Introduction

Rust resistance gene Lr34 is one of the most important disease resistance genes in wheat. It was initially described by Dyck et al. (1966) as an adult plant leaf rust (caused by Puccinia triticina eriks.) resistance gene. This locus also confers resistance to stripe rust (Yr18, Singh 1992b), stem rust (Dyck et al. 1985; Hiebert et al. 2010), powdery mildew (Pm38, Spielmeyer et al. 2005), and barley yellow dwarf virus (Bdv1, Singh 1993). Lr34 conditions adult plant, slow rusting resistance and is completely linked with leaf tip necrosis (Singh 1992a). This resistance has not been overcome by P. triticina in a race-specific manner, over many years of widespread use throughout the world. This locus also enhances resistance conditioned by other leaf rust (German and Kolmer 1992) and stem rust (Hiebert et al. 2010) resistance genes. The Lr34/Yr18 locus was cloned (Krattinger et al. 2009), and molecular markers were developed that are either closely linked to Lr34, or within the coding region of the gene (Dakouri et al. 2010; Lagudah et al. 2009).

The Western Canadian provinces of Alberta, Saskatchewan, and Manitoba produce approximately 96% of the wheat in Canada (Statistics Canada 2011). Canada Western Red Spring (CWRS) cultivars are grown on an average of 70% of the wheat area in Western Canada. Thatcher, the most popular CWRS cultivar from the 1940s to 1970s, is susceptible to leaf rust (McCallum and DePauw 2008). Breeding and research efforts resulted in the incorporation of resistance into subsequent cultivars, many of which were derived from Thatcher. Commonly used resistance genes in this class include Lr10, Lr13, Lr14a, Lr16, Lr21, Lr22a, and Lr34 (McCallum et al. 2007a). Glenlea, registered in 1972, was the first major cultivar in Western Canada to carry Lr34 (Dyck et al. 1985), although it did not have the end-use functionality to be a CWRS cultivar (McCallum and DePauw 2008). Since the 1970s, Glenlea and subsequent cultivars with Lr34 have maintained a moderate level of resistance. Combinations of resistance genes involving Lr34 have often been very effective, partially due to the ability of Lr34 to enhance the expression of other resistance genes (German and Kolmer 1992). Laura, registered in 1986, was the first CWRS cultivar with Lr34 (McCallum and DePauw 2008) and the gene was subsequently incorporated into more recently released CWRS cultivars. Determining the presence of Lr34 in current cultivars should help to predict the field reaction and durability of cultivars and to aid wheat breeder’s decisions on selecting parents for future crosses. The other main bread wheat classes in Western Canada include the Canada Prairie Spring (CPS) and Canada Western Extra Strong (CWES) which are distinguished from CWRS cultivars principally by their different end-use functionalities. Germplasm from these different classes tends to be relatively distinct (McCallum and DePauw 2008).

It is possible to postulate the presence or absence of Lr34 in cultivars based on adult plant reaction. However, this is complicated by the presence of other resistance genes, the race non-specific nature of Lr34, and effects of environmental variability on expression. Molecular markers, tightly linked to Lr34, can help postulate the presence of this important gene. The csLV34 marker was reported to be tightly linked to Lr34 (Lagudah et al. 2006) and diagnostic for the presence of the gene; two alleles were detected; csLV34a (229 bp) which is usually associated with the susceptibility allele and csLV34b (150 bp), normally associated with the Lr34 resistance allele (Lagudah et al. 2006). This marker was used, in conjunction with phenotypic analysis, to postulate the presence of Lr34 in 84 Australian wheat cultivars (Singh et al. 2007). The authors found good agreement between the presence of csLV34b and adult plant leaf rust resistance, although some susceptible lines also had a csLV34b allele in which recombination between Lr34 and csLV34 could have taken place, as was subsequently reported in other wheat lines (McCallum et al. 2008; Lagudah et al. 2009). The csLV34 marker was also used to assay the presence of Lr34 in a larger, more diverse group of cultivars from many countries (Kolmer et al. 2008). Additional markers linked to Lr34 that have been used for mapping include wms1220 (Lagudah et al. 2006), csLVMS1 (Spielmeyer et al. 2008), and swm10 (Bossolini et al. 2006).

