Search

Common beans (CB), a vital source for high protein content, plays a crucial role in ensuring both nutrition and economic stability in diverse communities, particularly in Africa and Latin America. However, CB cultivation poses a significant threat to diseases that can drastically reduce yield and quality. Detecting these diseases solely based on visual symptoms is challenging, due to the variability across different pathogens and similar symptoms caused by distinct pathogens, further complicating the detection process. Traditional methods relying solely on farmers' ability to detect diseases is inadequate, and while engaging expert pathologists and advanced laboratories is necessary, it can also be resource intensive. To address this challenge, we present a AI-driven system for rapid and cost-effective CB disease detection, leveraging state-of-the-art deep learning and object detection technologies. We utilized an extensive image dataset collected from disease hotspots in Africa and Colombia, focusing on five major diseases: Angular Leaf Spot (ALS), Common Bacterial Blight (CBB), Common Bean Mosaic Virus (CBMV), Bean Rust, and Anthracnose, covering both leaf and pod samples in real-field settings. However, pod images are only available for Angular Leaf Spot disease. The study employed data augmentation techniques and annotation at both whole and micro levels for comprehensive analysis. To train the model, we utilized three advanced YOLO architectures: YOLOv7, YOLOv8, and YOLO-NAS. Particularly for whole leaf annotations, the YOLO-NAS model achieves the highest mAP value of up to 97.9% and a recall of 98.8%, indicating superior detection accuracy. In contrast, for whole pod disease detection, YOLOv7 and YOLOv8 outperformed YOLO-NAS, with mAP values exceeding 95% and 93% recall. However, micro annotation consistently yields lower performance than whole annotation across all disease classes and plant parts, as examined by all YOLO models, highlighting an unexpected discrepancy in detection accuracy. Furthermore, we successfully deployed YOLO-NAS annotation models into an Android app, validating their effectiveness on unseen data from disease hotspots with high classification accuracy (90%). This accomplishment showcases the integration of deep learning into our production pipeline, a process known as DLOps. This innovative approach significantly reduces diagnosis time, enabling farmers to take prompt management interventions. The potential benefits extend beyond rapid diagnosis serving as an early warning system to enhance common bean productivity and quality., (© 2024. The Author(s).)

Background: Common bean is an important staple crop in the tropics of Africa, Asia and the Americas. Particularly smallholder farmers rely on bean as a source for calories, protein and micronutrients. Drought is a major production constraint for common bean, a situation that will be aggravated with current climate change scenarios. In this context, new tools designed to understand the genetic basis governing the phenotypic responses to abiotic stress are required to improve transfer of desirable traits into cultivated beans., Results: A multiparent advanced generation intercross (MAGIC) population of common bean was generated from eight Mesoamerican breeding lines representing the phenotypic and genotypic diversity of the CIAT Mesoamerican breeding program. This population was assessed under drought conditions in two field trials for yield, 100 seed weight, iron and zinc accumulation, phenology and pod harvest index. Transgressive segregation was observed for most of these traits. Yield was positively correlated with yield components and pod harvest index (PHI), and negative correlations were found with phenology traits and micromineral contents. Founder haplotypes in the population were identified using Genotyping by Sequencing (GBS). No major population structure was observed in the population. Whole Genome Sequencing (WGS) data from the founder lines was used to impute genotyping data for GWAS. Genetic mapping was carried out with two methods, using association mapping with GWAS, and linkage mapping with haplotype-based interval screening. Thirteen high confidence QTL were identified using both methods and several QTL hotspots were found controlling multiple traits. A major QTL hotspot located on chromosome Pv01 for phenology traits and yield was identified. Further hotspots affecting several traits were observed on chromosomes Pv03 and Pv08. A major QTL for seed Fe content was contributed by MIB778, the founder line with highest micromineral accumulation. Based on imputed WGS data, candidate genes are reported for the identified major QTL, and sequence changes were identified that could cause the phenotypic variation., Conclusions: This work demonstrates the importance of this common bean MAGIC population for genetic mapping of agronomic traits, to identify trait associations for molecular breeding tool design and as a new genetic resource for the bean research community.

HarvestPlus, part of the Consultative Group on Internation Agriculture research (CGIAR) Program on Agriculture for Nutrition and Health (A4NH) uses conventional plant breeding techniques to develop staple food crops that are rich in micronutrients, a food-based approach to reduce micronutrient malnutrition known as biofortification. The nutritional breeding targets are established based on the food intake of target populations, nutrient losses during storage and processing and bioavailability. This review collates the evidence on the retention of provitamin A carotenoid (pVAC) after processing, cooking, and storing of the staple crops targeted for pVAC biofortification: cassava, maize, and sweet potato. Sun drying was more detrimental to the pVAC levels (27-56% retention) in cassava than shade (59%) or oven (55-91%) drying, while the pVAC retention levels (66-96%) in sweet potato were not significantly different among the various drying methods. Overall, boiling and steaming had higher pVAC retention (80-98%) compared to baking (30-70%) and frying (18-54%). Gari, the most frequently consumed form of cassava in West Africa had the lowest pVAC retention (10-30%). The pVAC retention of maize grain and cassava and sweet potato flour reached levels as low as 20% after 1-4 months of storage and was highly dependent on genotype. Therefore, we recommend that an evaluation of the pVAC degradation rate among different genotypes be performed before a high pVAC crop is promoted.