• differential set;
  • diversity;
  • leaf rust;
  • SSR;
  • Triticum durum;
  • virulence


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

No internationally agreed differential set is available for characterization of virulences in populations of Puccinia triticina causing wheat leaf rust on durum wheat. In a first step, 73 potentially differential host genotypes were tested with 96 durum leaf rust isolates collected in France. A differential set, adapted to the local epidemiological context and useful for comparison with international studies was selected, including French commercial cultivars, Thatcher lines with Lr genes, and international cultivars. In the second step, a sample of 310 isolates collected in France from 1999 to 2009 was characterized on this set. Diversity was very low, as only five pathotypes were distinguished. Genotyping of a subset of 76 isolates according to 20 SSR markers confirmed this low diversity, with 73 isolates belonging to a single dominant genotype. Population was strongly shaped by cultivars, and the findings explain the successive breakdown of resistance sources deployed in French durum wheat cultivars. The gene Lr14a, suggested to be an efficient source of resistance in several European and American countries, was overcome by pathotypes frequent in France since 2000. Postulation of resistance genes in the commercial cultivars led to a proposed simplified version of the differential set. This study, providing new information about leaf rust resistance genes present in the French durum wheat germplasm, highlights the need to diversify sources of resistance to P. triticina in this germplasm. The results are also discussed in terms of relatedness and intercontinental migration of P. triticina on durum wheat.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Durum wheat is grown on approximately 500 000 Ha in four distinct regions of France. The annual production averages 2 million tons. Between 1994 and 2010, France exported 75% of its yearly production, and in 2010, it was ranked the second exporter of durum wheat in the world (Anonymous, 2011). Whenever weather conditions are favourable, severe wheat leaf rust epidemics occur on susceptible cultivars. Although the spraying of fungicides can bring this disease under control, breeders are seeking durable genetic resistance, especially now that national as well as European Union objectives have been set to curb fungicide use for both environmental and economic reasons.

Up until the end of the 20th Century, knowledge about the genetic basis of resistance to leaf rust in durum wheat, in comparison to bread wheat, was relatively limited (Kolmer, 1996). Severe leaf rust epidemics on durum wheat in northwestern Mexico in 2001 and thereafter, due to the breakdown of the most popular Mexican cultivar Altar C84 (Singh et al., 2004), emphasized the need for resistant germplasm and a better understanding of its genetic basis (Herrera-Foessel et al., 2005). In terms of the host population, it is not generally known if durum wheat shares the same resistance genes as bread wheat (Ordonez & Kolmer, 2007b). In terms of the pathogen populations, differences between populations of Puccinia triticina growing on bread and durum wheat were first pointed out in the early 1990s (Singh, 1991; Huerta-Espino & Roelfs, 1992). Anikster et al. (1997) established that the same species, P. triticina, with Thalictrum sp. as an aecial host, was found on both durum and bread wheat. An additional species, P. recondita, can be found on durum wheat, with Anchusa italica as an aecial host, but up to now it has only been identified in limited areas in Morocco (Ezzahiri & Diouri, 1992), Portugal (D’Oliveira & Samborski, 1964) and southern Italy (Casulli et al., 1996). Further studies (Martinez et al., 2005; Goyeau et al., 2006; Ordonez & Kolmer, 2007b) have confirmed that the differences between populations growing on bread and durum wheat were such that the differential set used for bread wheat leaf rust was no longer adequate to differentiate virulences within the durum wheat leaf rust populations. Today, there is still no internationally agreed set of lines to use as differentials for durum wheat leaf rust.

Given the limited knowledge about resistance genes to leaf rust in durum wheat, most race surveys of durum wheat leaf rust populations have been conducted using the isogenic Thatcher (Tc) lines with Lr genes developed for the study of bread wheat leaf rust populations (Singh et al., 2004; Ordonez & Kolmer, 2007b; Mantovani et al., 2010). As one of the main purposes of a race survey is to monitor changes in virulence that affect commercial cultivars and breeding programmes, the genes for resistance that are present in commercial cultivars should be represented in the differential host series (Samborski, 1985; Knott, 1989). In their studies of Mexican populations, Singh et al. (2004) completed the set of Tc isogenic lines with cultivars Gaza and Altar C84.

The objective of the present study was to accurately characterize the pathogen population of P. triticina relative to the sources of resistance present in durum wheat cultivars grown in France, and its evolution from 1999 to 2009. To achieve this objective, a preliminary step aiming to establish a differential set was required. Genotyping of the durum leaf rust isolates was also undertaken to investigate whether or not it would provide additional information to virulence phenotyping. Once the characterization of the durum wheat leaf rust pathotypes had been achieved, it was possible to postulate the resistance genes of the commercial cultivars included in the differential set, which led to the proposal of a simplified version of this set. The results are discussed in terms of relatedness and intercontinental migration of P. triticina on durum wheat.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

To establish a differential set, 73 host genotypes were inoculated with a sample of 96 leaf rust isolates. This step led to the selection of a set comprising 28 host lines and cultivars. In order to describe the population of P. triticina, a total of 310 isolates (including the 96 mentioned above) were then phenotyped at the seedling stage on this differential set. A further characterization of the population included genotyping with SSRs, and phenotyping at the adult stage. Postulation of resistance genes in the commercial cultivars of the differential set was achieved through the analysis of infection types.

Spore production

Wheat seedlings were grown in air-filtered cabinets in a glasshouse at temperatures between 15 and 25°C for 7 days, using a 14 h photoperiod (daylight supplemented by 400 W Na-lamps). Up until 2002 cv. Michigan Amber was used. After that date, cvs Local Red and/or Lebanon were used depending on the availability of seeds. Seedlings were treated with 20 mL of a maleic hydrazide solution (0·25 g of maleic hydrazide per litre of H2O) to prevent emergence of secondary leaves and increase the size of uredinia. Seven-day-old seedlings were inoculated by gently wiping them with infected field-collected leaves. Inoculated seedlings were placed in a dew chamber at 15°C for 24 h, and then put back in the glasshouse. One week after inoculation, the seedlings were trimmed so that only one plant with one uredinium remained in each pot. Cellophane bags were placed over the pots to prevent contamination between isolates. Ten days after inoculation, spores from a single uredinium were collected in a 00 gelatin capsule using a cyclone collector. The quantity of spores was increased by adding 0·5 mL of light mineral oil to each capsule, and spraying the resulting spore suspension onto 7-day-old Michigan Amber, Local Red or Lebanon seedlings. After 10 more days, spores were collected and divided into two batches, one to inoculate a differential set and one to keep in storage at −80°C.

