Bacterial leaf streak (BLS) caused by Xanthomonas oryzae pv. oryzicola (Xoc) is one of the most devastating diseases in rice production areas, especially in humid tropical and subtropical zones throughout Asia and worldwide. A genome-wide association study (GWAS) analysis conducted on a collection of 236 diverse rice accessions, mainly indica varieties, identified 12 quantitative trait loci (QTLs) on chromosomes 1, 2, 3, 4, 5, 8, 9 and 11, conferring resistance to five representative isolates of Thai Xoc. Of these, five QTLs conferred resistance to more than one Xoc isolates. Two QTLs, qBLS5.1 and qBLS2.3, were considered promising QTLs for broad-spectrum resistance to BLS. The xa5 gene was proposed as a potential candidate gene for qBLS5.1 and three genes, encoding pectinesterase inhibitor (OsPEI), eukaryotic zinc-binding protein (OsRAR1), and NDP epimerase function, were proposed as candidate genes for qBLS2.3. Results from this study provide an insight into the potential QTLs and candidate genes for BLS resistance in rice. The recessive xa5 gene is suggested as a potential candidate for strong influence on broad-spectrum resistance and as a focal target in rice breeding programs for BLS resistance.

Bacterial leaf blight (BLB) is a serious disease affecting global rice agriculture caused by Xanthomonas oryzae pv. oryzae (Xoo). Most resistant rice lines are dependent on single genes that are vulnerable to resistance breakdown caused by pathogen mutation. Here we describe a genome-wide association study of 222 predominantly Thai rice accessions assayed by phenotypic screening against 20 Xoo isolates. Loci corresponding to BLB resistance were detected using >142,000 SNPs. We identified 147 genes according to employed significance thresholds across chromosomes 1–6, 8, 9 and 11. Moreover, 127 of identified genes are located on chromosomal regions outside estimated Linkage Disequilibrium influences of known resistance genes, potentially indicating novel BLB resistance markers. However, significantly associated SNPs only occurred across a maximum of six Xoo isolates indicating that the development of broad-spectrum Xoo strain varieties may prove challenging. Analyses indicated a range of gene functions likely underpinning BLB resistance. In accordance with previous studies of accession panels focusing on indica varieties, our germplasm displays large numbers of SNPs associated with resistance. Despite encouraging data suggesting that many loci contribute to resistance, our findings corroborate previous inferences that multi-strain resistant varieties may not be easily realised in breeding programs without resorting to multi-locus strategies.

To help farmers manage limited resources, rice disease diagnosis must be accurate, timely, and affordable. This study addresses challenges in rice field images, such as environmental variability and differences in rice leaf sizes. The proposed technique combines convolutional neural network object detection with image tiling, using estimated rice leaf width as a reference for image division. An 18‐layer ResNet model was trained using ground truth leaf width values for regression in leaf width estimation. Experiments used a dataset of 4960 images representing 8 rice diseases. The leaf width prediction model achieved a mean absolute percentage error of 11.18% and was used to generate a tiled dataset for training advanced rice disease detection models. The tiling technique was evaluated using YOLOv4, YOLOv8n, YOLOv8l, DINO‐5scale Swin‐L, and Co‐DINO‐5scale Swin‐L models by comparing detection performance on original and tiled datasets. Mean average precision improved significantly: YOLOv4 increased from 87.56% to91.14%, YOLOv8n from 89.80% to 91.70%, and YOLOv8l from 89.80% to 93.20%. More advanced models, such as DINO5scale Swin‐L and Co‐DINO‐5scale Swin‐L, achieved even higher precision, at 93.40% and 94.20%, respectively. In conclusion, the tiling technique improved detection efficiency and addressed object size variability, enhancing rice disease detection accuracy inreal‐world scenarios.

Burkholderia glumae is one of the most critical rice-pathogenic bacteria, and it causes bacterial panicle blight (BPB) in rice plants. In 2017, BPB symptoms were observed from rice fields in Chiang Rai, Northern Thailand. Sixty-one isolates obtained from the symptomatic panicles of rice were initially identified as B. glumae by polymerase chain reaction (PCR) using species-specific primers. Among them, six selected strains isolated from the susceptible japonica rice cultivar DOA2 were characterized in terms of morpho-physiology, pathology, phylogenetics, and genomics. Our genome sequence analysis of the six selected strains revealed the presence of multiple prophages, which may reflect the high level of diversity in this bacterial species through dynamic horizontal gene transfer processes, including phage infection. This notion was supported by the results of phylogenetic and phylogenomic analyses, which showed the formation of several subgroups not related to the years of isolation or the geographical origins. This study reports the isolation of B. glumae as the causal pathogen of BPB disease in japonica rice in Thailand and provides genomic resources to better understand the biology and diversity of this plant pathogenic bacterium. Further studies with a vast collection of B. glumae strains from various rice-growing regions around the world are needed to elucidate the evolution, variability, and lifestyle of the pathogen.

Rice is the most important grain in Thailand for both consuming and exporting. One of the critical problems in rice cultivation is rice diseases, which affects directly to the yield. Early disease recognition is handled by a human, which is difficult to achieve high accuracy and the performance depends on the farmer’s experience. To overcome this problem, we did three folds of contributions. First, an infield rice diseases image dataset, named K5RD, was created. Second, a number of additional techniques to enhance the classification scores including data augmentations and learning rate adjustment strategies were carefully surveyed. Third, a set of selective deep learning models including ResNets and DenseNets were applied to classify such rice diseases. The experimental results reveal that the proposed framework can achieve high performance, which its F1 score is higher than 98% on average, and has the potential to be implemented as a practical system to provide to Thai farmers in the future.