We used 138 F₄-derived F₅ RILs from the cross of Apav#1 × Mucuy. The susceptible parent, Apav#1 (CIMMYT Germplasm ID: 1853706), derived from a cross of “Avocet-YrA” × “Pavon 76,” was susceptible to YR, LR, and stem rust (SR) at all growth stages, against predominant Pst and Pt races used in various trials in Mexico. In contrast, Mucuy (CIMMYT Germplasm ID: 5663955), derived from the cross “Mutus”^*2 × “Akuri,” showed intermediate resistance during seedling evaluations in the greenhouse but high resistance to both YR and LR at the adult plant stage in field trials, in both cases against races Pst and Pt. The RILs were developed following a bulk advancement of the population until the F₄ generation and then harvesting random plants individually to obtain F₅ RILs (Basnet et al., 2014a). We multiplied the RIL seed and used it in all studies.

Disease severity (DS) of the parents and RILs were recorded on 3 occasions using the modified Cobb's Scale (Peterson et al., 1948). Initial data were recorded when the DS of Apav#1 was around 80% and repeated after 7 days. The last data set was recorded when it reached 100%. For multiple disease readings, the area under the disease progress curve (AUDPC) was calculated as per Bjarko and Line (1988). The correlation analysis of final disease severity (FDS) in each environment was conducted using SAS 9.4 software (SAS Institute, Cary, NC).

DNA was extracted using the CTAB method (Dreisigacker et al., 2012) from one-week-old seedlings of the parents and RILs grown in a greenhouse and genotyped with the DArT-GBS platform (Reference for SAGA). In total, we genotyped 40,519 GBS-in-Silico and 39,849 GBS-SNP, in combination with closely linked molecular markers for different rust genes (Xgwm210 for Lr16, cslv46G22 for Lr46/Yr29, two Yr17-linked markers WGGB156 and WGGB159 by Wang et al. (2018), and one Yr17-linked marker provided by Evans Lagudah), in the entire RIL population. Genetic linkage maps were constructed using Joinmap 4.1 (Van Ooijen, 2006) with 12,681 polymorphic markers, and 56 linkage groups were constructed. In addition, QTL analysis was performed by inclusive composite interval mapping (ICIM) using IciMapping 4.2 (Meng et al., 2015) with the DS of each tested environment and the mean of FDS (referred to as YrM for YR and LrM for LR). The logarithm of odds (LOD) score was determined based on the 1,000 permutation test and a significance level of α = 0.05. The percentages of phenotypic variance explained (PVE) were determined using stepwise regression at the LOD peaks (Somers et al., 2004; Francki et al., 2009; Huang et al., 2012; Wilkinson et al., 2012).

The FDS and host response to YR for Apav#1 ranged from 70 to 100 S and for Mucuy ranged from 0 to 5 MS at the adult plant stage over 9 environments. The continuous distribution of YR DS for RILs in each environment indicated the polygenic inheritance of APR (Figures 2A–C).

LR FDS and host responses were 100 S for Apav#1 and 0 for Mucuy over all 3 seasons. The LR DS frequency distributions were distorted and skewed toward the resistance (Figure 2D), indicating the segregation of at least one large-effect LR resistance locus in the population.

The correlation coefficients of FDS for RILs varied from 0.23 to 0.93 in the nine YR environments (Table 1), while the phenotypic correlation coefficient was 0.90 to 0.93 across the three Mexican environments. Low phenotypic correlation coefficients among Mexico, Kenya, and India were attributed to the presence of different rust isolates in these locations. For LR FDS, the correlation coefficients among the three test environments in Mexico were high, ranging from 0.86 to 0.89 (Table 1). In addition, significant phenotypic correlations, ranging from 0.30 to 0.75, occurred in all environments between LR and YR (Table 1), indicating the presence of pleiotropic/co-located resistance loci in the population.