Establishing a differential set

A sample of 96 isolates, collected between 1999 and 2007, was selected to inoculate the differential lines to be tested. These isolates were chosen to maximize the diversity of cultivars, years and regions of collection (Fig. 1). Three categories of potential differential lines were considered: (i) 17 lines and cultivars (Acme, Altar C84, Arnautka, Bordos-61, CD 40473, Entrelargo de Montijo, Gaza, Golden Ball, Karkov, Kubanka, Langdon, Lebanon, Local Red, Medea Ap 9d, Medora, Peliss, Yavaros, seeds obtained by courtesy of Jim Kolmer and Ravi Singh) previously used in other P. triticina virulence studies (Huerta-Espino & Roelfs, 1989, 1992; Singh et al., 2004; Ordonez & Kolmer, 2007b); (ii) Lr genes, in 29 Thatcher isolines (Lr1, Lr2a, Lr2b, Lr2c, Lr3, Lr3bg, Lr3ka, Lr9, Lr10, Lr11, Lr13, Lr14a, Lr14b, Lr15, Lr16, Lr17, Lr19, Lr20, Lr21, Lr23, Lr24, Lr25, Lr26, Lr29, Lr30, Lr32, Lr33, Lr38, Lr44); and (iii) 26 cultivars registered in France, reported by breeders or extension services (Arvalis) to be resistant, or having shown a change in their susceptibility in the field over the last 10 years (Acalou, Akenaton, Amarillo, Amosis, Arcalis, Artimon, Biensur, Brennur, Byblos, Claudio, Combo, Dakter, Joyau, Karur, Latinur, Lloyd, Nautilur, Nefer, Neodur, Orjaune, Orlu, Orobel, Pescadou, Reaumur, Sachem, Silur). Untreated seeds of the lines and cultivars were sown in pressed-peat pots (3 × 3 cm2) containing a commercial compost (peat substrate, Gebr. Brill Substrate), with four seedlings per pot and three pots per line (12 seedlings for each cultivar). Spore suspension (3 mg of spores per millilitre of mineral oil) was sprayed on the first leaves of each cultivar, using 0·7 mL per tray of 240 seedlings. After inoculation, the sets were placed in a dew chamber in the dark at 15°C for 24 h, and then in a glasshouse with a 16 h photoperiod (daylight supplemented by 400 W Na-lamps) at 18/22°C for 9 days. Infection types were scored according to the 0–4 infection type (IT) scoring system (Stakman et al., 1962): IT 0 = no visible symptoms, IT ; = hypersensitive flecks, IT 1 = small uredinia with necrosis, IT 2 = small to medium sized uredinia with green islands and surrounded by necrosis or chlorosis, IT 3 = medium to large sized uredinia with chlorosis, IT 4 = large uredinia without chlorosis, IT X = heterogeneous ITs, evenly distributed over a given leaf, and IT Y = heterogeneous ITs with larger uredia toward the tip. Plus and minus signs are used to indicate variation in ITs, and the letters C and N are used to indicate more than normal chlorosis or necrosis, respectively. Relevant lines were sorted out to constitute the differential set, which included only the lines resistant to at least one isolate. The lines susceptible to all the isolates, i.e. displaying a high IT 3 to 3+, were dropped, except three lines kept as susceptible controls.


Figure 1.  Areas of durum wheat cultivation in France (Ha per state in 2009), as well as distribution across year and area of collection of the 96 isolates selected for the establishment of a differential set. Overall, the isolates were collected from 30 different cultivars.

Download figure to PowerPoint

Phenotyping the population for virulences

The French population was represented by 310 isolates collected between 1999 and 2009. Most of these isolates were collected on three cultivars (Table 1): Neodur, susceptible, registered in 1987; Nefer, registered as resistant in 1997 but which became susceptible around 2001; and Biensur, registered in 2001 as resistant and the most widely grown cultivar in France from 2005 to 2008. Samples from these three cultivars represented 49% of the 310 single-pustule isolates analysed; the remaining 51% were collected from various other cultivars. Isolates were phenotyped by inoculation of the above given differential set, using the described methods. This led to the identification of five different pathotypes.

Table 1.   Year and cultivar of origin for 310 durum leaf rust isolates collected in France and characterized for their virulence phenotype
Nefer  4141 111145555
Biensur  1 32511111539
Pescadou  1   11 51018
Orjaune35111111 3 17
Joyau    111145417
Dakter      1 55516
Karur    1 12 5110
Lloyd22111111   10
Miradoux         189
Other cvs46841157647466


Given the clonal reproduction of P. triticina in France (Goyeau et al., 2007), a subsample of isolates including the five pathotypes identified with the differential set was considered as representative of the genetic diversity of the population. A subset of 76 isolates distributed among the five pathotypes (see Table 4) was genotyped using 20 microsatellite markers: RB8, RB11, RB12, RB17, RB25, RB26 and RB29, developed by Enjalbert et al. (2002), and PtSSR13, PtSSR50, PtSSR55, PtSSR61, PtSSR68, PtSSR91, PtSSR92, PtSSR152, PtSSR154, PtSSR158, PtSSR164, PtSSR173 and PtSSR186, developed by Szabo & Kolmer (2007). DNA extraction was carried out by shaking, in a FastPrep shaker (FP24; MP Biomedicals, Holbrook), each sample of 10 mg of urediospores with a 250 mg matrix (silica) and a ceramic bead in 400 μL of AP1 Buffer supplied in the QIAGEN kit. The OmniPrep genomic DNA extraction kit (GenoTech) was used according to the manufacturer’s instructions. DNA concentration was determined using a NanoDrop ND-1000 spectrometer (Labtech). A final concentration of P. triticina DNA at 2 ng μL−1 was used for DNA amplification. The primers were combined into two multiplex reactions. Amplification of DNA was performed with an Applied Biosystem 9600 Thermal Cycler, with a temperature profile of 95°C for 5 min; 32 cycles of 95°C for 30 s, 58°C for 1 min 30 s and 72°C for 30 s; and 30 min of final extension at 60°C. Primers were labelled (WellRED- PROLIGO) with three fluorochromes (D2, D3 and D4, respectively black, green and blue) and used for each multiplex reaction. Each reaction (20 μL) contained 4 μL of DNA solution (approximately 10 ng μL−1), 10 μL of QIAGEN Type-it microsatellite PCR kit Mix (2 × ), 2 μL of Q-solution (5 × ), 2 μL of the primer mix (2 μm each), and 2 μL of H2O. PCR products were diluted to obtain between 1 and 2 ng of amplified fragments to prevent signal saturation. Capillary electrophoresis and fragment size determination were performed on a CEQ 8000 Genetic Analysis System (Beckman Coulter).

