Resistance to leaf rust in cultivars of bread wheat and durum wheat grown in Spain

A set of bread wheat and durum wheat cultivars adapted to Spanish conditions was tested for resistance against leaf rust caused by different pathotypes of Puccinia triticina in field trials and in growth chamber studies. Lower levels of resistance were found in durum wheat than in bread wheat. The most frequent Lr genes found in bread wheat were Lr1, Lr10, Lr13, Lr20, Lr26 and Lr28. In durum wheat, additional resistance genes that differed from the known Lr genes were identified. The level of partial resistance to leaf rust was in general low, although significant levels were identified in some bread wheat and durum wheat cultivars.

Keywords: Triticum aestivum; T. turgidum; Puccinia triticina; resistance

Leaf rust caused by Puccinia triticina is an important disease in bread wheat (Triticum aestivum ssp. aestivum) and durum wheat (T. turgidum ssp. durum) in most wheat‐growing areas ([28], [3]). It should be noted that not all researchers agree that P. triticina is the causal agent of leaf rust of durum wheat. [1] isolated a type of leaf rust from durum wheat in Morocco different from the P. triticina type collected on bread wheat, with a different substomatal vesicle and alternative host (Anchusa azurea). [33] also confirmed this finding. They stated that some Moroccan isolates collected on durum wheat are indeed clearly different from P. triticina and very similar to P. recondita f. sp. secalis, based on DNA sequence analysis. However, most of the isolates collected worldwide on durum wheat match the P. triticina type. Histological observations were carried out on some of the pathotypes collected on durum wheat and all of them displayed a substomatal vesicle similar to P. triticina, suggesting, with a high probability, that most of the Spanish durum wheat leaf rust is caused by P. triticina.

Leaf rust is an important disease in Spain ([18]) and in recent times it has been particularly severe on durum wheat ([2]). Breeders have most commonly used resistance of a hypersensitive type. This is easily recognizable by the presence of macroscopically visible necrosis, and is usually associated with complete resistance. Therefore, it is very effective in controlling the disease. So far, 50 resistance loci conferring resistance to P. triticina (Lr genes) have been reported ([3]), most of which originate from bread wheat and some of which were introgressed into wheat from wild relatives or even from rye. Knowledge of the Lr genes used in wheat cultivars is important to predict their reaction to pathotypes present in a particular region or to other pathotypes from neighbouring areas that could, eventually, be introduced.

A drawback to the use of Lr genes is their limited durability; they are often overcome by new pathotypes of the pathogen. It is often accepted that the use of so‐called 'partial resistance' offers a better chance of achieving durable resistance ([5], [7]). Partial resistance is characterized by a reduction of the epidemic build‐up in spite of a high infection type (absence of a hypersensitive reaction) ([20]). Partial resistance to leaf rust is common in commercial barley varieties ([19]) but it is less common in wheat ([15]).

The aim of this study was to identify Lr resistance genes present in a collection of bread wheat and durum wheat cultivars grown in Spain and to determine the level of partial resistance to leaf rust in these cultivars.

Plant materials:  A collection of 28 commercial cultivars of bread wheat, Triticum aestivum L., and 41 cultivars of durum wheat, T. turgidum ssp. durum, was used in this study. These were cultivars believed to be more adapted to southern Spanish conditions that had been selected for testing prior to recommendation to farmers by Red Andaluza de Experimentación Agraria (RAEA).

Field experiments:  Resistance to leaf rust was tested in RAEA field trials at six locations across Andalusia (southern Spain) during three consecutive field seasons (1997–1998, 1998–1999, 1999–2000). Four replicates of each cultivar were established in 10 × 1 m plots at each location. No artificial inoculation was performed, as heavy infections of leaf rust are known to occur in the area ([18]). Disease severity was assessed once when leaf rust reached its peak, which was about 14–20 days after heading, as a percentage of infected leaf area according to the modified Cobb's scale ([22]). To calculate a balanced average, original disease scores were converted into relative values, referred respectively to the susceptible bread wheat cultivar 'Yécora' and the durum wheat cultivar 'Arcobaleno'. The mean relative disease severity (RDS) per accession per location was calculated as a mean of the scores.

