English (pdf)
Article in xml format
Article references
Automatic translation
Permalink
1. Introduction
Leaf rust (LR) of bread wheat (Triticum aestivum L.), caused by the fungus Puccinia triticina (P. triticina) Eriks, has been reported as one of the most important diseases in Paraguay1, the Southern Cone of America2 and worldwide3. At present, it causes severe epidemics and losses on an annual basis in Paraguay. Two or more fungicide applications are necessary to control the disease in susceptible cultivars. Yield losses up to 50% were estimated in Alto Paraná Norte and Canindeyú1, where sowing is earlier and epidemics are generally more severe.
There are scarce precedents of characterization of the population of P. triticina in Paraguay. From 23 samples collected in Paraguay during 2011, races TDT-10,20 and MFP were most frequently isolated, and other races identified in smaller proportion were MFP-20, MDT-10,20, TDT-10, TFT-10,20, MDP, MFT-10,20, MFP-10,20, MDP-20. These races have also been identified in Uruguay4, illustrating the similarity of the pathogen population present in Paraguay and Uruguay. Furthermore, the races present in the Southern Cone countries that share the same epidemiological zone east of the Andes are generally similar since there are no geographical barriers that prevent the inoculum from moving from one country to another5.
LR is the main cause for the replacement of commercial cultivars in Paraguay6 and is also considered one of the main reasons for the increase in foliar fungicide applications in the crop, which increases production costs1. However, the main strategy to manage this disease is through genetic resistance7.
Genetic resistance of wheat to LR is conditioned by a high number of genes. Most of the over 80 catalogued Lr resistance genes 8) are major genes expressed from the seedling to the adult plant stage (all-stage resistance, ASR), that produce a hypersensitivity response9. Resistance based on these genes has been widely used by breeders; however, most often it has not been durable since initially resistant varieties carrying one or few ASR genes become susceptible when the pathogen develops new virulent races to these genes10. Some major genes expressed in adult plants, which produce a hypersensitivity response (adult plant resistance, APR-HR, Lr12, Lr13, Lr22a, Lr22b, Lr35, Lr37), have similar characteristics to the ASR genes. Other genes that express in adult plants (APR-PR) have a minor and additive effect, condition quantitative resistance, are race-non-specific, and have been the focus of greater interest because these are presumed to condition durable resistance11. In the field, minor genes determine slow disease development12 and do not express high levels of resistance when present alone. However, the combination of four or five genes confers resistance levels close to immunity11. This resistance has been called partial resistance 13) (PR), adult plant resistance14, and slow rusting 12)(14) . Four genes conditioning APR-PR to LR have been widely studied: Lr3415, Lr4616, Lr6717, and Lr6818. An outstanding feature of Lr34, Lr46, and Lr67 is that they have pleiotropic effects on other pathogens 11)(12)19)(20) 21. Lr34, located on chromosome 7DS, was first described in the Brazilian cultivar Frontana22.
One of the methodologies used to study the genetic resistance conditioned by ASR genes is the postulation of their presence in wheat genotypes based on the reaction to different races of the pathogen. This method is based on the gene-by-gene concept 23)(24) and has been widely used 25)(26) 27 because it is a fast, low-cost, and convenient method for identifying ASR conferred by one or two genes, but may not be appropriate when resistance is more complex. It is not possible to use this methodology when races do not have the virulence combination that determines compatible reaction or susceptibility, nor postulate the presence of APR-PR genes, due to the absence of specific virulence to these genes.
The use of molecular markers is another alternative to postulate or confirm the presence of disease resistance genes28. In the case of wheat LR, there are suitable markers for several resistance genes, for example, major genes Lr1, Lr10, Lr19, Lr21, Lr22a, Lr25, Lr29, Lr32, Lr35, Lr37, Lr39, Lr47, Lr50, Lr51 and APR-PR genes Lr34, Lr46, Lr67, Lr68, Lr75 29)(30) . The Lr34 molecular marker csLV3431 has been used by several researchers 32)(33)34)(35) .