Subsequently the coding region for Lr34/Yr18 was identified, and reported to be an ABC transporter gene (Krattinger et al. 2009). Three mutations were originally described within the ABC transporter sequence, namely single nucleotide polymorphisms (SNPs) in intron 4 and exon 12, and a 3 bp indel in exon 11 (Krattinger et al. 2009). Gene specific markers within the ABC transporter gene were developed based on these sequence differences between the resistance and susceptibility alleles (Lagudah et al. 2009). Fine mapping of this region on chromosome 7D confirmed that the ABC transporter was the most likely valid candidate gene conditioning leaf rust resistance (Dakouri et al. 2010). Dakouri et al. (2010) described 10 new molecular markers spanning the Lr34 locus, four of which were within the ABC transporter gene. To determine haplotypes of wheat lines at the Lr34 locus and vicinity the gene-based markers described by Lagudah et al. (2009), those described by Dakouri et al. (2010), and previously reported linked markers, are useful. The objective of our study was to determine the Lr34 molecular marker profiles of a large number of Canadian wheat cultivars, related cultivars and lines, advanced breeding lines and other potential source materials to determine Lr34 haplotype variation.

Materials and methods

Plant materials

Three hundred and seventy-five wheat lines were assembled for this study. They included nearly all Canadian cultivars from each of the major spring bread wheat (Triticum aestivum L.) classes, including CWRS, CPS, CWES, and, Canada Western Hard White spring (CWHWS) which represents most of the bread wheat production in Canada (McCallum and DePauw 2008). The collection also included breeding lines within each of these end-use functionality classes, parental lines for many of the cultivars, some unrelated cultivars, and lines with a diversity of origins throughout the world.

Genomic DNA extraction and molecular marker analysis

Plants from each line were grown to the 3–4 leaf stage and DNA was extracted from frozen, lyophilized leaf tissue, as described by Dakouri et al. (2010). The molecular markers linked to Lr34 which were used in this study (in map order along chromosome 7D) were: wms1220 (Lagudah et al. 2006), caSNP12, caIND11, caIND10, caSNP4, caISBP1, cam11 (Dakouri et al. 2010), csLVMS1 (Spielmeyer et al. 2008), swm10 (Bossolini et al. 2006), csLV34 (Lagudah et al. 2006). Markers caSNP12, caIND11 and caSNP4 correspond to Krattinger et al. (2009) sequence variation within the ABC transporter gene, namely a C/T SNP in exon 12, a 3 bp indel in exon 11, and an A/T SNP in intron 4. The markers inside the ABC transporter coding sequence are caSNP12, caIND11, caIND10, and caSNP4, where caIN10 characterized a 1 bp indel in exon 10 (Dakouri et al. 2010). All other markers are located outside the coding region of Lr34.

Microsatellite markers wms1220, cam11, csLVMS1, swm10, as well as indel marker caIND11 and caIND10, were resolved on an ABI3100 using a modified amplification procedure with M13 tail primers (Schuelke 2000). Markers csISBP1, csLV34, caSNP12, caSNP4 were resolved on agarose gels as either co-dominant or dominant (SNP) markers. Allele sizes were determined by fragment analysis using a modified version of Genographer (Benham et al. 1999) and include an M13 tail of 19 bp.

Results

Haplotype groups and molecular marker profiles for 375 wheat cultivars and lines

The numbers of alleles differed among the markers for this diverse collection of germplasm (Table 1). There were only two alleles for all markers except wms1220, cam11, csLVMS1, and swm10 which had 9, 6, 4 and 5 alleles each, respectively. The markers within the coding sequence of the gene: caSNP12, caIND11, caIND10, caSNP4, divided the lines into four main groups. The assignment of these groups into those carrying the resistance or a susceptibility allele at the Lr34 locus was based on the alleles of caSNP12 and caIND11 because these were shown to be diagnostic for resistance conditioned by Lr34 in previous studies (Krattinger et al. 2009; Lagudah et al. 2009; Dakouri et al. 2010). Group 1 lines had the “C” allele at caSNP12, and the “null” allele at caIND11, whereas the other three groups had the “T” allele at caSNP12 and the “TTC” allele at caIND11. Group 1 consisted of all 108 lines that had the Lr34 resistance allele, whereas the other three groups had a susceptibility allele (Table 2). The three groups with susceptibility alleles were differentiated from each other by alleles at caIND10 and caSNP4 with group 2 having the “A” and “T” alleles, group 3 had the “null” and “A” alleles, and group 4 had the “null” and “T” alleles, for caIND10 and caSNP4, respectively. These four main groups defined by the gene-based markers were then further divided into sub-groups based on data from the linked markers (Table 2).