Characterizing interactions at the adult stage

The expression of rust resistance genes at the adult stage can be different from that observed at the seedling stage. Interactions between the pathotypes and the lines/cultivars selected as differentials were thus further characterized at the adult stage, in the glasshouse for the five pathotypes identified, and in the field for one of those pathotypes. In the glasshouse, individual plants of the differential lines were grown in pots filled with 0·7 L of commercial compost (Klasmann Substrat 4, Klasmann France SARL) to which 3·0 g of slow-release fertilizer (Osmocote 10-11-18 N-P-K) were added. The plants were vernalized at the two-leaf stage for 2 weeks at 6°C with a photoperiod of 8 h. They were then grown in the glasshouse under the same conditions as those described above for healthy seedlings. A set of five isolates, one per identified pathotype (see Table 3), was used. For each line and each pathotype, the flag leaves of two plants with three tillers per plant (i.e. six flag leaves) were inoculated by applying a 1:160 mixture of leaf rust urediospores and Lycopodium spores on the leaf surface with a soft brush. Immediately after inoculation, plants were placed in a dew chamber (15°C) for 24 h and then returned to the greenhouse until the end of the experiment, with an average day/night temperature set at 22/15°C. Infection types were recorded 2 weeks after inoculation using the same scoring system as for seedlings.

In the field, three adjacent rows, 1·5 m long for each entry, were sown in two randomized blocks in Lectoure (France) in October 2009. Once all flag leaves had emerged, the nursery was inoculated by spraying a suspension of spores in mineral oil Soltrol 170® on 9 April 2010. A single pathotype was used, and pathotype identification of eight isolates collected by the end of the epidemic confirmed that no other pathotype had naturally infected the nursery. Disease was scored twice, on 25 May and 8 June using the modified Cobb scale (Peterson et al., 1948).

Postulation of resistance genes

The profiles of infection types displayed by the commercial cultivars when inoculated with the five pathotypes were compared to those displayed by lines of the differential set carrying known resistance genes. Available information on the lines included in the differential set enabled the postulation of genes Lr14a, Lr23 and ‘Alt’ (unnamed gene of the cultivar Altar C84, Singh et al., 2004).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Establishing a differential set

With the exception of Altar C84 and Gaza, all the other international cultivars tested displayed a 3+ (susceptible) infection type to almost all the isolates used, or, in a few cases, an X++3 infection type (Table 2). Lines with uniform infection type were not selected for a differential set. Among the Thatcher isolines, lines with genes Lr1, Lr2a, Lr2b, Lr3, Lr3bg, Lr3ka, Lr9, Lr11, Lr13, Lr15, Lr16, Lr17, Lr19, Lr24, Lr25, Lr26, Lr29, Lr30, Lr32 and Lr38 were resistant to all the isolates, whereas lines with genes Lr2c, Lr10, Lr14b, Lr20, Lr21, Lr33 and Lr44 displayed a 3 to 3+ infection type to most of the isolates, with a few exceptions having a high mesothetic infection type, in particular Lr2c (Table 2). All these Thatcher lines, either resistant or susceptible to all the isolates, were not considered as informative differential hosts, and thus were dropped. Lines with genes Lr14a and Lr23 were resistant, intermediate or susceptible, and thus retained as differentials (Table 2). The line with Lr20, although being susceptible to all isolates (except one with an X++3 IT), was retained in the differential set because it provided a means to clearly detect isolates avirulent on Lr20 (all isolates displaying a characteristic ;N IT on Lr20 turned out to be pathotypes frequently found on bread wheat). Among the cultivars registered in France, Karur, Neodur, Amarillo, Claudio, Combo and Orjaune were dropped because of their high infection type to all isolates. Cultivar Lloyd was kept as a susceptible control. Another susceptible control was added, Local Red or Lebanon, depending on the seed availability. All the other French cultivars were kept because they were either differential, or resistant, thus carrying unknown but effective resistance genes. Acalou was resistant to all the isolates; however, its poor vegetation rendered a reliable assessment of infection types difficult. It too was dropped from the set.

Table 2.   Distribution of infection types (IT) among three sets of lines or cultivars (cvs), inoculated with 96 leaf rust isolates collected from durum wheat. Interactions were categorized as resistant, mesothetic or susceptible depending on infection type
SetCultivaraInfection typebTotalc
  1. aCultivars or lines tested.

  2. bFor each cultivar, lowest to highest infection type displayed (range), and percentage of interactions within that range, over the 96 isolates tested. 0 = no visible symptoms, ; = hypersensitive flecks, 1 = small uredinia with necrosis, 2 = small to medium sized uredinia with green islands and surrounded by necrosis or chlorosis, 3 = medium to large sized uredinia with chlorosis, 4 = large uredinia without chlorosis, X = heterogeneous ITs, evenly distributed over a given leaf, Y = heterogeneous ITs with larger uredia toward the tip. Plus and minus signs indicate variation in ITs. C indicates more than normal chlorosis.

  3. cTotal: Number of interactions tested. Ninety-six isolates were used, but the results of some interactions are missing because the corresponding plants were missing or in too bad a condition to ensure the reliability of the assessment of infection type.