We identified two co-located resistance loci for YR and LR in the RIL population. The first locus was located on chromosome 1BL and designated QYr.cim-1BL/QLr.cim-1BL. This locus was detected in all tested YR and LR environments and accounted for 10.4–33.3% of YR phenotypic variation and 20.6–33.6% of LR phenotypic variation (Table 2, Figure 3A). Based on the closely linked molecular markers, we developed a KASP marker, such as Kasp_hzau_MuYLr-1BL, and genotyped the entire RIL population (Supplementary Figure 2, Supplementary Table 2). The single marker analysis showed highly significant mean differences of both YR and LR for RILs carrying the positive allele and those lacking it (Supplementary Table 3). Kasp_hzau_MuYLr-1BL was one of the flanking markers of QLr.cim-1BL (Table 2). Because the known pleiotropic multi-pathogen slow-rusting resistance gene Lr46/Yr29/Sr58/Pm39 is also located on 1BL, we genotyped the RIL population with the closely linked molecular marker cslv46G22. The result showed that cslv46G22 was one of the flanking markers of QYr.cim-1BL/QLr.cim-1BL as well.

The second QTL was QYr.cim-2AS/QLr.cim-2AS, located on the short arm of chromosome 2A and which explained 23.3–54.1 and 18.3–25.5% of YR and LR DS variations, respectively. QYr.cim-2AS was detected in all Mexican YR environments and 1 year of the Kenyan environment and was located in the interval of (DArT)-GBS markers 1208841 and 978751 (Table 2, Figure 3) within 2.1 cM from the seedling resistance gene YrMu. We also mapped a leaf rust resistance QTL, QLr.cim-2AS, on 2AS, which was flanked by molecular markers 100033379|F|0_5:A>G, 3952334, 997868 and 1085721 (Table 2, Figure 3C). We developed an InDel marker named InD_hzau_MuYLr-2AS that was genotyped on the F₅ RIL population (Supplementary Table 2); the single marker analysis showed that it was significantly correlated with both YR and LR phenotypes and co-located with YrMu (Supplementary Table 4). Thus, the InDel marker InD_hzau_MuYLr-2AS can be used in wheat breeding to select for QTL QYr.cim-2AS/QLr.cim-2AS.

In addition to the 2 co-located resistance loci mentioned above, we found 4 more resistance loci derived from Mucuy that confer APR to YR, named QYr.cim-3AS, QYr.cim-3BS, QYr.cim-3DS and QYr.cim-6BS, in combination with a locus, QYr.cim-1AL, contributed by Apav#1. Identified at both the Mexican and Kenyan testing locations in 2 years, QYr.cim-3AS was associated with molecular markers 4010188, 1092360, 4989420, 1140071, 1139244, 3937315, 4990593, 3951957, 4398142, 4536273, and 4009657 (Table 2). Its corresponding physical locations on the Chinese Spring (CS) reference genome (International Wheat Genome Sequencing Consortium, 2018) ranged from 6.6 to 10.2 Mb and it explained 3.7–12.6% of the YR phenotypic variation. QYr.cim-3BS was in vicinity of the molecular markers 1051249|F|0_64:T>G, 1128851|F|0_5:C>T, 1056536|F|0_58:A>G, 1004919|F|0_40:G>C, 1315407|F|0_8:G>A, 1318182, 1109710|F|0_29:C>T, and 1076654|F|0_12:T>C, and the corresponding physical locations on the CS reference genome (International Wheat Genome Sequencing Consortium, 2018) spanned from 10.6 to 24.9 Mb. It was detected in 2 years at each of the Mexican and Indian locations and accounted for 3.3–10.7% of YR phenotypic variation (Table 2). QYr.cim-3DS was detected in 2 years at the Kenyan location, explained 3.6–7.7% of YR phenotypic variation, and was linked to markers 1091508, 1215873, 2261207, and 1143346 in physical positions spanning from 78.7 Mb to 90.7 Mb (International Wheat Genome Sequencing Consortium, 2018; Table 2). QYr.cim-6BS was located near markers 2261971|F|0_23:A>C, 1000134|F|0_15:T>C, 1128426|F|022:C>T, 1109468|F|0_15:G>A, 1239693, 1218710, 7353260|F|0_50:A>T and 2261971|F|0_23:A>C, with physical positions ranging from 115.7 to 153.4 Mb (International Wheat Genome Sequencing Consortium, 2018). QYr.cim-6BS was detected for 2 years at the Mexican location and 1 year at the Kenyan location, and it explained 2.5–10.4% of the YR phenotypic variation. QYr.cim-1AL for YR resistance was the only locus identified that was derived from Apav#1. It explained 2.2–14.6% of the phenotypic variation, had a physical position ranging from 565.5 to 580.0 Mb (International Wheat Genome Sequencing Consortium, 2018), and was flanked by molecular markers 1125323|F|0_58:T>G, 22599648|F|0_5:A>G, 1101176, 4989882, 3064615|F|0_6:C>T, 996385|F|0_7:C>G, 989816|F|0_24:A>C, 126500, 987869|F|0_42:A>C and 989816|F|0_24:A>C (Table 2).