It is important to characterize the resistance and know the genes present in the germplasm of a breeding program to allow identifying sources of resistance with different genes to be introduced to increase genetic variability, as well as better characterizing the leaf rust reaction of commercially used cultivars. This study aims to characterize the field resistance and postulate LR resistance genes present in wheat cultivars and lines from Paraguay.
2. Materials and methods
2.1 Seedling tests
For the postulation of ASR genes, 116 wheat varieties from the Paraguayan Wheat Research Program's Wheat Breeding Program, some of which are introductions from CIMMYT, were evaluated at the seedling stage (Table S1, Supplementary material). Thatcher (Tc) was included as the susceptible control and monogenic differential lines with Lr1, Lr2a, Lr2b, Lr2c, Lr3a, Lr3bg, Lr3ka, Lr9, Lr10, Lr11, Lr14a, Lr14b, Lr16, Lr17a, Lr19, Lr20, Lr21, Lr23, Lr24, Lr26, Lr30 in the Tc background developed in Canada36 and Lr39 (TAM107*3/TA2460), Lr42 (KS91WGRC11=CENTURY*3/T.TAUSCHII), Lr47 (PAVON753) were used to perform the gene postulation. These lines were selected since they represent genes commonly present in improved bread wheat germplasm.
2.2 Races of Puccinia triticina
Nineteen races of P. triticina isolated from samples collected in Uruguay were used in this study (Table 1). Two races were selected for their high frequency in the pathogen population during 2012 (MFP and TDT-10,20)4 and the rest were selected to represent different combinations of avirulence/virulence that allow discriminating the presence of different ASR genes.
The inoculum of these races was preserved in vacuum glass tubes in a refrigerator at 5 ºC in the Rust Laboratory of the Uruguayan National Institute of Agricultural Research (INIA) La Estanzuela. To increase the inoculum of the races, 12 to 15 seeds of the susceptible genotype Little Club (LC) were sown in a 10-cm-diameter pot with a mixture of soil, vermiculite, sand, and substrate (Biofer almácigos, Riverfilco; Biofer Ltd., Montevideo) in a 1:1:1:1 ratio. When plants emerged, each pot was treated with 20 cm3 of a maleic hydrazide solution (0.36 g/l) to stop their development and intensify spore production. Each pot of LC was inoculated with spores of a different race, suspended in Soltrol 170 mineral oil (Phillips Petroleum Co., Borger, TX). Pots were placed in a wet chamber (100% relative humidity) for about 16 hours. Subsequently, pots were moved to the greenhouse with a temperature of 20-25 ºC and six to eight hours of supplementary light (high pressure sodium Son T 400w). To prevent cross-contamination, a PVC cage was placed on each pot and connected to a hose that released a light air flow. Approximately two weeks after inoculation, the inoculum was collected and placed in glass tubes that were vacuum sealed and stored in a refrigerator at 4-6 °C37.