International cvsAcme 0X++313+9996
Altar;− to ;1277 0X++3+ to 3+2396
Arnautka 0 03+10095
Bordos 61 0 03+10095
Cd 40473 0X++313+9995
Entrelargo 0X++313+9995
Gaza0; to ;12100 0 096
Golden Ball 0 03+10096
Karkov 0 03+10096
Kubanka 0 03 to 3+10096
Langdon 0 03+10096
Lebanon 0 03+10043
Local Red 0 03+10053
Medea Ap 9d 0 03+10096
Medora 0 033+ to 3+10096
Peliss 0X++323+9896
Yavaros 0 03+10096
Thatcher isolines with Lr genesLr10; to ;−100 0 095
Lr2a0; to ;12−100 0 095
Lr2b;1+ to ;12+33X−− to X++367 096
Lr2c 0X++ to X++312X++3+ to 3+8894
Lr3; to ;12100 0 096
Lr3bg; to ;12100 0 096
Lr3ka; to ;1++100 0 096
Lr90; to ;−−100 0 096
Lr10 0X++323 to 3+9896
Lr110; to ;12+99Y−1 095
Lr13;1C to 2+C100 0 096
Lr14a 0X+ to X++34533+ to 3+5596
Lr14b 0 03 to 3+10077
Lr150; to ;1−C100 0 095
Lr16;1 to 12C100 0 096
Lr17;1 to ;12+74Y−− to Y++26 090
Lr190; to ;100 0 096
Lr20 0X++3133+ to 3+9995
Lr21 0 033+ to 3+10096
Lr23;1 to 2++40X−− to Y++33633+ to 3+2492
Lr24;−− to ;1C100 0 096
Lr250; to ;12−100 0 096
Lr260; to ;1C100 0 096
Lr290; to ;12C100 0 078
Lr30;1 to ;3−−88Y− to Y++313 096
Lr32; to ;3−100 0 080
Lr33 0 03+10096
Lr380; to ;12−100 0 074
Lr44 0 033+ to 3+10095
French cvsNeodur 0 03+10096
Lloyd 0 03+10096
Karur 0 03 to 3+10095
Combo 0X++313 to 3+9996
Orjaune 0X++ to X++333 to 3+9796
Claudio 0X++ to Y++353 to 3+9596
Amarillo 0X++ to X++323 to 3+9896
Orobel;−− to ;3−8X−− to Y++3383+5496
Artimon;121X−− to Y++3453+5496
Nefer;−− to ;122X−− to X++443+5496
Amosis;−−1X−− to Y++453+5496
Reaumur0; to ;12+11X−− to X++3433+ to 3+5496
Arcalis;−− to ;12+20X−− to X++2633+ to 3+5496
Joyau;−− to ;12+4X−− to X++34233+ to 3+5496
Akenaton;12 to ;12+8X−− to X++34433+ to 3+4896
Biensur0; to ;1+75X−− to X−23+2395
Pescadou; to ;1235X−− to X++343X++3+ to 3+2296
Orlu0; to ;175X−− to Y++53+2095
Dakter0; to ;1+74X−− to X++3633+ to 3+2095
Sachem0; to ;12+51X−− to X++345X++3+ to 3+495
Nautilur0; to ;12+78X−− to X++3193+293
Latinur0; to ;12+82X−− to X++143+492
Silur0; to ;1279X− to X++3163 to 3+492
Brennur0; to ;12+82X− to X++3173+192
Acalou;3−− to ;3−4X−− to X++396 096
Byblos0; to ;3−−73X−− to X++327 095

CS2A/2M (Lr28) and cv. Gatcher (Lr27+31), although not included in the set of 73 potentially differential host genotypes tested, were added to the differential set in 2008 and 2009 respectively, because taking into account virulence to these genes opens the door to comparisons with populations from other countries (Huerta-Espino et al., 2009). The final differential set comprised 28 lines and cultivars (Table 3).

Table 3.   Differential set for virulence phenotyping of Puccinia triticina from durum wheat, and infection types (IT)a for the five pathotypes identified in France, at the adult and seedling stages in the glasshouse, and in the field for isolate BD07 012
Differential setPathotype (isolate)
[0] (BD01 219)[14a] (BD07 084)[Altar, 23] (BD01 117)[Altar, 23, (Gaza)] (B00 020-B)[(Altar), 14a, 23] (BD07 012) 
  1. Lines/cultivars retained in a simplified differential set are in bold.

  2. a0 = no visible symptoms, ; = hypersensitive flecks, 1 = small uredinia with necrosis, 2 = small to medium sized uredinia with green islands and surrounded by necrosis or chlorosis, 3 = medium to large sized uredinia with chlorosis, 4 = large uredinia without chlorosis, X = heterogeneous ITs, evenly distributed over a given leaf, Y = heterogeneous ITs with larger uredia toward the tip. Plus and minus signs were used to indicate variation in ITs, and the letters C and N were used to indicate more than normal chlorosis or necrosis, respectively.

  3. bCultivars within a group were postulated to have the same gene combination, indicated between brackets. ‘Alt’ is used to designate the gene combination of cultivar Altar C84.

  4. cField assessment according to the modified Cobb Scale (Peterson et al., 1948).

  5. dInteraction not tested.

  6. eThe four lines/cultivars marked with an asterisk are those sufficient to distinguish the five pathotypes.

Group Ib (Lr14a)
 ArtimonX++03+3+X++0X+03+3+100 Sc
 AmosisX+03+3+X0X03+3+80 S
 JoyauX03+3+X0X03+3+70 S
 OrobelX++03+3+X0X−03+3+90 S
 NeferX03+3+X0X−−03+3+70 S
 ArcalisX03+3+X−0;1203+3+80 S
Group II (Alt)
 Biensur;−0;03+3+3+3+3+X60 MS
 Dakter;−−0;−03+X−−3+X−−XX−−40 R-MR
 Orlu;−−0;−03+X++3+X++X++X−−40 R-MR
 Altar C84*;X−−;X−−3+3+3+3+X++3X++50 R-MR
Group III (Alt, Lr14a)
 Silur;−−0;;;X0X;−X++33+70 S
 Latinur;−−0;;−X0X−03+X70 S-MS
 Sachem;−0;12;−−X0X03+X80 S
 Brennur0;0;−0;10X−−0X+X50 MR
 Nautilur0;0;−−0;120XX++X++3X50 MR
 Pescadou;1X+YX++3+3+3+3+3+3+70 S
 Tc Lr23*;12X;12X3+3+3+3+3+3+
 ByblosXX−−X++X−−X+X−−0;X−−;X−−5 R
 Tc Lr2033+3+33+3+3+3+3+3+3+3+
 Lloyd3+3+3+3+3+3+3+3+3+3+100 S
 Local Red3+3+3+3+3+3+3+3+3+3+

Identification of pathotypes

Among the 310 isolates from different years, areas and cultivars of origin, only five different pathotypes were identified (Table 3). Four lines, TcLr14a, TcLr23, Altar C84 and Gaza, were enough to distinguish these five pathotypes. For the purposes of this study, a simple nomenclature system was therefore used, based only on these four lines/cultivars. Pathotype [0] (47 isolates) was avirulent on the four lines. Pathotype [14a] (123 isolates) was virulent only on Lr14a. Pathotype [(Altar), 14a, 23] (36 isolates) was virulent on both Lr14a and Lr23. It was avirulent on Gaza and Altar C84, but displayed a high IT X++3 on Altar C84 at the seedling stage, while pathotypes [0] and [14a] displayed an IT ; to ;1 on Altar C84. Pathotype [Altar, 23] (74 isolates) was virulent on both Altar C84 and Lr23. In addition, pathotype [Altar, 23, (Gaza)] was distinguished, virulent on both Altar C84 and Lr23, and displaying an intermediate or high IT on Gaza, whereas the other four pathotypes displayed a very low IT (;− to ;) on Gaza. Pathotype [Altar, 23, (Gaza)] was found in the case of only three isolates: at the seedling stage, two of them displayed a mesothetic infection type on Gaza, while the remaining one displayed a susceptible infection type.