We found 2 more resistance loci derived from Mucuy that conferred APR to LR, named QLr.cim-2BS and QLr.cim-5AL. Both loci were consistently identified in all LR environments except for QLr.cim-5AL in LrY17. QLr.cim-2BS explained 9.0–17.5% of LR phenotypic variation and was located on the short arm of chromosome 2B. It was flanked by molecular markers 4989699, 1016414, 1100485|F|0_14:C>T, 1224458, and 1126406|F|0_59:A>G (Table 2). The seedling LR resistance gene Lr16 was also identified on 2BS. The physical location of Lr16 was 13.7–23.9 Mb, based on the CS reference genome (International Wheat Genome Sequencing Consortium, 2018), meaning it overlapped with QLr.cim-2BS at 13.9–23.9 Mb. This confirmed that QLr.cim-2BS and Lr16 should be the same gene that provided all-stage resistance to LR in the RIL population. QLr.cim-5AL was located in the interval of molecular markers 1204040|F|0_64:G>C and 1141498|F|0_63:T>C and explained 4.7–6.9% of LR phenotypic variation (Table 2).

Due to the differences among rust races in different countries, resistance effects provided by individual loci varied greatly at different locations. For example, QYr.cim-2AS provided significant resistance in Mexico, could not be detected in India and was identified in Kenya only in 2016. Therefore, we analyzed the phenotypic effects of QTL combinations among 4 stably detected QTLs according to the average DS in the three countries.

The F₅ RILs were divided into 16 groups according to the genotypes of the 4 stable YR resistance QTL on chromosomes 1Bl, 2AS, 3AS, and 6BS derived from Mucuy. The presence of resistance alleles for the QTL in each RIL was inferred with the QTL combination based on the flanking molecular markers. In Mexico, the disease severity of lines with QYr.cim-2AS ranged from 4.5 to 21.3% (Figure 4A). In addition, QYr.cim-1BL played a great role in reducing YR severity: the mean DS of YR was 63.3% with QYr.cim-1BL present alone, whereas YR DS ranged from 4.5 to 52.7% when QYr.cim-1BL was present with other QTL (Figure 4A). Although the resistance effects of QYr.cim-3AS and QYr.cim-6BS were not significantly different from lines without any resistance QTL, they conferred a significantly lower YR DS when combined with QYr.cim-2AS or QYr.cim-1BL. In Kenya and India (Figure 4A), QYr.cim-1BL showed a higher effect than QYr.cim-2AS, when both were present alone but the presence of both loci reduced DS in the line, overall, with the highest average DS of 40.6%.

The F₅ RILs were classified into 16 groups according to the genotype of 4 LR resistance QTL (Figure 4B). When present alone, QLr.cim-1BL, QLr.cim-2AS, QLr.cim-2BS and QLr.cim-5AL reduced the LR DS from 90.0% to 51.9, 60.3, 84.4, and 76.9%, respectively. Although the resistance effect provided by QLr.cim-2BS alone was small, it had a significant additive effect when combined with other loci. The DS ranged from 2.8 to 27.5% when QLr.cim-2BS was present with other QTL. In general, the number of QTL was negatively correlated with DS for RILs.

So far, chromosome 1AL lacks a formally designated Yr gene. In this study, the only resistance locus derived from susceptible parent Apav#1, QYr.cim-1AL, explained 2.2–5.4% of the YR phenotypic variation in Mexican environments and 3.8–14.6% of the YR phenotypic variation in Kenyan environments; however, no YR resistance conferred by this locus was detected in Indian environments, suggesting that QYr.cim-1AL provides either small race-specific resistance or is environmentally unstable. The physical location range of QYr.cim-1AL on the CS reference genome was 565.5–580.0 Mb (International Wheat Genome Sequencing Consortium, 2018). Several QTL have been reported near this interval (Ren et al., 2012a; Rosewarne et al., 2012; Basnet et al., 2014b). Thus, QYr.cim-1AL might be the same YR resistance locus from Pastor, Naxos, and TAM112, based on their physical positions of the CS reference genome (International Wheat Genome Sequencing Consortium, 2018). This needs to be further confirmed through gene cloning.

