL sampling. Samples will include soil, crop debris, weeds, corn leaves and cobs. Additionally, since 2021 ARS has received subsamples of corn used in this survey for isolation and characterization of associated mycotoxigenic fungi. As part of this project we will quantify, isolate, genetically characterize, and compare Fusarium and Aspergillus populations of 2021- 2023 corn from IL counties of high and low mycotoxin contamination levels. A high throughput quantitative digital polymerase chain reaction (PCR) method previously validated by the team will be used to distinguish between fungal species and estimate relative abundances of predominant toxigenic and non-toxigenic Aspergillus and Fusarium species.