Postulation of Lr resistance genes:  The collection was inoculated with several pathotypes of known virulence, many of which were collected from the field, plus other clearly distinct pathotypes. Bread wheat cultivars were inoculated with 12 isolates (Table 1) and durum wheat cultivars with seven isolates (Table 2). Fourteen‐day‐old seedlings, grown in 0.1‐l pots, were inoculated with spores mixed with talcum powder (1 : 30 v/v) resulting in a deposition of about 30 spores/cm2. The inoculation was performed by blowing the mixture of spores and talcum over the seedlings. Four seedlings were inoculated per accession in one replication. Plants were incubated for 24 h in darkness at 20°C and relative humidity at saturation. They were then transferred to a growth chamber at 20°C, with 112 μmol/m2 s of light intensity and a 14‐h photoperiod. Twelve days after inoculation, infection type (IT) was recorded using a 0–9 scale adapted from [16]. ITs 7 or higher were regarded as compatible, and were referred to as high IT. ITs lower than 7 were referred to as low IT, indicative of hypersensitive resistance. Most of the time there was agreement in the IT of the four leaves. Only a few times one leaf presented an IT different from the other three. In such an uncommon event, the IT of the majority of the leaves was taken. The presence of leaf rust (Lr) resistance genes in the seedling stage was postulated by comparing the low and high ITs displayed by the cultivars with the IT of known Lr genes in the Thatcher isolines. For example, a pathotype avirulent on Lr26, which produces a high IT on a certain cultivar indicates the absence of Lr26 in that cultivar. Bread wheat pedigrees of the cultivars 'Babui', 'Barbol', 'Dariel', 'Perico' and 'Surco' were very useful in discarding several Lr genes. The candidate genes used for postulation were the most common Lr genes in bread wheat (Lr1, Lr2a, Lr2b, Lr2c, Lr3, Lr10, Lr13, Lr16, Lr17, Lr20, Lr23, Lr24, Lr26 and Lr28; [29], [34]). As tests were performed on seedlings, adult plant resistance genes such as Lr12, Lr22b, Lr35 and Lr37 could not be postulated.

1  Seedling reactions of bread wheat cultivars grown in Spain when tested with 12 pathotypes of P. triticina; the relative disease severity in the field (RDS) is also shown

1 1Infection types according to [16].

• 2 2Latency period measured in first leaves with the leaf rust pathotype MCSTTM (in case the cultivar shows a compatible infection type). Data referred to susceptible check 'Little Club' (189 h = 100%).
• 3 3Mean relative disease severity in the field was referred to the cultivar Yecora in each location (=100%) during 1998 and 1999.
• 4 4Nomenclature according to [12].
• 5 5Cultivars with the letter 'a' displayed a higher latency period than Little Club (LSD 0.05).
• 6 6m: missing data.
• 7 7Different combinations of Lr genes explaining resistance to P. triticina are separated by a bar.
• 8 8?: unknown Lr gene.
• 2  Seedling reactions of durum wheat cultivars grown in Spain when tested with seven pathotypes of P. triticina (infection type and latency period); relative disease severity (RDS) in the field and postulation of virulence patterns are also shown

• 9 1Infection type according to [16].
• 10 2Latency period measured in first leaves with the leaf rust pathotype DBGCPL (in this case the cultivar shows a compatible infection type). Data referred to susceptible check 'Little Club' (184 h = 100%).
• 11 3Relative disease severity in the field was referred to the cultivar 'Arcobaleno' for each location and year (1999 and 2000).
• 12 4Nomenclature according to [12].
• 13 5Different combinations of Lr genes (or virulence pattern) explaining resistance to P. triticina are separated by a bar. Postulated virulence pattern (An = 'Antón', Ar = 'Ardente', Bo = 'Bólido', Br = 'Bravadur', Ca = 'Canyon', Cl = 'Claudio', Cr = 'Creso', Ma = 'Manolete', Ne = 'Nefer', Sen = 'Senadur', Si = 'Simeto').
• 14 6m: missing data.
• 15 7Cultivars with letter 'a' displayed a higher latency period than Little Club (LSD 0.05).