A co-located resistance locus QYr.cim-1BL/QLr.cim-1BL was identified on the long arm of wheat chromosome 1B. QYr.cim-1BL/QLr.cim-1BL showed stable and significant resistance effects in all tested environments for both YR and LR. We genotyped the RIL population with cslv46G22, a closely linked molecular marker for Lr46/Yr29/Sr58/Pm39 located on 1BL, which showed that cslv46G22 was loosely linked with QYr.cim-1BL/QLr.cim-1BL as one of the flanking markers. To further verify the relationship among QYr.cim-1BL/QLr.cim-1BL and Lr46/Yr29/Sr58/Pm39, we removed the effect of cslv46G22 and re-did the QTL analysis but were unable to detect any other resistance locus on 1BL. Therefore, we conclude that QYr.cim-1BL/QLr.cim-1BL should be the known APR gene Lr46/Yr29/Sr58/Pm39. Over the last two decades, many CIMMYT derived bread and durum wheat have been reported to possess Lr46/Yr29/Sr58/Pm39: for example, “Pavon 76” (William et al., 2006), “Saar” (Lillemo et al., 2008), “Pastor” (Rosewarne et al., 2012), “Quaiu 3” (Basnet et al., 2013), “Francolin#1” (Lan et al., 2014), “Sujata” (Lan et al., 2015), “Kundan” (Ren et al., 2017), “Bairds” (Lan et al., 2017b), “Chilero” (Ponce-Molina et al., 2018), and Arableu#1 (Yuan et al., 2020). Based on all reported studies, Lr46/Yr29/Sr58/Pm39 should be placed from 662.1 to 684.8 Mb, according to the CS reference genome (International Wheat Genome Sequencing Consortium, 2018). The relatively diffused localization of Lr46/Yr29/Sr58/Pm39 might be due either to the genetic background effect, phenotyping errors, different genotyping platforms, or population size. In addition, recent studies have reported that more than one pleiotropic APR locus could be present in the 1BL region (Yuan et al., 2020; Zhou et al., 2021), indicating that the 1BL region might carry a gene cluster composed of multiple APR genes, which will be confirmed by future gene cloning.

Another co-located resistance locus, QYr.cim-2AS/QLr.cim-2AS, and the seedling YR resistance gene YrMu were located on 2AS (Supplementary Tables 5, 6). Yr17 is located on a translocation on wheat chromosome 2AS derived from the Aegilops ventricosa 2NvS segment, which is known to confer resistance against multiple wheat diseases but also plays role in increasing wheat yields (Gao et al., 2021) and has been used in CIMMYT wheat breeding. This translocation was initially reported to provide significant resistance to all three rusts due to the loci of Yr17, Lr37, and Sr38 (Bariana and McIntosh, 1993; Chen, 2005). However, races virulent to Yr17 rapidly evolved, once this gene was deployed in Europe (Bayles et al., 2000), Mexico (Randhawa et al., 2018), and in the Indian Pst population, due to the widespread cultivation of the Yr17-carrying variety HD2967 during the mid-2010's (unpublished results). However, we detected significant YR resistance on 2AS that was stable over 3 years in Mexico against the Pst isolates used in phenotyping, and this resistance was also identified in Kenya in 2016. PstS1 was an invasive strain that originated in East Africa in the early 1980's; PstS2 evolved from PstS1 and the two strains have become the dominant races in East Africa (Walter et al., 2016). As of 2019, the new genetic group PstS11 was reported as dominant in East Africa by the Global Rust Reference Center and, unlike PstS1 or PstS2, PstS11 is virulent to Yr17. Several reports have indicated that Yr17 might be considered a race-specific APR gene due to the difficulty and inconsistent seedling phenotypes, especially with the aggressive Pst1 lineage (Fang et al., 2011; Milus et al., 2015). In the present study, the seedling reaction of Mucuy was very low, while the single gene line of Yr17 (Avocet+Yr17) was susceptible to the Mexican Pst isolate Mex.14.191. However, YrMu was mapped on wheat chromosome 2AS at 16.6–19.1 Mb (International Wheat Genome Sequencing Consortium, 2018) and it co-segregated with 79 molecular markers including InD_hzau_MuYLr-2AS, WGGB156, and WGGB159, the latter of which were closely linked to Yr17 (Wang et al., 2018). Thus, YrMu should be the known YR resistance gene Yr17 and the enhanced expression of Yr17 in Mucuy seedlings might be due to the background effect of other APR loci, or due to the residual effect of the ineffective Yr17 on APR. Similarly, an LR APR QTL was also located on the translocation, in addition to Lr37, because Pt races MCJ/SP and MBJ/SP used in field trials are known to be virulent to Lr37 (Huerta-Espino et al., 2011).

