Since the early 1960’s, the fungal pathogen Aspergillus flavus (Link ex Fr.) has been the focus of intensive research due to the production of carcinogenic and highly toxic secondary metabolites collectively known as aflatoxins following pre-harvest colonization of crops. Given this recurrent problem and the occurrence of a severe aflatoxin outbreak in maize (Zea mays L.), particularly in the Southeast U.S. in the 1977 growing season, a significant research effort has been put forth to determine the nature of the interaction occurring between aflatoxin production, A. flavus, environment and its various hosts before harvest. Many studies have investigated this interaction at the genetic, transcript, and protein levels, and in terms of fungal biology at either pre- or post-harvest time points. Later experiments have indicated that the interaction and overall resistance phenotype of the host is a quantitative trait with a relatively low heritability. In addition, a high degree of environmental interaction has been noted, particularly with sources of abiotic stress for either the host or the fungus such as drought or heat stresses. Here, we review the history of research into this complex interaction and propose future directions for elucidating the relationship between resistance and susceptibility to A. flavus colonization, abiotic stress, and its relationship to oxidative stress in which aflatoxin production may function as a form of antioxidant protection to the producing fungus.

Drought stress decreases crop growth, yield, and can further exacerbate pre-harvest aflatoxin contamination. Tolerance and adaptation to drought stress is an important trait of agricultural crops like maize. However, maize genotypes with contrasting drought tolerances have been shown to possess both common and genotype-specific adaptations to cope with drought stress. In this research, the physiological and metabolic response patterns in the leaves of maize seedlings subjected to drought stress were investigated using six maize genotypes including: A638, B73, Grace-E5, Lo964, Lo1016, and Va35. During drought treatments, drought-sensitive maize seedlings displayed more severe symptoms such as chlorosis and wilting, exhibited significant decreases in photosynthetic parameters, and accumulated significantly more reactive oxygen species (ROS) and reactive nitrogen species (RNS) than tolerant genotypes. Sensitive genotypes also showed rapid increases in enzyme activities involved in ROS and RNS metabolism. However, the measured antioxidant enzyme activities were higher in the tolerant genotypes than in the sensitive genotypes in which increased rapidly following drought stress. The results suggest that drought stress causes differential responses to oxidative and nitrosative stress in maize genotypes with tolerant genotypes with slower reaction and less ROS and RNS production than sensitive ones. These differential patterns may be utilized as potential biological markers for use in marker assisted breeding.

The maturity of fruits and vegetables such as tomatoes significantly impacts indicators of their quality, such as taste, nutritional value, and shelf life, making maturity determination vital in agricultural production and the food processing industry. Tomatoes mature from the inside out, leading to an uneven ripening process inside and outside, and these situations make it very challenging to judge their maturity with the help of a single modality. In this paper, we propose a deep learning-assisted multimodal data fusion technique combining color imaging, spectroscopy, and haptic sensing for the maturity assessment of tomatoes. The method uses feature fusion to integrate feature information from images, near-infrared spectra, and haptic modalities into a unified feature set and then classifies the maturity of tomatoes through deep learning. Each modality independently extracts features, capturing the tomatoes’ exterior color from color images, internal and surface spectral features linked to chemical compositions in the visible and near-infrared spectra (350 nm to 1100 nm), and physical firmness using haptic sensing. By combining preprocessed and extracted features from multiple modalities, data fusion creates a comprehensive representation of information from all three modalities using an eigenvector in an eigenspace suitable for tomato maturity assessment. Then, a fully connected neural network is constructed to process these fused data. This neural network model achieves 99.4% accuracy in tomato maturity classification, surpassing single-modal methods (color imaging: 94.2%; spectroscopy: 87.8%; haptics: 87.2%). For internal and external maturity unevenness, the classification accuracy reaches 94.4%, demonstrating effective results. A comparative analysis of performance between multimodal fusion and single-modal methods validates the stability and applicability of the multimodal fusion technique. These findings demonstrate the key benefits of multimodal fusion in terms of improving the accuracy of tomato ripening classification and provide a strong theoretical and practical basis for applying multimodal fusion technology to classify the quality and maturity of other fruits and vegetables. Utilizing deep learning (a fully connected neural network) for processing multimodal data provides a new and efficient non-destructive approach for the massive classification of agricultural and food products.