Table 1: Avirulence/virulence formula of Puccinia triticina races used in seedling tests
| Pt racesa | Avirulence / virulence |
| CHT | 1,2a,2b,2c,9,10,19,20,21,24,39,42,47/3a,3bg,3ka,11,14a,14b,16,17a,26,30 |
| DBB-10,20 | 1,2a,2b,3a,3bg,3ka,9,11,16,17a,19,21,24,26,30,39,42,47/2c,10,14a,14b,20,23 |
| KDG-10,20 | 1,3bg,3ka,9,16,17a,19,21,26,30,39,42,47/2a,2b,2c,3a,10,11,14a,14b,20,23,24 |
| LPG-10 | 2a,2c,3a,3ka,16,17a,19,20,21,30,39,42,47/1,9,10,11,14a,14b,23,24,26 |
| MCD-10,20 | 2a,2b,2c,3ka,9,11,16,19,21,23,24,30,39,42,47/1,3a,3bg,10,14a,14b,17a,20,26 |
| MCP-10 | 2a,2b,2c,9,11,16,19,20,21,23,24,39,42,47/1,3a,3bg,3ka,10,14a,14b,17a,26,30 |
| MCR-10 | 2a,2b,2c,3bg,9,16,17a,19,20,21,24,39,42,47/1,3a,3ka,10,11,14a,14b,23,26,30 |
| MCT-10 | 2a,2b,2c,9,16,19,20,21,23,24,39,42,47/1,3a,3bg,3ka,10,11,14a,14b,17a,26,30 |
| MDT | 2a,2b,2c,9,10,16,19,20,21,23,26,39,42,47/1,3a,3bg,3ka,11,14a,14b,17a,24,30 |
| MFP | 2a,2b,2c,9,10,11,16,19,20,21,39,42,47/1,3a,3bg,3ka,14a,14b,17a,23,24,26,30 |
| MFP-10,20 | 2a,2b,2c,9,11,16,19,21,39,42,47/1,3a,3bg,3ka,10,14a,14b,17a,20,23,24,26,30 |
| MFP-20 | 2a,2b,2c,9,10,11,16,19,21,39,42,47/1,3a,3bg,3ka,14a,14b,17a,20,23,24,26,30 |
| MFR | 2a,2b,2c,9,10,16,17a,19,20,21,39,42,47/1,3a,3bg,3ka,11,14a,14b,23,24,26,30 |
| MFT-10,20 | 2a,2b,2c,9,16,19,21,39,42,47/1,3a,3bg,3ka,10,11,14a,14b,17a,20,23,24,26,30 |
| MHP-10 | 2a,2b,2c,9,11,19,20,21,23,24,39,42,47/1,3a,3bg,3ka,10,14a,14b,16,17a,26,30 |
| MKD-10 | 2a,2b,2c,3ka,9,11,19,20,21,23,30,39,42,47/1,3a,3bg,10,14a,14b,16,17a,24,26 |
| MMD-10,20 | 2a,2b,2c,3bg,3ka,11,16,19,21,23,24,30,39,42,47/1,3a,9,10,14a,14b,17a,20,26 |
| SPG-10 | 3a,3bg,3ka,16,17a,19,20,21,30,39,42,47/1,2a,2b,2c,9,10,11,14a,14b,23,24,26 |
| TDT-10,20 | 9,16,19,21,26,39,42,47/1,2a,2b,2c,3a,3bg,3ka,10,11,14a,14b,17a,20,23,24,30 |
aLong & Kolmer, 1989. The inclusion of 10 and/or 20 after the denomination of the race indicates virulence to these genes.
2.3 Postulation of all-stage resistance genes
Twenty-eight genotypes (six to eight seeds per genotype) were planted in 45.5 × 28.0 × 7.7 cm pots filled with the abovementioned substrate. The susceptible control Tc was included in each pot. The genotypes were inoculated with the 19 races individually, following the procedure described for the increase of races' inoculum. At 12 days after inoculation, the infection type (IT) was evaluated according to the scale described by Stakman and others38, where IT 0 = immune response, without uredinia or necrosis; IT; (fleck) = necrotic lesions without sporulation; IT 1 = small uredinia surrounded by necrosis; IT 2 = small uredinia surrounded by chlorosis; IT 3 = moderate uredinia without chlorosis or necrosis; IT 4 = large uredinia without chlorosis or necrosis. The symbols + and - were used to indicate larger and smaller uredinia compared to the typical IT, respectively. IT X = a mesothetic response of flecks, small and large uredinia. IT 0-2+ and X were considered low ITs and IT 3-4 were considered high ITs.
Two replications were planted for each genotype and race. For avirulent ITs, if the ITs differed by less than one IT unit between replications, the highest IT was considered. If the ITs differed by more than one IT unit, the data was discarded (n/a). For intermediate ITs, if one replicate had low IT and the other had high IT, the information was also discarded.