Postulation of resistance genes in commercial cultivars of the differential set

Analysis of the infection types allowed the cultivars to be put into groups displaying similar profiles at the seedling and adult stages (Table 3). The first group comprises cultivars Artimon, Akenaton, Amosis, Reaumur, Joyau, Orobel, Nefer and Arcalis. Given their infection type profile similar to those of TcLr14a (Table 3), it can be postulated that all the cultivars of this group have Lr14a. A second group of cultivars, Biensur, Dakter, Orlu and Altar C84, displayed an identical profile at the seedling stage for four of the five pathotypes (Table 3). The ITs of Altar C84 and Biensur appeared higher than those of Dakter and Orlu for the fifth pathotype [(Altar), 14a, 23] at the seedling stage, but this difference was not confirmed at the adult stage in the glasshouse where the four cultivars displayed intermediate ITs (X = to X++) with this pathotype (Table 3). It was postulated that Biensur, Dakter and Orlu have the same wheat leaf rust resistance gene(s) as Altar C84. A third group of cultivars displaying similar profiles included Silur, Latinur, Sachem, Brennur and Nautilur. In the glasshouse, their IT profile with isolates avirulent on Altar C84 suggests that these cultivars could possibly have the same gene(s) as Altar C84 for resistance to wheat leaf rust. With the two pathotypes virulent on Altar C84, the resistant ITs displayed by these cultivars was compatible with the expression of Lr14a. The combination (Alt, Lr14a) was then suggested for the cultivars of this third group, where Alt refers to the gene combination of cultivar Altar C84. Two of the French cultivars retained in the differential set did not fall into any one of the above defined groups. Pescadou displayed a profile very similar to TcLr23 at the seedling and adult stages in the glasshouse. It is postulated that Pescadou is carrying gene Lr23. Byblos was the only cultivar displaying low infection types to the five pathotypes at the seedling and adult stages. It was very resistant in the field trial (5R), as is the case in nurseries and commercial fields in France. At this stage, given the available information, it is not possible to postulate the gene(s) providing this efficient resistance to leaf rust.

Reduction of the differential set

After the postulation of resistance genes in the commercial cultivars of the differential set, it appeared that some of them were likely to be redundant. Thus the differential set was simplified, keeping only cultivars that are representative of each group of postulated gene combinations. However, this selection was based on in-depth analysis of the profile of infection types, taking into account slight differences within the groups defined above. Within group I composed of eight cultivars (Table 3), three subgroups were established mainly based on minor differences in infection types at the seedling stage with pathotypes [Altar, 23] and [Altar, 23, (Gaza)]. Cultivars Artimon, Joyau and Nefer were taken as representative for each of the three subgroups. Artimon is also useful for the detection of bread wheat pathotypes, which were avirulent to all the differential commercial cultivars except Artimon (results not shown).

All the cultivars were kept in group II, because of small differences observed within this group with pathotype [Altar, 23] at the adult stage, and with pathotype [(Altar), 14a, 23] in the field (Table 3).

In group III, for which the combination (Alt, Lr14a) was postulated, the low ITs displayed by Brennur and Nautilur to pathotype [Altar, 23, (Gaza)] in the glasshouse is compatible with the presence of Lr14a, but suggests a gene in addition to Lr14a. Moreover, field findings did not fully confirm that the five cultivars in group III had the same resistance genes. Brennur and Nautilur were scored 50MR, whereas Latinur, Sachem and Silur were scored 70 S-MS to 80S. Cultivars Brennur and Nautilur were therefore kept, in addition to cv. Silur, as representative of this third group.

Structure and evolution of the French durum P. triticina population

As mentioned above, the 310 isolates chosen to represent the French durum wheat leaf rust population between 1999 and 2009 displayed only five different pathotypes. Genotyping of 76 isolates, chosen so as to represent these five pathotypes (Table 4), showed that all 76 isolates had the same genotype, except for three variants, differing by only one allele for two of them, and three alleles at two loci for the third one (Table 5).

Table 4.   Number of isolates selected for genotyping for each of the five pathotypes characterized (see Table 3 for the description of the pathotypes), and distribution across the four genotypes identified (see Table 5 for the description of the genotypes)
[0]9   9
[14a]16   16
[Altar, 23]1311 15
[Altar, 23, (Gaza)]2   2
[(Altar), 14a, 23]33  134
Table 5.   Dominant genotype (73 out of 76 isolates) of French durum leaf rust isolates for 20 SSR loci. Three variants of genotype 1 were identified: genotype 2, displaying 282/290 at locus RB12, genotype 3 displaying 175/200 at locus RB11, and genotype 4, displaying 175/200 at locus RB11, and 230/230 at locus RB25
LocusGenotype 1
Allele 1Allele 2

Pathotype [0] was dominant from 1999 to 2001, and then decreased and was no longer found in 2008 and 2009 (Fig. 2a). It was replaced from 2001–2002 by pathotype [14a], which was dominant up to 2007 (Fig. 2a). Pathotype [Altar, 23], present at a low to medium frequency from 1999 to 2007, strongly increased in 2008 and 2009 (Fig. 2a). Pathotype [(Altar), 14a, 23] emerged in 2007 and represented approximately 50% of the population in 2008 and 2009 (Fig. 2a). The pathogen population appeared strongly structured by the cultivars. Cultivar Neodur, susceptible to all the isolates (Table 2) and thus probably carrying no resistance gene, hosted a single pathotype, [0], avirulent on all the differentials up to the year 2000 (Fig. 2b). After that point in time, it hosted all five pathotypes. Cultivar Nefer, postulated to carry Lr14a (Table 3) hosted almost exclusively a single pathotype, [14a], first detected in 2001 (Fig. 2c). Cultivar Biensur, postulated to carry Alt (Table 3), hosted almost exclusively up until 2006 pathotype [Altar, 23], avirulent on Lr14a and virulent on Altar C84 (Fig. 2d). As of 2007, all the cultivars started to host the new pathotype [(Altar), 14a, 23] (Fig. 2a). This pathotype was virulent on most of the French cultivars at the seedling stage.