Table 1 Alleles for 10 molecular markers within or linked to the Lr34/Yr18 coding region
Full size table
Table 2 Profiles for 375 wheat cultivars and breeding lines based on 12 molecular markers within or linked to the Lr34 coding region
Full size table

Both caISBP1 and cam11 are linked tightly to the ABC transporter gene (Dakouri et al. 2010). There were only two alleles for caISBP1, a 509 bp fragment associated with the resistance allele for Lr34 and a 391 bp fragment associated with the susceptibility allele at this locus. The only exceptions to this pattern among the 375 lines were RL6058 (Thatcher*6/PI 58548), a Thatcher near-isogenic line containing Lr34, and its PI 58548 donor parent, which both had the 391 bp fragment (sub-group 1E, Table 2, Supplementary Table 1). Similarly, these two lines also had a 300 bp allele for cam11 whereas all the other lines with the resistance allele were null. This appears to be a relatively rare haplotype which could be the result of a recombination event between the caSNP4 of the ABC transporter and cam11. There were four alleles for csLVMS1 (224, 226, 228, 230 bp) (Table 1). All the lines that carried Lr34 had the 226 bp fragment, whereas most lines in the susceptible groups had the 228 bp fragment. Four sub-groups of lines without Lr34 had the 226 bp fragment, but this represented only 13 of the 267 lines without Lr34. RL6058 and PI 58548 which differed at caISBP1 and cam11, had the same allele at csLVMS1 as the rest of the lines with Lr34 suggesting a second recombination event between caISBP1 and csLVMS1. SSR marker swm10 had five alleles (Table 1) but only the 208 bp fragment was present in the lines that carried Lr34. The 208 bp allele was relatively rare among the lines without Lr34, found in only four lines from three groups. Lines without Lr34 typically had the 210, 212 or 214 bp fragments although one line (Synthetic W7984, a parent of the ITMI population) had a rare 194 bp fragment.

Only two alleles (csLV34a and csLV34b) were reported to date for the csLV34 marker. The csLV34a allele (229 or 255 bp in our study) was associated with the susceptible allele of Lr34, whereas the csLV34b allele (150 or 175 bp in our study) was associated with the resistance allele of Lr34 in lines and cultivars known to contain Lr34 (Lagudah et al. 2006). The susceptible cultivar Thatcher had the expected 255 bp csLV34a allele, whereas the near-isogenic line Thatcher-Lr34 (RL6058) had the 175 bp csLV34b allele. Recombination between csLV34 and Lr34 was previously reported (McCallum et al. 2008; Lagudah et al. 2009). Lines without Lr34 generally had the 255 bp allele except four sub-groups of lines, namely 2B (HY476), 2D (HY344), 4C (BW366 and BW807), and 4D (51 lines). Most of the 51 lines in sub-group 4D were related to the pre-harvest sprouting resistant line RL4137 (DePauw et al. 2009), also in this sub-group (Supplementary Table 1). AC Domain, previously postulated to have Lr34 (Liu and Kolmer 1997), was also in this sub-group. All lines with Lr34 had the 175 bp fragment except for group 1D (a breeding line designated 94B30-C6D) which had the 255 bp fragment. Recombination between csLV34 and Lr34 was previously thought to be rare (Lagudah et al. 2006) although in the current study there were five sub-groups (1D, 2B, 2D, 4C, 4D) in which there appeared to be a recombination between csLV34 and Lr34.