Several previous studies have shown that chromosome 2BS possesses various race-specific and quantitative resistance loci to LR (Messmer et al., 2000; Xu et al., 2005). So far, six formally named LR resistance genes have been identified on chromosome 2BS, including Lr13 (Seyfarth et al., 1999), Lr16 (McCartney et al., 2005), Lr23 (Datta et al., 2008), Lr35 (Seyfarth et al., 1999), Lr48 (Bansal et al., 2008), and Lr73 (Park et al., 2014). In this study, polymorphism was found in both parents and RILs for Xgwm210, the closely linked molecular markers of Lr16 and the seedling reaction for Mucuy were similar to that of Thatcher+Lr16 against Pt races MBJ/SP. So, by comparing the physical locations, QLr.cim-2BS and Lr16 should be the same gene that conferred LR resistance in Mucuy. Lr16 was a widely deployed LR resistance gene and conferred moderate resistance at both seedling and adult plant stages, and also showed high additive effects with other resistance genes in the field (Lan et al., 2014), which can be considered useful for breeding.

QYr.cim-3AS was detected in both Mexico and Kenya for multiple years, but not in India. Therefore, it was also considered an unstable or race-specific APR gene. So far, only one formally named gene, Yr76 (Xiang et al., 2016), is mapped on 3AS, but QYr.cim-3AS should be distinct from Yr76, given that the latter was a seedling resistance gene. For similar reasons, YrEDWL (Liu et al., 2017) and several YR resistance-linked SNPs (Jighly et al., 2015) were identified on 3AS but are unlikely to be QYr.cim-3AS. In addition, two APR QTL have been mapped on 3AS in the CIMMYT wheat line “Saar” (Lillemo et al., 2008) and the Swiss winter wheat cultivar Arina (Buerstmayr et al., 2014). QYr.cim-3AS might be the same as the two QTLs, according to its physical position based on the CS reference genome (International Wheat Genome Sequencing Consortium, 2018).

There are 4 officially named genes on 3BS: Yr4 (Bansal et al., 2010), Yr30 (Singh et al., 2000b), Yr57 (Randhawa et al., 2015), and Yr58 (Chhetri et al., 2016). QYr.cim-3BS should be different from Yr4 and Yr57 because the latter two provide resistance at the seedling stage and their physical positions are around 3.3 Mb (Xbarc75), which is at least 7.3 Mb away from QYr.cim-3BS. Yr58 starts to express and confer resistance at the four-leaf stage. Although Yr30 is a common APR gene in CIMMYT materials, the approximate physical interval of this gene is in the telomeric region distal to 10 Mb (William et al., 2006; Rosewarne et al., 2012; Basnet et al., 2014a; Lan et al., 2014; Wu et al., 2017; Jia et al., 2020), but QYr.cim-3BS is positioned at 10.6–24.9 Mb. Other YR resistance QTL located on 3BS, considered distinct from Yr30, include QYrhm.nwafu-3BS (Yuan et al., 2018), QYrsk.wgp-3BS (Liu et al., 2019), and Qyrto.swust-3BS (Zhou et al., 2019). The physical locations of QYrhm.nwafu-3BS and QYrsk.wgp-3BS overlapped with QYr.cim-3BS; it is possible that they are the same gene/allele, but this needs further verification. It seems that there is more than one APR gene with significant resistance to YR in the 3BS region, other than just Yr30, but we cannot rule out the possibility of the presence of Yr30 in Mucuy, since Yr30 is not cloned yet.