Figure 2.  Number of isolates of the different pathotypes in samples collected from durum wheat from 1999 to 2009. (a) Overall sample from all durum wheat cultivars, (b) Sample from cv. Neodur only, (c) Sample from cv. Nefer only, and (d) Sample from cv. Biensur only. Five pathotypes were distinguished, according to their virulences on the following four lines/cultivars: Altar C84, TcLr14a, TcLr23 and Gaza. Pathotype [0] was avirulent on these four lines. Pathotype [14a] was virulent only on Lr14a. Pathotype [(Altar), 14a, 23] was virulent on Lr14a and Lr23, and avirulent on Gaza and Altar C84, but with a high IT (X++3) on Altar C84 at the seedling stage. Pathotype [Altar, 23] was virulent on both Altar C84 and Lr23. Pathotype [Altar, 23, (Gaza)] was virulent on both Altar C84 and Lr23, and displayed an intermediate or high IT on Gaza.

Download figure to PowerPoint

Twenty-seven of the 310 isolates displayed a pathotype different from the five above-mentioned pathotypes: they were avirulent on all the durum commercial cultivars except Artimon, and virulent on Lr14a. Fifteen of them were checked on the Thatcher lines differential set used for bread wheat (Goyeau et al., 2006), and they all happened to be common bread wheat pathotypes. All isolates avirulent to Lr20 were common bread wheat pathotypes.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

A first investigation of the diversity and evolution in the population of P. triticina growing on durum wheat in France was achieved thanks to the establishment of an adapted differential set. In addition, this study provides information about leaf rust resistance genes in the French durum wheat germplasm, and access to pathotypes that are relevant for testing new breeding lines.

From 1999 to 2009, a small percentage of the isolates collected on durum wheat (27 out of 310) were bread wheat pathotypes, indicating that some bread leaf rust pathotypes can grow on durum wheat. However, overall, they remained at very low frequencies and cannot be considered as responsible for epidemics. This study confirms that the population of leaf rust sampled from durum wheat cultivars is clearly different, given its phenotypes and genotypes, from the population growing on bread wheat, as pointed out by previous studies (Singh, 1991; Huerta-Espino & Roelfs, 1992; Singh et al., 2004; Ordonez & Kolmer, 2007b). In the present study, the year 1999 appeared as an exception: a significant proportion (42%) of the isolates collected from durum wheat was in fact bread wheat pathotypes (Fig. 2a). However, it was after 1999 that the protocols were greatly improved to avoid contamination both in the lab and during collection in the field, which might explain why 1999 appears as an exception.

Four host lines, TcLr14a, TcLr23, Altar C84 and Gaza, proved sufficient to differentiate the five pathotypes. The differential set used for wheat leaf rust in North America (Kolmer et al., 2008) does not include all of these four lines. This holds true for the set used by CIMMYT in Mexico (Singh, 1991). It explains why the nomenclature based on those sets could not be used, and thus a simple but informative nomenclature is given here, based on the four given lines, knowing that the emergence of new pathotypes will require the system to be adapted. Gaza and Lr23 are not independent as differentials, as Gaza has Lr23 (McIntosh et al., 1995). However the pathotypes identified in this study confirm that Gaza has additional resistance other than Lr23 (Singh et al., 2004), as it is resistant to isolates virulent on Lr23.

To ensure that the differential set proposed here, comprising 20 lines and cultivars, is all encompassing, cv. Acalou, dropped after the first stage of the study because of its poor vegetation but resistant to all isolates, has to be included as an unknown but efficient source of resistance. The variation of infection types observed on TcLr2c suggests that this line might also be useful as a differential.

Diversity in the durum leaf rust population in France between 1999 and 2009 was very low, as revealed by both genotyping and phenotyping. The population was almost monomorphic, according to the set of 20 SSR markers used, with a single genotype representing 96% of the isolates analysed, two other genotypes differing from the dominant genotype at a single locus, and a fourth genotype differing at only two loci. The SSR employed to carry out the genotyping were identified using bread wheat leaf rust isolates (Enjalbert et al., 2002; Szabo & Kolmer, 2007). It cannot be excluded that this SSR set has a low discrimination power for durum wheat leaf rust isolates, but on most loci, the four genotypes differed from the genotypes identified in the French bread wheat leaf rust population (results not shown). This confirms the divergence between leaf rust populations on bread and durum wheat on a country scale, as already shown at the intercontinental scale by Ordonez & Kolmer (2007a). These authors also noted the high genetic similarity between the durum leaf rust isolates from South America, North America and Europe. Within-country genetic diversity was already found to be low, yet higher than in the present study; Ordonez & Kolmer (2007a) distinguished five genotypes in a set of 14 durum French isolates using 11 SSR markers, and Mantovani et al. (2010) characterized seven genotypes using 15 SSR markers (of which 12 were in common with those used in the present study) in a set of 15 durum Italian isolates.

Phenotyping also revealed a low level of diversity: the five identified pathotypes were very close, as they differ from each other by at the most three virulences. Moreover, the study of their emergence over time, as compared to pathotypes identified elsewhere in the world, raises issues of relatedness and intercontinental migration. At the beginning of the period under study, in 1999–2000, the French population was mainly composed of pathotype [0], avirulent on all the differential lines, and almost all the tested Lr genes, except Lr2c, Lr10, Lr14b, Lr20, Lr21, Lr33 and Lr44. Race BBB/BN, identified as the most common race on durum wheat in Mexico until 2001 (Singh et al., 2004) is closely related to pathotype [0], but differs in that BBB/BN is virulent on Lr23 whereas pathotype [0] had an IT of ;12 on Lr23. These two pathotypes also displayed different ITs on Lr21, Lr2c and Gaza. Although all the isolates tested in this study displayed a high (33+ to 3+) infection type to line TcLr21, it is not considered that their virulence on Lr21 is proven; testing at the adult stage would be necessary to establish virulence, as avirulent responses for Lr21 can be mistakenly interpreted as virulent based on the seedling reactions (McIntosh et al., 1995).

In 1999 and 2000, one and three isolates of pathotype [Altar, 23] were collected, respectively. In Mexico, virulence for Altar C84 appeared in 2001 when the new race BBG/BN emerged (Singh et al., 2004). However, there is an isolate of this pathotype collected in 1997 (not included in the present study). Thus virulence for Altar C84 was present in France before its emergence in Mexico in 2001. Pathotype [Altar, 23] and race BBG/BN differ by their ITs on Lr11, Lr21 and Gaza. In France, the frequency of pathotype [Altar, 23] was very high on cultivar Biensur. However, this pathotype probably first emerged on another cultivar, as cultivar Biensur was only registered in 2001.