To postulate which resistance genes are probably present in the genotypes, the IT patterns of the Paraguayan genotypes were compared with the IT of the lines carrying unique LR resistance genes.
2.4 Adult plant (field) resistance
Paraguayan genotypes were evaluated under field conditions and natural infection of the pathogen during the winter-spring of 2012 at INIA La Estanzuela, Colonia Department (LE: latitude 34.3° S, Longitude 57.7° W, elevation 70 masl) and at Young, Río Negro Department (Y: latitude 32.7° S, longitude 57.6° W, elevation 76 masl). Planting dates were July 12 at Young and July 24 at La Estanzuela.
The experimental design used was incomplete randomized blocks with two replications. Tc and Avocet, used as susceptible controls, were repeated six times in each replication. Plot size was two one-meter-long rows. Spreader rows (mixture of different susceptible genotypes) were planted perpendicular to the plots to homogeneously increase the natural LR inoculum.
LR severity and reaction were assessed starting at stem elongation and approximately every two weeks (four times in La Estanzuela, and three times in Young). LR severity was determined according to the modified Cobb scale39. Small uredinia surrounded by distinct necrosis were considered resistant (R); moderate-large sized uredinia surrounded by necrosis were considered moderately resistant (MR); moderate to large uredinia surrounded by chlorosis were considered moderately susceptible (MS); large uredinia lacking necrosis or chlorosis were considered susceptible (S); and a mixture of large and small uredinia were considered as a mixed (M) response. The coefficient of infection (CI) was calculated as severity × reaction, using a coefficient for each reaction: R = 0.2, MR = 0.4, M, MRMS or MSMR = 0.6, MS = 0.8, and S = 1.0.
The area under the disease progress curve (AUDPC)40 was calculated based on the LR CI. The adjusted AUDPC calculation and the ANOVA were performed using a mixed linear model with the package lme441 in the software R42, using the following model:
Y ijkl = G i + E j + R k(j) + B l(jk) + ε ijkl
where Y: LR AUDPC values, G: effect of genotype i-th (fixed), E: effect of j-th location (fixed), R: repetition within location (random), B: incomplete block within location and repetition (random), and ε: experimental error with iid N (0, σ2 ε).
The genotype × location effect was disregarded and included in the experimental residuals since we were interested in the expression of resistance over locations. Based on this, the adjusted mean AUDPC across locations and the minimum significant difference (MDS, P < 0.05) was calculated.
The adult plant reaction of the genotypes recorded in the field was classified into different categories based on the values of the adjusted AUDPC. According to the distribution of the AUDPC of all genotypes and comparing these with the AUDPC of the susceptible checks, those genotypes with values in the range of 0-375 were considered resistant (R), 376-780 moderately resistant (MR), 781-1000 moderately resistant to moderately susceptible (MRMS), and 1001-1500 moderately susceptible (MS).
2.5 Confirmation of the presence of Lr34 based on the csLV34 molecular marker
The molecular marker csLV3431 was used to postulate the presence of the APR-PR Lr34 gene. DNA extraction, PCR amplification and determination of alleles in agarose gel electrophoresis were performed according to CIMMYT protocols43. The wheat line Parula was used as the positive control for the expected band of the allele associated with the presence of Lr34.
3. Results
In seedling tests, the susceptible control Tc had high IT (3 to 4). It was not possible to postulate the ineffective (Lr14a, Lr14b) and effective genes (Lr19, Lr21, Lr39, Lr42, Lr47) to all races. The rest of the tested genes had high and low IT to different races (Table S2, Supplementary material). The seedling IT information of the studied genotypes is presented in three tables, according to their reaction pattern to the races: resistant to all races (Table 2), resistant to the most frequent races (Table 3), and susceptible to one or both most frequent races, MFP and TDT-10,20 (Table 4).
