Leaf rust, caused by the fungus Puccinia triticina, is one of the most destructive diseases affecting global wheat production. Developing disease-resistant wheat varieties is the most cost-effective and environmentally friendly approach to managing this disease. We phenotyped a collection of 559 wheat accessions from five continents for resistance to leaf rust in field trials at three locations in China (Zhoukou, Henan; Wuhan, Hubei; and Xinxiang, Henan) during the 2020–2021, 2021–2022, and 2022–2023 cropping seasons, followed by best-linear-unbiased-estimation analysis across environments. These accessions were genotyped using the MGISEQ-2000 re-sequencing platform, and a genome-wide association analysis was subsequently performed. Twenty-four stable leaf rust resistance loci across 15 chromosomes were identified. Among these, 11 loci may represent new sources of resistance. Notably, Lr.hzau-2BS.1 and Lr.hzau-7AL were consistently detected across all three environments and BLUE. Lr.hzau-2BS.1 has the highest frequency in European wheat accessions, whereas Lr.hzau-7AL is most prevalent in South American accessions. Gene-expression analysis identified 101 candidate genes associated with these loci. Closely linked Kompetitive Allele Specific PCR (KASP) markers, 2B-209172 and 7A-348992, were developed for Lr.hzau-2BS.1 and Lr.hzau-7AL, respectively. Chinese wheat varieties Mianmai 45 and Liaomai 16, which carry resistance alleles at both loci and exhibit < 5% leaf rust severity, represent valuable sources of leaf rust resistance for wheat breeding programs. These newly identified resistance loci and their KASP markers provide valuable resource for their exploitation in wheat breeding.

Wheat stripe rust poses a marked threat to global wheat production. Accurate and effective disease severity assessments are crucial for disease resistance breeding and timely management of field diseases. In this study, we propose a practical solution using mobile-based deep learning and model-assisted labeling. StripeRust-Pocket, a user-friendly mobile application developed based on deep learning models, accurately quantifies disease severity in wheat stripe rust leaf images, even under complex backgrounds. Additionally, StripeRust-Pocket facilitates image acquisition, result storage, organization, and sharing. The underlying model employed by StripeRust-Pocket, called StripeRustNet, is a balanced lightweight 2-stage model. The first stage utilizes MobileNetV2-DeepLabV3+ for leaf segmentation, followed by ResNet50-DeepLabV3+ in the second stage for lesion segmentation. Disease severity is estimated by calculating the ratio of the lesion pixel area to the leaf pixel area. StripeRustNet achieves 98.65% mean intersection over union (MIoU) for leaf segmentation and 86.08% MIoU for lesion segmentation. Validation using an additional 100 field images demonstrated a mean correlation of over 0.964 with 3 expert visual scores. To address the challenges in manual labeling, we introduce a 2-stage labeling pipeline that combines model-assisted labeling, manual correction, and spatial complementarity. We apply this pipeline to our self-collected dataset, reducing the annotation time from 20 min to 3 min per image. Our method provides an efficient and practical solution for wheat stripe rust severity assessments, empowering wheat breeders and pathologists to implement timely disease management. It also demonstrates how to address the “last mile” challenge of applying computer vision technology to plant phenomics.

Leaf rust (LR) and stripe rust (YR) are important diseases in wheat producing areas worldwide and cause severe yield losses under favorable environmental conditions when susceptible varieties are grown. We determined the genetic basis of resistance to LR and YR in variety Borlaug 100 by developing and phenotyping a population of 198 F₆ recombinant inbred lines derived from a cross with the susceptible parent Apav#1. LR and YR phenotyping were conducted for 4 and 3 seasons, respectively, at CIMMYT research stations in Mexico under artificial epidemics. Mendelian segregation analyses indicated that 3–5 LR and 2 YR genes conferred resistance in Borlaug 100. Lr46/Yr29 (1BL), Yr17 (2AS) and Yr30 (3BS) were present in the resistant parent and segregated in the RIL population based on characterization by molecular markers linked to these genes. When present alone, Lr46/Yr29 caused average 13% and 16% reductions in LR and YR severities, respectively, in RILs. Similarly, Yr17 and Yr30 reduced YR severities by 57% and 11%, respectively. The Yr30 and the Yr17 translocation were also associated with 27% and 14% reductions, respectively, in LR severity, indicating that the 3BS and 2AS chromosomal regions likely carry new slow rusting LR resistance genes, temporarily designated as LrB1 and LrB2, respectively. Additive effects of Yr30*Yr17, Yr29*Yr17 and Yr29*Yr30 on YR and LR were significant and reduced YR severities by 56%, 55%, and 45%, respectively, and LR severities by 34%, 40%, and 45%, respectively. Furthermore, interaction between the three genes was also significant, with mean reductions of 70% for YR and 54% for LR severities. Borlaug 100, or any one of the 21 lines with variable agronomic traits but carrying all three co-located resistance loci, can be used as resistance sources in wheat breeding programs.