Progress 10/01/23 to 09/30/24
Outputs
PROGRESS REPORT Objectives (from AD-416): Develop methodologies and technologies that prevent and/or defend against the presence of aflatoxin in corn grown in the USA. Approach (from AD-416): Aflatoxins belong to the general class of mycotoxins produced by Aspergillus flavus and A. parasiticus fungi. These fungi are ubiquitous in many soils where corn is grown in the United States. As secondary metabolites aflatoxins are a serious human and animal health problem affecting short and long-term health, trade and export markets of corn- based products. When consumed in low dosages over prolonged periods, aflatoxins may lead to poor nutrient absorption, retarded child growth and development by contributing to malnutrition, increase the incidence and severity of infectious diseases, and suppress the immune system. Chronic exposure is a major risk factor for liver cancer (hepatotoxic carcinoma) in particular, in areas where hepatitis B virus infection is endemic. Ingestion of higher doses of aflatoxin can result in acute aflatoxicosis, which exhibits as hepatotoxicity, or in severe cases, complete liver failure and subsequent death. Corn is a dietary staple in the United States and is one of the cereal crops most susceptible to infection by A. flavus and contamination by aflatoxin. Contamination of corn with aflatoxin continues to be a major issue for industry, in- particular during the on-farm preharvest phase. Two approaches are the development of aflatoxin resistant corn varieties, utilizing various genomic strategies such as classical breeding and interference RNA. However, to date, the most successful is the biological control approach through the application of nontoxigenic strains of A. flavus and A. parasiticus to soils where they competitively exclude naturally toxigenic strains. This research addresses issues outlined in the 2021-2025 NP108 Action Plan, under Problem Statement 5. Develop, Validate and Implement Intervention and Control Strategies to Reduce or Eliminate Pathogens in the Food System. In Texas, weekly models provided an overall accuracy of 67% and higher in a single year predictions of aflatoxin contamination. Between corn growing seasons influence mycotoxin contamination levels at harvest. Soil depth, calcium carbonate, plant available water storage and organic matter are negatively correlated with aflatoxin contamination, while soil conductance is positively correlated. Early in the year surveillance of aflatoxin risk index (ARI) in Texas showed that high levels of risk index between corn seasons (March and November) leads to predicted high levels of mycotoxin contamination at harvest time in corn. Models showed that soil health is a key element in the probability of mycotoxin contamination at harvest time in corn. For example, high calcium carbonate levels in the soil from Illinois and Texas tend to have lower levels of aflatoxin contamination. In Texas, high levels of organic material present in the soil lead to prediction of lower levels of aflatoxin in hot-dry, mixed dry and mixed-humid regions in the state. Displacement of aflatoxigenic fungi by application of non-aflatoxigenic biocontrol strains of Aspergillus flavus to crops is an effective method for reducing pre-harvest aflatoxin contamination. Two aflatoxin biocontrol products, each comprised of a single non-aflatoxigenic A. flavus biocontrol strain, have been applied annually in Texas corn fields for over a decade. Recently, a new multi-strain biocontrol product comprised of strains endemic to Texas was registered by U.S Environmental Protection Agency (EPA), and it became available to producers during the 2024 growing season. However, the extent to which aflatoxin contamination risk and performance of different biocontrol products is influenced by biotic and abiotic factors has not been well characterized. In addition, potential adaptation of genetically distinct biocontrol strains to different agroecosystems, including superior ability to out-compete aflatoxin-producing fungi and remain a dominant component of crop- associated A. flavus populations, has not been elucidated, and this information could be used to improve aflatoxin biocontrol. Thus, field studies were established in Texas with the aim of optimizing management recommendations for biocontrol-based mitigation of aflatoxins in corn. Specific goals of the studies are to 1) compare performance of different aflatoxin biocontrol products / non-aflatoxigenic strains in different regions of Texas; 2) quantify area-wide, multi-year effects of biocontrol application in Texas corn; and 3) collect empirical field data to support models predicting mycotoxin contamination risk that are being developed by ARS scientists in New