Pathotype [Altar, 23, (Gaza)] was first collected in 2000. It never spread, but remained in the population, as the only other two isolates of this pathotype were collected in 2004 and 2007. It is very close to the Mexican BBG/BN, as it differs only by its ITs on Lr11 and Lr21. As already discussed by Ordonez & Kolmer (2007b) for Lr11, these few differences might well be due to the use of different seed stocks. Thus it cannot be excluded that the French pathotype [Altar, 23, (Gaza)] is the same as BBG/BN. Although Ordonez & Kolmer (2007b) have not used Altar C84 and Gaza as differential lines, comparisons can be made from the virulence phenotype with Lr genes, given the low diversity level in the French population. The French pathotype [Altar, 23] or [Altar, 23, (Gaza)] most likely corresponds to the virulence phenotype labelled 10, 14b, 20, 22a, 23, 33, 34, 41, 44 in Table 1 in Ordonez & Kolmer (2007b). These authors found that this virulence phenotype was the most frequent one in Mexico, Argentina and the United States. Thus it is likely that the Americas and France share this common virulence phenotype. However, providing definite proof of a single genotype, as well as its migration routes, would require molecular markers that provide better discrimination than the available set of SSRs.

The findings here related to pathotypes [Altar, 23] and [Altar, 23, (Gaza)] taken alone would tend to confirm that the migration of P. triticina on durum wheat (Ordonez & Kolmer, 2007b) is most likely quick and intercontinental. However, further findings from the detailed study at the scale of a country, France, reveal that there is an exception: pathotype [14a]. It was first isolated in 2001, and represented 90% of the total number of isolates collected from cultivar Nefer from 2001 to 2009. This cultivar was registered in 1997 and was completely resistant to wheat leaf rust until 2000. In 2001, French wheat breeders and Arvalis (French Technical Institute for Crops) observed strong epidemics on Nefer. Thus virulence for Lr14a in durum leaf rust pathotypes probably emerged only around 2000. The cultivars in which the gene Lr14a was postulated, including Nefer, were registered between 1995 (Arcalis) and 2005 (Akenaton). It is thus likely that isolates virulent for this gene emerged locally due to selection by the host. Virulence for Lr14a is not widespread in durum wheat leaf rust populations; it was present in isolates collected from Argentina, Chile and Ethiopia (Ordonez & Kolmer, 2007b), but it was not found in Mexico, the United States (Ordonez & Kolmer, 2007b), or Italy (Mantovani et al., 2010). Status of this virulence in Spain is unclear; all eight Spanish durum wheat leaf rust isolates collected in 2000–2003 were found avirulent by Ordonez & Kolmer (2007b), whereas all 14 isolates collected on durum wheat in Spain in 1999 and 2000 were found virulent by Martinez et al. (2005). The discrepancy between these two studies could be due either to the very low sample size, or to a different interpretation of infection types, as already suggested by Ordonez & Kolmer (2007b), especially as the avirulent infection type for Lr14a at the seedling stage is known to be mesothetic with a high environmental variability (McIntosh et al., 1995). Although pathotype [14a] has been in France for 10 years, where cv. Creso and other cultivars carrying Lr14a proved to be highly susceptible in the field to this pathotype (Goyeau et al., 2010), it has never spread to neighbouring countries (Spain, Italy), as suggested by the high level of resistance of cv. Creso in these countries. Still today, cv. Creso is considered by durum breeders as the most important source of durable resistance to leaf rust in the Mediterranean areas (Maccaferri et al., 2008, 2010). These findings illustrate that quick long-distance migration of this fungus should not be taken as a general rule.

In 2007, a new pathotype [(Altar), 14a, 23] emerged, combining virulence for Lr14a and Lr23, and displaying a very high IT on seedlings of Altar C84, but avirulent at the adult stage. Its frequency was low, 22, 8 and 11% respectively on Nefer, Biensur and Neodur on which the sampling was focused, and 23% given the overall sample. Thus no hypothesis can be formulated about a particular cultivar favouring the emergence of this pathotype. Pathotypes combining virulence for Lr14a and Lr23 were also found in Chile and Argentina (Ordonez & Kolmer, 2007b), but they differed by a few virulence genes from pathotype [(Altar), 14a, 23]. Emergence of pathotype [(Altar), 14a, 23] illustrates the continuous evolution of the population over time; however, the present knowledge does not allow one to conclude whether this emergence was driven by local selection only, or by long-distance migration followed by selection.

The pace of emergence of new durum leaf rust pathotypes is similar in France and Mexico; their number went up from one to five and six, respectively (Huerta-Espino et al., 2009) between 2000 and 2009, with new pathotypes gaining one additional virulence at each step. This illustrates a rapid evolution of the pathogen population in response to changes in the host population, despite its clonal reproduction in France, as suggested by a very low level of genetic diversity. This rather quick evolution emphasizes once again the need to survey these populations annually to provide breeders with up-to-date knowledge of the virulences in the pathogen populations. From the point of view of breeding, this study shows that there were few efficient sources of resistance to leaf rust in the cultivars tested, as only cultivars Acalou, Brennur, Byblos, Dakter, Nautilur and Orlu were resistant at the adult stage to the five pathotypes present in the French durum wheat leaf rust population. The sources of resistance present in Acalou and Byblos is unknown. The four other above-mentioned cultivars might be related as they were all postulated to carry resistance gene(s) of Altar C84.