Wms1220 had nine alleles, the highest number of alleles among the markers in this study. This was the only marker tested on the proximal side of the ABC transporter gene (Dakouri et al. 2010). Haplotypes 1A, 1B, 1C/1D, and 1E with Lr34 were differentiated by different alleles at wms1220. There was diversity among the haplotypes with the susceptibility allele, but a substantial proportion of the members from each of groups 2, 3 and 4 (as defined by the gene-based markers caSNP12, caIND11, caIND10, and caSNP4) belonged to a small number of sub-groups or haplotypes. Group 2 lines were CPS cultivars and breeding lines (2B, 2D, 2E), parental lines from CIMMYT (2C) and North Dakota (2A), and early Canadian cultivars such as Garnet and Ruby (2F). Group 3 lines were a mixture of Canadian cultivars and parental lines from South America (Buck Manantial) and the U.S.A. (Ceres). Group 4 was the largest group, containing 220 lines. There were two large sub-groups, viz. 4D, with 51 lines represented by RL4137 and the cultivar AC Domain, and group 4E with 140 lines represented by Thatcher and AC Barrie. Many lines had the 4E haplotype because Thatcher was used extensively as a recurrent parent in backcrossing to recover its excellent end-use functionality and stem rust resistance. The 4D haplotype was common because RL4137 and AC Domain with resistance to pre-harvest sprouting were used frequently as parents (Townley-Smith and Czarnecki 2008).

Introduction and inheritance of Lr34 in Canadian wheat cultivars

Over 70% of the bread wheat grown in Western Canada is from the CWRS class, which has good end-use functionality for producing high quality flour and bread (McCallum and DePauw 2008). The major cultivars which initiated this type of wheat in the early days of production were Red Fife (introduced in 1870) and Marquis (1909) (McCallum and DePauw 2008). Neither of these cultivars had Lr34 (Fig. 1) and they are susceptible to leaf rust. Thatcher (released in 1935) replaced Marquis, due to stem rust resistance, but was very susceptible to leaf rust. Thatcher was partially replaced by Selkirk (1953), which had improved resistance to leaf rust due to Lr10, Lr14a, Lr16 (Martens and Dyck 1989) but did not have Lr34 (Fig. 1). Thatcher derivatives Manitou (1965) and then Neepawa (1969) replaced both Thatcher and Selkirk as dominant wheat cultivars in Western Canada and both were more resistant to leaf rust than Thatcher due to Lr13, donated by Frontana, but neither inherited Lr34 from Frontana (Fig. 1). Neepawa was replaced as the most popular cultivar in 1986 by Katepwa which had a similar level of resistance to leaf rust but also lacked Lr34 (Fig. 2). The first major CWRS cultivar to have Lr34 was Laura (registered in 1986) (Fig. 3), which could have inherited Lr34 from either Tobari 66 or Carazinho, as all three have the 1C haplotype (Table 2, Supplementary Table 1). CDC Teal (1991), another widely grown cultivar in Western Canada has Lr34 from BW38 (Sonora 64/Tezanos Pintos Precoz//Neepawa). BW38 inherited Lr34 from Tezanos Pintos Precoz since both Sonora 64 (2C) and Neepawa (4E) lack Lr34. CDC Teal, BW38 and Tezanos Pintos Precoz all belonged to haplotype 1C and carried the resistance allele of Lr34 (Fig. 2, Supplementary Table 1). Roblin (registered in 1986) and Pacific (1983) were both important cultivars and parental lines with Lr34 derived from the same parents (Fig. 3). Roblin inherited the 3C haplotype found in BW40, BW38 and BW15, which could have been inherited from either Tezanos Pintos Precoz (BW40 and BW38) or Tobari 66 (BW15) (Supplementary Table 1). However, Pacific inherited the 1C haplotype from RL4353 which likely was initially from Era, since it was reported to carry Lr34 (Ezzahiri and Roelfs 1989) and the csLV34b allele (Kolmer et al. 2008). After 1998, AC Barrie became the leading CWRS cultivar, replacing CDC Teal, Roblin and other cultivars (McCallum and DePauw 2008). Since AC Barrie did not have Lr34 (4E), the proportion the CWRS area seeded to cultivars with Lr34 declined during the years 1998 to 2005 (Fig. 4). The proportion of the total CWRS area seeded to cultivars with Lr34 increased slightly from 2007–2009 because many of the recently registered cultivars have Lr34, such as 5602HR, CDC Alsask, and AC Lillian, (Figs. 3, 6).