Average field infection of susceptible controls Avocet and Tc was high (final infection severity of 99% and 90%, adjusted AUDPC of 5196 and 4746, respectively). The infection of both susceptible controls was consistently high over locations and reps. The phenotypic correlation of AUDPC between locations was 0.96. The estimated AUDPC of the genotypes ranged from 0 to 1438 (Table 2, Table 3, and Table 4) and these values were significantly lower than the AUDPC of the susceptible controls (MDS0.05 1163). The percentage of field R, MR, MRMS and MS genotypes was 61, 22, 11 and 6, respectively.
Forty-five Paraguayan genotypes were resistant in the seedling stage to the 19 races (Table 2), which did not allow identifying the genes expressed at that stage. Their AUDPC in the field ranged from 0 (R) to 994 (MRMS).
The presence of 12 ASR genes: Lr1, Lr2(b, c alleles), Lr3(a, bg, ka alleles), Lr9, Lr10, Lr11, Lr16, Lr17, Lr23, Lr24, Lr26, and Lr30 was postulated in 46 genotypes (Table 3 and Table 4). Additional resistance that could not be identified was present in many of the genotypes.
Table 2: Seedling infection type to 19 Puccinia triticina races, AUDPC, field reaction and presence of the csLV34 marker in Paraguayan genotypes resistant to all races
a+ presence, - absence of the allele associated with Lr34. H: heterozygous. bn/a: not available information
Table 3: Postulated genes, seedling infection types to 19 Puccinia triticina races, AUDPC, field reaction and presence of the csLV34 marker in Paraguayan genotypes resistant to prevalent races (MFP and TDT-10,20)
a+ presence, - absence of the allele associated with Lr34. H: heterozygous. b+: presence of additional unidentified resistance gene(s). cn/a: not available information
Table 4: Postulated genes, seedling infection types to 19 Puccinia triticina races, AUDPC, field reaction and presence of the csLV34 marker in Paraguayan genotypes susceptible to one or both prevalent races (MFP and TDT-10,20)
a+ presence, - absence of the allele associated with Lr34. H: heterozygous. b+: presence of additional unidentified resistance gene(s). cn/a: not available information
The number of genes postulated in each genotype ranged from 1 to 7. One gene was postulated in 14 genotypes, seven with Lr23, three with Lr10 and one with Lr3a, Lr3bg, Lr26 or Lr30; all these genotypes had additional resistance conferred by ASR genes which was not possible to identify (+). Two genes were postulated in 12 genotypes, three genes were postulated in four genotypes, four genes were postulated in eight genotypes, five genes were postulated in five genotypes, six genes were postulated in one genotype, and seven genes were postulated in two genotypes. Other 25 genotypes only possessed seedling resistance that could not be identified with the races used in the study.
Most genotypes have different genes or gene combinations, although some might carry the same or similar resistance base. Genotypes 32, 46, and 72 probably have Lr10, Lr11, Lr23, and Lr24, while genotypes 24 and 67 could carry Lr26 in addition to these genes; genotypes 1 and 96 probably have Lr23 and Lr26; genotypes 20 and 94 probably have Lr23 and Lr30; genotypes 50 and 82 might share Lr3(a or bg), Lr17a, and Lr26; genotypes 26 and 58 could share Lr3(a, bg or ka), Lr17a, Lr23, and Lr30; genotypes 50 and 82 probably share Lr3(a or bg), Lr17a and Lr26; genotypes 26 and 58 might share Lr3(a, bg or ka), Lr17a, Lr23, and Lr30, and genotypes 6, 18, 47 and 115 were postulated to have Lr3a and Lr10.
Thirty-five genotypes were resistant to the races prevalent during 2012 (MFP and TDT-10,20) 4) and were susceptible to at least one of the other races tested (Table 3). These genotypes had a range of AUDPC from 13 (R) to 1438 (MS). The 36 seedling S genotypes to MFP and/or TDT-10,20 had a range of AUDPC from 64 (R) to 1321 (MS) (Table 4).