QYr.cim-3DS was identified only in Kenya for 2 consecutive years and accounted for 3.6–7.7% of the phenotypic variation. Singh et al. (2000b) and Boukhatem et al. (2002) both mapped a QTL on 3DS in the CIMMYT wheat “Opata 85.” The tightly linked marker, Xbcd532, is physically located at least 37.1 Mb away from QYr.cim-3DS. There are several other YR resistance genes identified on chromosome 3DS, such as Yr49 (McIntosh et al., 2017), Yr66 (McIntosh et al., 2013), YrY206 (Zhang et al., 2008), QYr.inra-3DS (Dedryver et al., 2009), YrS1 (Sun et al., 2019), and QYrsn.nwafu-3DS (Huang et al., 2020). Among them, QYrsn.nwafu-3DS provides resistance only in seedlings and YrY206 originated from Aegilops tauschii (Coss.) Schmal. According to the physical location of the flanking markers (International Wheat Genome Sequencing Consortium, 2018), Yr49 (Xgwm161) and YrS1 (Xcfd79) were mapped 71.7 and 65.6 Mb distal to QYr.cim-3DS, respectively. Yr66 (Xgwm341) and QYr.inra-3DS (Xgwm456) are at least 18.8 and 122.9 Mb away from QYr.cim-3DS, respectively. Based on the source of resistance, resistance characteristics, and physical location comparisons, QYr.cim-3DS might be a new YR APR QTL, in Kenyan environments.

QLr.cim-5AL was located on 5AL, where no other LR resistance gene has been officially designated. Rosewarne et al. (2012) detected a QTL on 5AL derived from Avocet that provided APR to LR. The LOD peak of this locus was near wPt-0837, which corresponds to a physical location of 621.6 Mb, based on the CS reference genome (International Wheat Genome Sequencing Consortium, 2018). QLr.cim-5AL detected in the present study was located on 585.2–589.2 Mb and was derived from Mucuy, so the two QTL should be different, given the ~35 Mb gap between them. Recently, Zhang et al. (2019) identified QLr.hebau-5AL flanked by AX-110679506 and AX-110996595, which corresponded to 589.3–591.4 Mb of the CS reference genome (International Wheat Genome Sequencing Consortium, 2018). This locus explained 6.6–7.1% of the variation in the LR resistance response and was derived from a resistant cultivar SW 8588, whose pedigree includes the CIMMYT variety Milan. Therefore, this locus is likely to be QLr.cim-5AL. However, QLr.hebau-5AL also had an effect on YR across four environments in China, whereas no effect of QLr.cim-5AL on YR was identified in the present study. We speculate that this could be due to different Pst isolates present in China and Mexico or a genetic background effect.

QYr.cim-6BS was identified on 6BS. There are three named YR resistance genes on this chromosome: Yr35 (Marais et al., 2003), Yr36 (Uauy et al., 2005), and Yr78 (Dong et al., 2017). Yr35 provides all-stage resistance and was transferred to wheat from T. turgidum ssp. dicoccoides. Yr36 also originated from T. dicoccoides and encodes a kinase-START protein that confers temperature-dependent broad-spectrum resistance (Fu et al., 2009). QYr.cim-6BS is unlikely to be Yr35 or Yr36, based on the source and its resistance characteristics. Yr78 provides APR only in the field (Dong et al., 2017). Comparing the genetic distance and physical position between the corresponding closely linked molecular markers (Somers et al., 2004), several QTL providing APR to YR that map on 6BS could likely be Yr78. These include QYrst.wgp-6BS.1 (Santra et al., 2008), QYr.inra-6B (Dedryver et al., 2009), QYr.caas-6BS (Lan et al., 2010), QYr.caas-6BS.3 (Ren et al., 2012b), and QYrMA.wgp-6BS (Liu et al., 2018). The physical locations of several markers closely linked to Yr78 are included in the range of QYr.cim-6BS detected in this study, such as Xwmc104 (149.1 Mb) and Xbarc136 (151.3 Mb). Therefore, it is likely that QYr.cim-6BS is the same as Yr78, but further verification of this is needed.

The evolution of new virulence and pathogen migration and adaptation to unconventional environments has been observed in the last decade. In addition to the rapid mutation from avirulence to virulence in rust fungi, global climate change and the limited use of resistance genes in complex combinations are important contributors. The average effective life of a race-specific resistance gene is 2–4 years, with the evolution of new races in Mexico. Breeding new varieties with durable resistance is the most effective way to control wheat diseases. Mucuy was distributed for international testing in 2013 and showed high resistance to both YR and LR in Mexico, India, Kenya, and China, suggesting that is a good choice as the donor for introducing resistance into other elite breeding materials. The new resistance loci identified in our study can be further studied to characterize their effects and interactions in other genetic backgrounds and thereby derive the best combinations of effective resistance genes to enhance durability.