Fig. 1
figure 1

Early CWRS wheat cultivars. Cultivars in blue font have the resistance allele for Lr34, those in red had the susceptibility allele, and those in black were not tested. The year of registration and/or release is designated in brackets for Canadian wheat cultivars

Full size image
Fig. 2
figure 2

Neepawa derived CWRS wheat cultivars. Cultivars in blue font have the resistance allele for Lr34, those in red had the susceptibility allele, and those in black were not tested. The year of registration and/or release is designated in brackets for Canadian wheat cultivars

Full size image
Fig. 3
figure 3

Pacific, Roblin and Laura derived CWRS wheat cultivars. Cultivars in blue font have the resistance allele for Lr34, those in red had the susceptibility allele, and those in black were not tested. The year of registration and/or release is designated in brackets for Canadian wheat cultivars

Full size image
Fig. 4
figure 4

Proportion of the total CWRS seeded area which was seeded to Lr34 carrying cultivars. Data from Canadian Wheat Board varietal surveys (McCallum and DePauw 2008). No prairie wide variety surveys were conducted 1993–1997

Full size image

Cultivar Grandin was used as a parent to develop several CWRS cultivars. It was heterogeneous for the presence of Lr34 (results not shown) and did not donate Lr34 to the CWRS cultivars Superb, CDC Abound or CDC Go (Fig. 5). A group of CWRS cultivars have resistance to stem sawfly by having solid stem, derived from the line S-615 (Fig. 6). Early solid stem cultivars such as Rescue (1946), Chinook (1952), Cypress (1962) and Chester (1976) did not have Lr34, but it was introduced into the later solid stem cultivars Leader (1981), Lancer (1985) and AC Eatonia (1993) through the U.S. cultivar Chris. AC Lillian (2003), the most popular CWRS cultivar in 2007–2009, inherited Lr34 through BW621 since AC Lillian and BW621 have the 1A haplotype. AC Lillian’s other potential Lr34 donor parent Pasqua had the 1C haplotype (Fig. 6, Supplementary Table 1).

Fig. 5
figure 5

Modern CWRS wheat cultivars. Cultivars in blue font have the resistance allele for Lr34, those in red had the susceptibility allele, and those in black were not tested. The year of registration and/or release is designated in brackets for Canadian wheat cultivars

Full size image
Fig. 6
figure 6

Solid stem CWRS wheat cultivars. Cultivars in blue font have the resistance allele for Lr34, those in red had the susceptibility allele, and those in black were not tested. The year of registration and/or release is designated in brackets for Canadian wheat cultivars

Full size image

Canada Prairie Springs cultivars were originally produced from CIMMYT semi-dwarf cultivars (McCallum and DePauw 2008), which commonly carry Lr34. Despite the relationship to CIMMYT germplasm, nearly all the CPS cultivars registered to date lack Lr34 (Fig. 7). The exceptions are Genesis (1990) which inherited Lr34 from Pitic 62, and 5701HR which inherited Lr34 from either N89-3004 or N87-0446. Most CPS lines are derived from HY320, which has the susceptibility haplotype (2E), inherited by many of its progeny including AC Taber, AC Crystal and AC Foremost (Supplementary Table 1). This haplotype was also in Romany, a parent of HY320, but was not found among the tested CWRS cultivars or breeding lines.

Fig. 7
figure 7

CPS wheat cultivars. Cultivars in blue font have the resistance allele for Lr34, those in red had the susceptibility allele, and those in black were not tested. The year of registration and/or release is designated in brackets for Canadian wheat cultivars

Full size image

CWES cultivars are primarily derived from Glenlea (registered in 1972). Glenlea was the first major cultivar in Canada to have Lr34; it continues to be moderately resistant to leaf rust. Glenlea donated Lr34 to CWES cultivars Bluesky (1987), Amazon (1998), AC Corinne (1999), Glenavon (1999), Burnside (2004), CDC Rama (2002), and CDC Walrus (2004) since all these cultivars belong to the group 1B haplotype, whereas other potential donors Kitt, McNeal and Pasqua had the 1C haplotype (Fig. 8, Supplementary Table 1).