From the 12 genes postulated in Paraguayan genotypes, Lr23 was the most frequent, postulated in 28 genotypes. Other genes frequently present in the genotypes were Lr10, Lr26 and Lr3 (jointly considering its three alleles). Lr11, Lr24, Lr30, Lr1, Lr17a, Lr2 (b or c) were postulated in progressively fewer genotypes, while Lr9 and Lr16 were postulated in only one genotype (Figure 1).
Based on the csLV34 molecular marker Lr34 was postulated in 31%, absent in 60% and heterozygous in 9% of the genotypes (Table 5). Lr34 was postulated in 20% of all-race-resistant genotypes in the seedling stage, 54% of prevalent race-resistant genotypes, and 22% of seedling susceptible genotypes to one or both prevalent races. Lr34 was postulated in 31% of field R genotypes, 32% of MR genotypes, 31% of MRMS genotypes, and 29% of the MS genotypes.
4. Discussion
The postulated ASR genes Lr1, Lr2(a, b alleles), Lr3(a, bg, ka alleles), Lr9, Lr10, Lr11, Lr16, Lr17, Lr23, Lr24, Lr26, and Lr30 were detected alone or in combinations of up to seven genes per genotype. From the three reported Lr2 alleles (Lr2a, Lr2b, Lr2c)44, two (Lr2b and Lr2c) were postulated in only one genotype each. All Lr3 alleles (Lr3a, Lr3bg, Lr3ka)45 were postulated in some genotypes, although these could not be differentiated in some cases. The methodology used in this study is not very precise for genotypes with complex resistances27. Other studies have postulated up to five rust resistance genes in wheat genotypes.
As expected, genotypes derived from the same cross or closely related generally had similar basis of resistance: genotypes 24, 25 and 72 (E-92225/FCEP30) share Lr10 and Lr24, while 24 and 72 also have Lr11 and Lr23; genotypes 96 and 97 (ITAPUA40/3/ITP35/PF84432//CORD4) have Lr23, and 96 also has Lr26; genotypes 12 and 66 (ITAPUA40/CARCOVE//JUP*5/AMIGO) share Lr11, Lr23 and Lr26, while 66 also has Lr1 and Lr24; genotypes 32, 33, 46 and 67 (ITAPUA45/CORDILLERA4) share Lr10, Lr11, Lr23, Lr24, while genotypes 33 and 67 also have Lr26; genotypes 48 and 58 (MILAN/KAUZ//PASTOR/3/PASTOR) share Lr3(a or ka), Lr23 and Lr30; genotypes 41 and 90 (PARULA/IAN10) had Lr23. All genotypes derived from ITAPUA 40 had either Lr23 or Lr26, or both genes.
Most of the studied genotypes were introduced from CIMMYT or derived from local crosses with lines of this origin or from Brazil. Both origins have traditional regional cultivars in their foundation germplasm. Therefore, the resistance genes identified in the studied genotypes most probably come from CIMMYT germplasm and/or regional breeding programs. Eight of the postulated genes in the studied genotypes were previously reported in CIMMYT germplasm (Lr1, Lr3a and Lr3bg alleles, Lr10, Lr16, Lr17a, Lr23, Lr24, Lr26) 9)(46)47)(48) . Lr1, Lr3a, Lr10, Lr16, Lr17a, Lr24, and Lr26 were reported in cultivars and lines from Argentina, Brazil and Uruguay 2)(25)33)(49) . Lr3bg and Lr23 are present in genotypes from Brazil and Uruguay 2)(25) 49. Lr3ka was first reported in the Argentinean cultivar Klein Aniversario45 and has also been postulated in genotypes from Argentina and Brazil 33)(49) . Lr9, postulated in a single genotype, was transferred from Aegilops umbellulata to wheat50. This gene is not widely distributed in breeding germplasm but it was reported in some Brazilian and Argentinean cultivars 33)(49) . Lr11 was described in the Argentinean cultivar El Gaucho51 and was reported in some Brazilian cultivars49. Lr30 was initially described in the Brazilian cultivar Terenzio52. The origin of Lr2b and Lr2c could not be traced back to known sources53.