Evaluation of the level of partial resistance:  Latency period (LP) was measured at the seedling stage. Pathotypes were selected for their wide virulence against either bread or durum wheat in order to overcome the effect of a hypersensitive reaction of the Lr genes that masks partial resistance. Pathotype MCSTTL (virulent on Lr1, Lr3, Lr3ka, Lr10, Lr13, Lr14a, Lr15, Lr17a, Lr18, Lr20, Lr21, Lr23 and Lr26), named according to [12], was used for bread wheat, whereas DBGCPL (virulent on Lr2c, Lr13, Lr14a, Lr18, Lr20, Lr21 and Lr23) was used for durum wheat. Plants were sown in trays (35 × 20 × 8 cm) that included eight cultivars plus the susceptible check cultivar 'Little Club' and the partially resistant bread wheat cultivar Akabozu. About 12 days after sowing, the first leaves were fixed in a horizontal position and inoculated in a settling tower to allow uniform deposition of spores. Three milligrams of spores were mixed with talcum powder (v/v 1 : 10) and applied in a settling tower. The deposition resulted in approximately 150 spores/cm2. The trays were incubated and transferred to a growth chamber with conditions as described above. Five seedlings per accession were evaluated in one replication. Pustule counts were carried out daily from the beginning of sporulation until the number of pustules no longer increased. LP was estimated by interpolation as the time period from the beginning of incubation until 50% of the total number of pustules appeared ([20]). LPs were referred in each tray to the cultivar 'Little Club' (=100%).

Leaf rust resistance in the field was common in bread wheat, with about half of the cultivars displaying RDS lower than 5 (Table 1). Complete resistance (RDS = 0) was observed at all locations in 26% of the cultivars ('Babui', 'Farak', 'Greina', 'Panabon', 'Perico' and 'Podenco').

All bread wheat cultivars were resistant (low IT) to at least one isolate. The results of postulation of Lr genes are also presented in Table 1. It was possible to postulate the presence of one Lr gene or a combination of Lr genes in 15 cultivars. A combination of two or more Lr genes was necessary to explain the resistance of nine cultivars. The resistance could not be explained by the presence of any of the Lr genes available for the study in three of the cultivars. Lr20 was detected in six cultivars, Lr1 in five and Lr26, Lr10 and Lr28 were postulated in two cultivars. Lr13 was detected in four cultivars although it might be present in up to 19 (60% of total). Although Lr13 is an adult plant gene, it can be detected at the second leaf stage ([24]).

Leaf resistance in the field was lower in durum wheat than in bread wheat, with only 9% of the cultivars displaying an RDS lower than 5 (Table 2) and only 6% of the cultivars displaying complete resistance. Two cultivars ('Manolete and Pedroso') out of 41 showed a high IT to all pathotypes. The remaining accessions showed a low IT to at least one pathotype. All cultivars except 'Senadur' and 'Simeto' showed an IT pattern that did not correspond with known Lr genes. The possible resistance gene (or gene combinations) was given a tentative designation based on one cultivar displaying that particular resistance pattern. The most frequent pattern was An (from Anton), shared by 17 cultivars. The cultivars with the pattern Ar ('Ardente' and 'Aronde') and Cr ('Creso' and 'Colosseo') had a very low disease severity in the field.

Two bread wheat cultivars (Babui and Farak) displayed longer LPs than the check and showed a low disease severity in the field. Three durum wheat cultivars ('Bolenga', 'Bravadur' and 'Duilio') displayed a longer LP than the check and intermediate disease severity in the field.