Orleans, Louisiana. In support of the goals of the project, ARS scientists in Maricopa, Arizona, conducted the first year of sampling for multi-year field studies in Texas. In cooperation with Texas corn producers, study fields were identified in the Coastal Bend, Central, North-Central, and Panhandle regions of Texas. These regions represent different agroecological zones in Texas that vary in climate, soil types, and cropping practices. Fields varied in biocontrol application history and planned biocontrol product application for the current growing season. Soil samples and debris were collected from corn fields between pre-plant and early emergence of the crop to assess the size and composition of A. flavus populations in early spring. In addition, soil and debris were collected from adjacent non-corn fields and non-cultivated areas to assess movement of applied biocontrol strains from treated to non-treated areas and to quantify carryover of applied biocontrol strains between growing seasons. Following corn maturity, soil, debris, and crop samples were collected from the same fields that were sampled in the spring. Corn ears were evaluated for fungal ear rots and insect injury. Soil, debris, and crop samples were homogenized, and A. flavus population densities are being determined using a combination of dilution plating and direct quantification using DNA-based methods. Genotypes of A. flavus recovered from samples are being characterized using molecular markers, and proportions of different biocontrol strain genotypes and aflatoxigenic A. flavus will be determined. Mycotoxins were quantified from crop samples, and soil is being analyzed for physical and chemical properties. Populations of mycotoxigenic Fusarium spp. are being characterized by ARS scientists in Peoria, Illinois. As expected, fungal populations, ear rot severity, insect injury, and mycotoxin concentrations in the harvested crop varied among fields and regions. High frequencies of A. flavus and/or mycotoxigenic Fusarium spp. were recovered from most samples. However, aflatoxin concentrations in all samples were low (<10 ppb), and a majority of A. flavus isolates characterized to date are biocontrol genotypes. Furthermore, biocontrol genotypes were recovered from non-treated fields and non-cultivated areas suggesting area-wide dispersal of applied biocontrol strains. Fusarium ear rot severity and fumonisin concentrations were high (>5 ppm) in several samples collected from the Coastal Bend region. Additional years of sampling will provide insight into 1) regional differences in aflatoxin biocontrol product performance, 2) the frequency with which biocontrol products need to be applied to maintain biocontrol strains as dominant components of crop-associated A. flavus populations, and 3) biotic and abiotic factors that influence mycotoxin contamination risk and aflatoxin biocontrol efficacy. Artificial Intelligence (AI)/Machine Learning (ML) ARS researchers in New Orleans, Louisiana, in collaboration with ARS researchers in Booneville, Arkansas, several state departments of agriculture in the U.S. and researchers in Wageningen University, The Netherlands have initiated a new modeling project using AI/ML to predict aflatoxin and fumonisin contamination in corn grown in Illinois, Iowa, and Texas by using weekly risk analysis. These models are based on a dynamic geospatial analysis of historical weekly toxin risk (an engineered feature that links weather and fungal growth), and weather parameters as well as soil physical properties. The models have shown over 95% overall and over 63% balanced prediction accuracy. These predictive approaches can be used to create a US-centric alert and predictive risk-assessment systems that provides stakeholders a proactive window of opportunity to deploy Integrated Pest Management (IPM) based mycotoxin mitigation strategies.
Impacts
(N/A)
Publications
- 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. Article 1039947. https:// doi.org/10.3389/fmicb.2022.1039947.
- Branstad-Spates, E.H., Castano-Duque, L.M., Mosher, G.A., Hurburgh, Jr., C. R., Owens, P.R., Winzeler, H.E., Rajasekaran, K., Bowers, E.L. 2023. Gradient boosting machine learning model to predict aflatoxins in Iowa corn. Frontiers in Microbiology. 14. Article 1248772. https://doi.org/10. 3389/fmicb.2023.1248772.
- Castano-Duque, L.M., Winzeler, H.E., Blackstock, J.M., Cheng, L., Vergopolan, N., Focker, M., Barnett, K., Owens, P.R., Van Der Fels-Klerx, I., Vaughan, M.M., Rajasekaran, K. 2023. Dynamic geospatial modeling of mycotoxin contamination of corn in Illinois: unveiling critical factors and predictive insights with machine learning. Frontiers in Microbiology. 14. Article 1283127. https://doi.org/10.3389/fmicb.2023.1283127.
|