Fig. 8
figure 8

CWES wheat cultivars. Cultivars in blue font have the resistance allele for Lr34, those in red had the susceptibility allele, and those in black were not tested. The year of registration and/or release is designated in brackets for Canadian wheat cultivars

Full size image

The Canada Western Hard White Spring wheat class was initiated with the registration of Snowbird and Kanata in 2001. Both have a susceptibility haplotype at the Lr34 locus (group 4D). Two more recent hard white cultivars in this class namely Snowhite 475 and 476 also have susceptibility haplotypes belonging to groups 2E and 2B, respectively (Supplementary Table 1). The different haplotypes for these cultivars reflect their parentage since Snowbird and Kanata were derived partially from AC Domain (group 4D), whereas Snowhite 475 and 476 were derived from CPS parental cultivars in groups 2E and 2B.

Discussion

The only markers truly diagnostic for the presence of the resistance allele for Lr34 were those within the coding sequence of the ABC transporter gene, namely caSNP12 and caIND11. However, in another analysis of a broader world wheat collection, recombination between SNP4 and caIND11 was observed in two lines, but such a rare event was not observed in the 375 lines studied herein (Dakouri et al. 2010). Markers caISBP1 and cam11, located just upstream of the ABC transporter gene, were also diagnostic for Lr34, with the exception of RL6058 and its donor parent PI 58548. All lines without Lr34 had the 391 bp fragment for caISBP1 and the 297–302 bp fragment for cam11. The csLV34 marker was previously thought to be fairly diagnostic for the presence of Lr34 (Singh et al. 2007; Kolmer et al. 2008). While only two alleles were found for csLV34 in our study this marker was not diagnostic for Lr34 because five groups appeared to have recombination events between csLV34 and Lr34 (1D, 2B, 2D, 4C, and 4D). Recombinants included one resistant line (94B30-C6D) with the 255 bp fragment and 55 susceptible lines with the 175 bp fragment usually associated with resistance. In this latter case, the large number of lines within this recombinant group, probably reflects the relatedness of the Canadian cultivars in Group 4D. McCallum et al. (2008) and Lagudah et al. (2009) also reported recombination between csLV34 and Lr34 in a small number of cultivars. These results highlight the problems of using linked markers to postulate the presence or absence of a resistance gene. Since gene-based diagnostic markers are now available for Lr34, they should be used preferentially to identify the presence of this resistance allele.

Resistance gene Lr34 was not found in Canadian cultivars registered prior to Glenlea in 1972. Although Frontana, which carries Lr34 (Fig. 1), was used to add leaf rust resistance gene Lr13 to Manitou (1965) and Neepawa (1969) (McCallum and DePauw 2008), it did not transfer Lr34 to these cultivars. Once Lr34 was transferred into CWRS cultivars such as Laura (1986), Roblin (1986) and CDC Teal (1991), the wheat growing area seeded to cultivars with Lr34 rapidly increased and has remained relatively high ever since (Fig. 4). Because many recently released CWRS cultivars have Lr34, such as AC Lillian, CDC Alsask and 5602HR (Figs. 3, 6), it is possible to fix Lr34 in crosses by choosing adapted parental lines that harbour it. Retaining it in segregating crosses should be much easier using the diagnostic markers reported (Lagudah et al. 2009, Dakouri et al. 2010). Cultivars with Lr34, like Glenlea and Laura, have maintained at least a moderate level of resistance over many years of production in Canada. However, cultivars without Lr34 such as AC Barrie and Superb were resistant when released, due to race-specific resistance genes, but became susceptible over time due to changes in the pathogen population (McCallum et al. 2007a). This type of “resistance breakdown” has not happened to Canadian cultivars with Lr34.

