This study examined the prevalence of seed-borne fungi in polyethylene plastic-packaged soybean seeds stored in a genebank and identified factors influencing their incidence. Twenty-four seed lots were sampled from the collection stored at 10 °C in the World Vegetable Center genebank in Arusha, Tanzania. The seeds used were those regenerated and harvested in 2015, 2016, and 2017. A seed health test was conducted for sterilized seeds on potato dextrose agar, sterilized seeds on top of the paper, and unsterilized seeds on top of the paper. Seven-year-old sterilized seeds plated on top of the paper exhibited the highest germination percentage (74 ± 7.09%) and lowest fungal incidence (7.00 ± 4.41%). Conversely, seven-year-old unsterilized seeds plated on top of the paper had the lowest germination (22.00 ± 6.97%) and highest fungal incidence (79.00 ± 8.31%). Older seed lots showed significantly higher equilibrium seed moisture content (Eqmc), though seed age had no significant influence on germination percentage or fungal incidence. Seed germination percentage showed a significant negative relationship with Eqmc, though Eqmc had no significant effect on fungal incidence. Meanwhile, germination percentage showed a significant negative relationship with fungal incidence. Six fungi isolates were identified using their morphological features from soybean seed samples: Absidia glauca, Aspergillus niger, Fusarium spp., Mucor hiemalis, Pestalotiopsis versicolor, and Sordaria macrospora. It is concluded that high fungal incidence negatively affects seed germination but is not significantly correlated with seed moisture content. The dry and cold storage conditions in the genebank, while effective in extending seed longevity, can unintentionally allow seed-borne mycoflora to persist. Although fungi may not actively proliferate under these conditions, they may do so when favorable conditions are restored, such as during germination tests, and consequently may compromise seed viability. Therefore, this study emphasizes the importance of increased systematic seed health checks before storage in genebanks to ensure optimal seed quality, and the replacement of all polyethylene plastic bags to aluminium foil packaging.

Deep learning is an emerging, powerful technology that facilitates object detection in plant science, enabling applications such as disease diagnostics and identification of plant species and crop varieties using image analysis. In this study, we demonstrated the success of deep learning algorithms when used in identifying Tanzania-released cassava varieties ( TARICASS4 , Kiroba , and Mkuranga1 ) through image analysis. Three deep learning algorithms (YOLO v8, SSD-MobileNet v2, and Faster R-CNN with ResNet-50) were trained using a dataset of 19,800 images from leaves, petioles and stems of cassava plants of the three varieties. Trained algorithms achieved promising performance in identifying the cassava varieties using each of the three plant parts (mAP50≥85.7%), with the highest values recorded for leaves (mAP50≥92.7%). Of the three models, the YOLO v8 model showed the consistent promising performance for the three plant parts, with an mAP50 of 97.3% for leaves. Further findings from confusion matrices confirmed the superior performance of YOLO v8 for leaves, with performance of ≥98% for each of the three varieties. YOLO v8 also achieved high levels of performance for stems and petioles for all varieties. Our results confirm that deep learning models can be effectively applied for AI-based identification of field-grown cassava. The approach developed for cassava has strong potential for scaling to additional cassava varieties as well as other Vegetatively Propagated Crops, and for use by seed producers and regulators who could benefit from the use of a field-usable tool for verifying variety ‘true-to-typeness’.