The seedling resistance that could not be identified probably corresponds to uncatalogued genes or known genes for which no genetic stocks were available. It is possible that the complementary genes Lr27+31, which have a low IT of X = to X+53, are present in the tested germplasm, since some genotypes showed this characteristic IT and these genes are present in CIMMYT germplasm 9)(47) 48. Another reason why the genes could not be identified probably relies on the genetics of the pathogen, since the races used might not possess the adequate virulence combination for the postulation.
It was not possible to postulate genes in 45 genotypes that were resistant to all races (Table 2 and Table 3), since by using this methodology only the presence of genes or combinations of genes that are ineffective to one or more races can be postulated. Some genotypes probably possess the resistance genes present in some of their parents. Genotypes 21, 35, 36, 37, and 38 probably possess Lr42, as they are derived from one parent with this gene. Lr42 was transferred from Triticum tauschii to wheat54; it is found in synthetic wheat and is present in modern CIMMYT lines. Genotype 37 is derived from a cross of parents with Lr42 and Lr47 (BABAX/LR42//BABAX*2/3/PAVON 7S3+LR47) and probably has Lr47, since it expressed very low IT to all races similar to its respective single gene line, or both genes. Lr47 was transferred from Triticum speltoides 55) and was effective to all races. Resistance in other genotypes in this group could be conferred by combinations of catalogued or uncatalogued ASR genes or by single Lr genes effective to the pathogen population of the region, and, therefore, valuable for LR resistance breeding. To identify or confirm the resistance genes present in these genotypes, it is necessary to use other methodologies as available molecular markers for ASR genes29, confirm the presence of the genes through allelism tests or perform genetic analysis to map the resistance.
The high, uniform, and consistent LR infection achieved in the experimental fields at both locations, indicated by the high phenotypic correlation between locations, allowed a reliable characterization of the resistance of Paraguayan genotypes in Uruguay in 2012. All genotypes evaluated in the field had LR infection significantly lower than the susceptible controls, expressing a degree of resistance from R to MS. Most genotypes of all categories (Table 2, Table 3 and Table 4) were R or MR (83%) and a low proportion was MRMS and MS (17%), which indicates that an effective selection for LR resistance was accomplished. Based on the phenotype, it was not possible to infer the presence of resistance expressed in adult plants in lines with seedling resistance to all races (Table 2), or resistant to the most frequent races (Table 3), since effective ASR masks the presence of APR-PR27.
However, using the molecular marker csLV34 APR-PR, the presence of the Lr34 gene was postulated in 31% of all genotypes. The csLV34 marker used to postulate the presence of Lr34 has also been used by other researchers 32)(33)34)(35) . Although it is not a perfect marker, it is very close to Lr34 (0.4 cM)31 and it is considered to have a high quality diagnostic power34. Lr34 was first described by Dyck and others22 in the Brazilian cultivar Frontana22 and it is present in the wheat germplasm from CIMMYT 11)(12) 56 and the region 2)(5)25)(33) . The presence of Lr34 in the regional germplasm is a relevant contribution to the control of wheat diseases since it has conferred durable resistance to LR and has a pleiotropic effect for resistance to stripe rust (Yr18)21, stem rust (Sr57)57, powdery mildew (Pm38) 19)(20) and spot botch (Sb1) 12)(58) .
Lr34 was postulated in 20% of the genotypes resistant to all races (Table 2) and in 54% of the genotypes resistant to the two prevalent races (Table 3). The significance of the presence of Lr34 in the germplasm with ASR relies on the enhanced resistance it confers when combined with major genes59. Additionally, the presence of Lr34 could help mitigate future losses if new virulent races that affect genotypes with ASR to all races emerge, or if low frequency virulent races to genotypes resistant to the most prevalent races increase in frequency.