Key impediments to increased corn yield and quality in the southeastern US coastal plain region are damage by ear-feeding insects and aflatoxin contamination caused by infection of Aspergillus flavus. Key ear-feeding insects are corn earworm, Helicoverpa zea, fall armyworm, Spodoptera frugiperda, maize weevil, Sitophilus zeamais, and brown stink bug, Euschistus servus. In 2006 and 2007, aflatoxin contamination and insect damage were sampled before harvest in three 0.4-hectare corn fields using a grid sampling method. The feeding damage by each of ear/kernel-feeding insects (i.e., corn earworm/fall armyworm damage on the silk/cob, and discoloration of corn kernels by stink bugs), and maize weevil population were assessed at each grid point with five ears. The spatial distribution pattern of aflatoxin contamination was also assessed using the corn samples collected at each sampling point. Aflatoxin level was correlated to the number of maize weevils and stink bug-discolored kernels, but not closely correlated to either husk coverage or corn earworm damage. Contour maps of the maize weevil populations, stink bug-damaged kernels, and aflatoxin levels exhibited an aggregated distribution pattern with a strong edge effect on all three parameters. The separation of silk- and cob-feeding insects from kernel-feeding insects, as well as chewing (i.e., the corn earworm and maize weevil) and piercing-sucking insects (i.e., the stink bugs) and their damage in relation to aflatoxin accumulation is economically important. Both theoretic and applied ramifications of this study were discussed by proposing a hypothesis on the underlying mechanisms of the aggregated distribution patterns and strong edge effect of insect damage and aflatoxin contamination, and by discussing possible management tactics for aflatoxin reduction by proper management of kernel-feeding insects. Future directions on basic and applied research related to aflatoxin contamination are also discussed.

The corn earworm, Helicoverpa zea (Boddie), is a major pest targeted by pyramided Bacillus thuringiensis (Bt) corn and cotton in the U.S. Cry1Ab is one of the first insecticidal toxins used in Bt crops, while Vip3A is a relatively new toxin that has recently been incorporated into Cry corn with event MIR 162 and Cry cotton varieties to generate pyramided Bt traits targeting lepidopteran pests including H. zea. The objectives of this study were to determine the current status and distribution of the Cry1Ab resistance, and evaluate the susceptibility to Vip3Aa20 expressed in MIR 162 corn in H. zea in the southeastern U.S. During 2018 and 2019, 32 H. zea populations were collected from non-Bt corn (19 populations), Cry corn (12), and Cry/Vip3A cotton (1) across major corn areas in seven southeastern states of the U.S. Susceptibility of these populations to Cry1Ab and Vip3Aa20 was determined using diet-overlay bioassays. Compared to a known susceptible insect strain, 80% of the field populations were 13- to >150-fold resistant to Cry1Ab, while their response to Vip3Aa20 ranged from >11-fold more susceptible to 9-fold more tolerant. Mean susceptibility to each Bt toxin was not significantly different between the two groups of the populations collected from non-Bt and Bt crops, as well as between the two groups of the populations collected during 2018 and 2019. The results show that resistance to Cry1Ab in H. zea is widely distributed across the region. However, the Cry1Ab-resistant populations are not cross-resistant to Vip3Aa20, and H. zea in the region is still susceptible to the Vip3Aa20 toxin. Vip3Aa20 concentrations between 5 and 10 µg/cm² may be used as diagnostic concentrations for susceptibility monitoring in future. Additional studies are necessary to elucidate the impact of the selection with Bt corn on resistance evolution in H. zea to Vip3A cotton in the U.S.