All bread wheat cultivars carried at least one Lr gene, which indicates that hypersensitive resistance to P. triticina is common in these cultivars. Lr13 is present in many bread wheat cultivars ([29]) and has been used frequently in bread wheat breeding ([30], [32], [4], [11]). The reason for the presence of Lr13 in many cultivars, apart for the resistance it confers, might be due to an interactive effect with genes such as Lr34 and Lr37 ([10]), although this effect has been questioned by other researchers ([23], [9]). Another gene identified very frequently in bread wheat is Lr1. This gives an imune IT both in the seedling and the adult plant and was very effective in some locations, although virulence to this gene is quite common. Lr1 and Lr13 are examples of resistance genes extensively used in the past in which virulence is rather a rule than an exception today ([13]). Lr26 was detected in at least three cultivars and confers a good level of resistance in the field to avirulent isolates. This gene originated from a translocation of a segment of chromosome 1RS of rye incorporated in the chromosome 1BL of wheat ([35]) and also carries Yr9 (yellow rust resistance), Sr31 (stem rust resistance) and probably Pm8 (powdery mildew resistance). The large number of wheat cultivars carrying this translocation resulted in the development of virulence to this gene in the USA and Eastern Europe ([17], [13]). Another gene is Lr20 that is effective at low temperatures but is ineffective above 25°C ([8]). To avoid resistance genes being overcome by the pathogen, it would be better to use them in combination (pyramiding). Various cultivars displayed a high latency period that could possibly be due to the presence of Lr34, common in CIMMYT material ([29]), or to other polygenes for partial resistance ([6]). Molecular markers associated with Lr genes or with QTLs for partial resistance could be very valuable in evaluating the presence of those genes in a collection of cultivars ([25]).

Most of the durum wheat tested carried at least one hypersensitive resistance gene. In two cultivars only, no hypersensitivity could be detected against any of the pathotypes used. The most interesting aspect of the hypersensitive resistance of durum wheat is that in only two cultivars did a combination of known Lr genes explain the resistance. The remaining cultivars might carry genes different from the known Lr genes. [36] studied the genetics of resistance to P. triticina in a collection of durum wheat cultivars and reported that the resistance gene combinations were different from known Lr genes, except for Lr16 and Lr17 that could be present in the some cultivars, although not detected in the present study. In fact, in the catalogue of resistance genes listed in the [3], the genes from durum wheat are listed separately from those of bread wheat. The resistance conferred by the pattern Ar and Cr is very effective in the field. The durum wheat 'Creso' is very resistant to leaf rust and its resistance is considered to be durable ([21]). There is evidence that its durability is based on a combination of hypersensitive resistance and partial resistance ([14]). The combination of different mechanisms of resistance in a cultivar can be very effective in increasing and protecting resistance ([27]).

Some cultivars have been found with a certain level of partial resistance, displaying moderate infection in the field and new references to such sources of partial resistance in durum wheat would be valuable.

The results obtained indicated that of the cultivars tested, modern durum wheat cultivars were more susceptible to leaf rust than bread wheat cultivars. Probably this is due to the fewer efforts made to incorporate leaf rust resistance in durum‐wheat breeding when compared with bread wheat. Breeding for resistance to wheat leaf rust has been one of the main goals in most bread wheat breeding programmes ([26], [31], [29]). However little attention has been given to durum wheat which has been cultivated to a smaller extent. Durum wheat cultivars such as 'Creso' or 'Colosseo', with high levels of combined mechanisms of resistance to P. triticina, seem to be very interesting and might be useful donors of resistance in breeding programmes.

The authors acknowledge RAEA (Red Andaluza de Experimentación Agraria) for establishing the field experiments and for providing seeds of the cultivars for this work and projects AGL2005‐01781, PTR95‐0635‐OP and PIA‐03‐059 for financial support. They also thank A. Moral, J.J. Perez and A. Castilla for technical assistance.