Genotypes that in the seedling stage were susceptible to the most frequent races (Table 4) most probably possess resistance expressed in advanced stages of development. MFP and TDT-10,20, the most frequent races identified in the 2012 survey4, were also the predominant races at La Estanzuela Experimental Station (17% and 32% of 65 analyzed samples, respectively) and at Young experimental field (20% and 34% of 79 samples, respectively). These two races are virulent to Lr12, Lr13 and Lr37 APR-HR genes, frequent in the regional germplasm5. Therefore, these genotypes’ field resistance is most probably related to the presence of APR-PR genes that confer PR. APR-PR genes do not express high levels of resistance on their own, however, the combination of four or five APR-PR genes confers resistance levels close to immunity11. Lines under field conditions with MS to R response would have a progressively greater number of APR-PR genes. Silva and others reported the presence of Lr34 and Lr68 APR-PR genes in line 59 (SUZ6/OPATA)60, which is consistent with the presence of the Lr34 molecular marker and R field reaction of this genotype.
Lr34 was postulated in 22% of the genotypes with APR-PR to LR (Table 5). According to the levels of resistance expressed in the field, other APR-PR genes would also be present in addition to Lr34, since this gene expresses moderate levels of resistance when present alone. Lr34 is probably present alone in entries with csLV34 that expressed MRMS field reaction (Table 4); therefore, higher levels of field resistance would be explained by the presence of additional APR-PR genes. Furthermore, there are a considerable number of genotypes with R to MS field reaction which lack csLVLr34, indicating the presence of APR-PR genes other than Lr34. Diversity in this characteristic is desirable to achieve high and stable PR levels. It would be desirable to confirm the presence and/or introgress other genes that confer this type of resistance12, such as Lr46, Lr67, Lr68, and other QTLs associated with PR that have not yet been characterized61. The introgression of Lr68 in Paraguayan germplasm would be especially interesting since this gene has a greater effect than Lr34 in some South American countries 32)(62) . Molecular markers could be used to accelerate the introgression of PR in the Paraguayan germplasm.
When using major resistance genes, it is essential to know which genes are present in the germplasm to combine them with different effective resistance, increasing the diversity and possible duration of the resistance. Lr42, possibly present in entries 21, 35, 36, 37, and 38, and Lr47, possibly present in entry 39, are effective to the leaf rust population in the region. These genes are not present in the traditional Paraguayan germplasm, and could be used to develop new resistant cultivars. While the presence of race-specific ASR genes is common in cultivars used by farmers and in the germplasm of wheat breeding programs, the use of APR-PR race-non-specific gene combinations that confer PR is the best alternative to achieve high levels of durable resistance. This decreases the need for LR chemical control, being environmentally and economically friendly, and it could also allow for better integrated disease management.
Before designing the resistance improvement strategy in the Paraguayan Wheat Improvement Program, the field resistance of the genotypes should be confirmed locally, since environmental conditions can affect the expression of PR62 and the resistance conferred by major genes63. The pathogen’s diversity can affect the effectiveness of the resistance of genotypes with major genes. While there is evidence of similarity in the population of P. triticina in the epidemiological zone comprising Uruguay and Paraguay, the pathogen’s population is highly variable both spatially and temporally, including many races during each growing season5, which also indicates the need to further test the genotypes locally and under the current pathogen population.
In conclusion, this study demonstrates that there is a significant number of genotypes with APR-PR to LR as a foundation to expand diversity and accumulate additional genes, to achieve stability of cultivars' field resistance and the pathogen population. Alternatively, genotypes with resistance conferred by ASR genes effective to all races of the pathogen may be used in combination with PR genes, to avoid high losses in case new virulent races of the pathogen to the ASR emerge. If only genotypes with effective ASR genes are used, it is advisable to increase the genetic diversity introducing other effective genes, and their strategic deployment should be associated with ongoing pathogen monitoring, as an essential tool to take early measures such as alerts for disease control and replacement of affected varieties if new virulent races are detected.