Most CPS cultivars do not have Lr34 and the level of leaf rust resistance has generally not been as good within this class as within the CWRS class. Recent cultivar 5701HR has Lr34 and was the third most popular cultivar with approximately 15% of the CPS area seeded to this cultivar in 2007 (McCallum and DePauw 2008). This cultivar should facilitate and accelerate the incorporation of Lr34 in this class as it becomes used as a parental line for the development of new CPS cultivars. By contrast, the CWES class, which was founded by cultivar Glenlea, is essentially fixed for Lr34, with the exceptions of Wildcat (1987) and Laser (1997). Laser was a relatively popular cultivar from 2000 to 2007, ranking third during many years, although much of its production was outside of the leaf rust prone area.

Resistance gene Lr34 works in combination with other genes such as Lr10, Lr13, Lr14a, Lr16, Lr21 and Lr22a to condition resistance in Canadian wheat cultivars (McCallum et al. 2007a). There is a synergistic effect of Lr34 on other leaf rust resistance genes when it is used in combinations of two or more genes (German and Kolmer 1992). Cultivar Pasqua, with resistance genes Lr11, Lr13, Lr14b, Lr30 and Lr34 (Dyck 1993), has remained immune since it was registered in 1991 (unpublished data). Of those genes Lr11, Lr13 and Lr14b do not condition resistance when used individually, and Lr30 only gives an intermediate level of protection. However, in backcross Thatcher lines derived from Pasqua, combinations of one or more of these genes with Lr34 resulted in good resistance, and combinations of three or four genes involving Lr34 resulted in near immunity (McCallum and Thomas 2011).

A few cultivars with susceptibility haplotypes at the Lr34 locus are still very resistant to leaf rust due to other resistance genes. McKenzie, for example, is currently resistant due to the combination of Lr10, Lr13, Lr16 and Lr21 (McCallum and Seto-Goh 2010). Since Lr10 and Lr13 are currently not effective when used alone and Lr16 conditions only partial resistance, Lr21 essentially provides all the resistance in McKenzie. If virulence to Lr21 were to develop in the pathogen population, McKenzie and other Canadian cultivars that depend on Lr21 could become susceptible. If Lr34 could be incorporated into cultivars with other effective genes such as Lr21, a more durable and effective level of resistance might be achieved. Virulence to Lr21 has recently been reported in the U.S.A. (Kolmer 2011) which could jeopardize the resistance of cultivars that currently rely on Lr21 for leaf rust resistance.

Most Canadian wheat cultivars are susceptible to stripe rust, except those with the moderate resistance conditioned by the Lr34/Yr18 locus (McCallum et al. 2007b). Stripe rust has become an increasing problem in the Central Great Plains of the U.S. since 2000 and could threaten wheat production in Canada (Chen et al. 2010). Therefore, increased incorporation of Lr34/Yr18 into future cultivars could be useful in protecting the Canadian wheat crop against stripe rust. Stem rust resistance is also conditioned by the Lr34/Yr18 locus (Dyck et al. 1985, Hiebert et al. 2010, McCallum et al. 2011). The central role of Lr34/Yr18 in multiple disease resistance makes knowledge of this gene an important consideration in any set of wheat germplasm. Given the availability of functional gene-based markers, a good initial step in developing rust resistance for wheat breeding programs is to determine the presence or absence of Lr34/Yr18 in cultivars, lines in development and potential parental lines. This was previously done for wheat cultivars from Australia (Singh et al. 2007) and an international collection (Kolmer et al. 2008) using linked markers. Gene pyramids which involve Lr34, such as Pasqua (Dyck 1993; McCallum and Thomas 2011), which have been demonstrated to be effective and durable, can be developed once Lr34 is fixed within parental lines or selected in progeny populations.

This study demonstrates the utility of gene-based molecular markers within the coding sequence of Lr34 and linked markers. The determination of the presence or absence of this important resistance allele in all the major bread wheat classes in Canada should be instructive in its incorporation and retention in these breeding programs. The introduction of Lr34 occurred repeatedly within these various classes, as CIMMYT derived cultivars, such as Tezanos Printos Precoz, were common donors in Canadian breeding programs. Recently released Canadian wheat cultivars increasingly rely, at least in part, on Lr34/Yr18 for rust resistance. To minimize losses due to cereal rusts, future breeding efforts should focus on its incorporation into wheat cultivars in Canada and internationally.