If durable resistance is to be brought about, there is a pressing need to diversify sources of resistance to wheat leaf rust in the durum wheat germplasm. Moreover, the strategy of using single Lr genes should be avoided. No matter how fully efficient they are against durum wheat leaf rust isolates, the breakdown of Lr genes is likely to occur within a few years as exemplified by the rapid changes observed in France in this study, as well as in Mexico (Huerta-Espino et al., 2009).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work could not have been achieved without the support and efficient collaboration provided by French durum wheat breeders, namely Carine André, Pierre Blanc, Laurent Gervais, Jean-Pierre Hardouin, Franck Lacoudre, Thierry Lefèvre, Jean-Paul Le Goff and Philippe Lonnet. Collaboration with Karim Ammar (CIMMYT, Mexico) greatly helped us in providing information on the genetic basis of resistance to durum wheat leaf rust in durum germplasm. The financial support of Contrat de Branche Ministère de l’Agriculture 2006–2008 and 2009–2011 is gratefully acknowledged. We are indebted to all the collaborators who have provided us with durum wheat leaf rust isolates, namely breeders and Arvalis Institut du Végétal. Anne-Catherine Zippert provided excellent technical assistance in the management of the collection of isolates. Technical help by Nicolas Lecutier is also acknowledged. We are especially grateful to Dr S. Tanis-Plant for a first review of this paper in English.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Anikster Y, Bushnell WR, Eilam T, Manisterski J, Roelfs AP, 1997. Puccinia recondita causing leaf rust on cultivated wheats, wild wheats, and rye. Canadian Journal of Botany 75, 208296.
  • Anonymous, 2011. The Durum Wheat Market – Worldwide, European Union, France. Montreuil, France: FranceAgriMer. [].
  • Casulli F, Pasquini M, Pancaldi D, Grifoni F, 1996. Epidemiological aspects and virulence of Puccinia recondita f. sp. tritici in Italy. In: Kema GHJ, Niks RE, Daamen RA, eds. Proceedings of the 9th European and Mediterranean Cereal Rusts and Powdery Mildews Conference. Lunteren, The Netherlands: European and Mediterranean Cereal Rusts Foundation, 137.
  • D’Oliveira B, Samborski DJ, 1964. Aecial stage of Puccinia recondita on Ranunculaceae and Boraginaceae in Portugal. In: Macer CF, Wolfe MS, eds. Proceedings of the First European Brown Rust Conference. Cambridge, UK: Plant Breeding Institute, 13349.
  • Enjalbert J, Duan X, Giraud T, Vautrin D, de Vallavieille-Pope C, Solignac M, 2002. Isolation of twelve microsatellite loci, using an enrichment protocol, in the phytopathogenic fungus Puccinia striiformis f. sp. tritici. Molecular Ecology Notes 2, 5635.
  • Ezzahiri B, Diouri S, 1992. Anchusa italica as an alternate host for wheat leaf rust in Morocco. Plant Disease 76, 1185.
  • Goyeau H, Park RF, Schaeffer B, Lannou C, 2006. Distribution of pathotypes with regard to host cultivars in French wheat leaf rust populations. Phytopathology 96, 26473.
  • Goyeau H, Halkett F, Zapater MF, Carlier J, Lannou C, 2007. Clonality and host selection in the wheat pathogenic fungus Puccinia triticina. Fungal Genetics and Biology 44, 47483.
  • Goyeau H, Ammar K, Berder J, 2010. Virulence in Puccinia triticina for durum wheat cultivar Creso and other durum wheat cultivars carrying resistance gene Lr14a in France. Plant Disease 94, 1068.
  • Herrera-Foessel SA, Singh RP, Huerta-Espino J, Yuen J, Djurle A, 2005. New genes for leaf rust resistance in CIMMYT durum wheats. Plant Disease 89, 80914.
  • Huerta-Espino J, Roelfs AP, 1989. Physiological specialization of leaf rust on durum wheat. Phytopathology 79, 1218 (Abstract).
  • Huerta-Espino J, Roelfs AP, 1992. Leaf rust on durum wheats. Vorträge für Pflanzenzüchtung 24, 1002.
  • Huerta-Espino J, Singh RP, Herrera-Foessel SA, Perez-Lopez JB, 2009. Phenotypic variation among leaf rust isolates from durum wheat in Northwestern Mexico. In: Akkaya MS, ed. Proceedings of the 12th International Cereal Rusts and Powdery Mildews Conference. Antalya, Turkey: European and Mediterranean Cereal Rusts Foundation, 29.
  • Knott DR, 1989. The Wheat Rusts – Breeding for Resistance. Berlin, Germany: Springer-Verlag.
  • Kolmer JA, 1996. Genetics of resistance to wheat leaf rust. Annual Review of Phytopathology 34, 43555.
  • Kolmer JA, Long DL, Hughes ME, 2008. Physiologic specialization of Puccinia triticina on wheat in the United States in 2006. Plant Disease 92, 12416.
  • Maccaferri M, Mantovani P, Tuberosa R, De Ambrogio E, Giuliani S, Sanguineti MC, 2008. A major QTL for durable leaf rust resistance widely exploited in durum wheat breeding programs maps on the distal region of chromosome arm 7BL. Theoretical and Applied Genetics 117, 122540.
  • Maccaferri M, Sanguineti MC, Mantovani P et al. , 2010. Association mapping of leaf rust response in durum wheat. Molecular Breeding 26, 189228.
  • Mantovani P, Maccaferri M, Tuberosa R, Kolmer JA, 2010. Virulence phenotypes and molecular genotypes in collections of Puccinia triticina from Italy. Plant Disease 94, 4204.
  • Martinez F, Sillero JC, Rubiales D, 2005. Pathogenic specialization of Puccinia triticina in Andalusia from 1998 to 2000. Journal of Phytopathology 153, 3449.
  • McIntosh RA, Wellings CR, Park RF, 1995. Wheat Rusts. An Atlas of Resistance Genes. Dordrecht, The Netherlands: Kluwer Academic Publishers.
  • Ordonez ME, Kolmer JA, 2007a. Simple sequence repeat diversity of a worldwide collection of Puccinia triticina from durum wheat. Phytopathology 97, 57483.
  • Ordonez ME, Kolmer JA, 2007b. Virulence phenotypes of a worldwide collection of Puccinia triticina from durum wheat. Phytopathology 97, 34451.
  • Peterson RF, Campbell AB, Hannah AE, 1948. A diagrammatic scale for estimating rust intensity of leaves and stems of cereals. Canadian Journal of Research Section C 26, 496500.
  • Samborski DJ, 1985. Wheat leaf rust. In: Roelfs AP, Bushnell WR, eds. The Cereal Rusts II. London, UK: Academic Press, 3959.
  • Singh RP, 1991. Pathogenicity variations of Puccinia recondita f. sp. tritici and P. graminis f. sp. tritici in wheat-growing areas of Mexico during 1988 and 1989. Plant Disease 75, 7904.
  • Singh RP, Huerta-Espino J, Pfeiffer W, Figueroa-Lopez P, 2004. Occurence and impact of a new leaf rust race on durum wheat in Northwestern Mexico from 2001 to 2003. Plant Disease 88, 7038.
  • Stakman EC, Stewart DM, Loegering WQ, 1962. Identification of physiologic races of Puccinia graminis var. tritici. USDA-ARS Bulletin 617, 153.
  • Szabo LJ, Kolmer JA, 2007. Development of simple sequence repeat markers for the plant pathogenic rust fungus Puccinia triticina. Molecular Ecology Notes 7, 70810.