Physiological specialization and pathotype distribution of Puccinia recondita in western Europe, 1995

Abstract

A pathogenicity survey of Puccinia recondita f.sp. tritici (Prt) was conducted in western Europe in 1995. Random urediospore isolates (850) of Prt were collected from the air by means of a jet spore sampler in wheat-growing regions of Austria, Belgium, France, Germany, northern Italy, Switzerland and the UK. Pathogenicity of the isolates was determined in tests of detached primary leaf segments maintained on water agar supplemented with benzimidazole (35 p.p.m.). The differential genotypes used were Thatcher, 20 near-isogenic Thatcher lines each with a single leaf rust resistance gene, and five cultivars/lines with additional resistance genes. All isolates were avirulent for the genes Lr9, Lr19, Lr21, Lr24, Lr25 and Lr29, and both virulence and avirulence were detected for the remaining 19 genes. Fifty-three pathotypes were identified, four of which predominated (64% of isolates) and were widespread throughout western Europe. Three of the four predominant pathotypes were also identified in collections of wheat leaf rust collected in Poland, Hungary, Estonia and Finland. One pathotype, which comprised 35% of isolates in the south of France, was not detected in any other region. This pathotype was indistinguishable from several isolates obtained from Morocco, which suggested that it may have originated from northern Africa. Comparisons with previously published data suggested that the four predominant pathotypes were very similar and possibly the same as pathotypes present in the former Czechoslovakia for up to 20 years. The results obtained provide evidence of migration of Prt over considerable distances in western Europe, stressing the need for a co-ordinated approach for genetical control of the disease in this region.

Introduction

Wheat leaf rust (caused by Puccinia recondita f.sp. tritici; Prt) is widespread throughout Europe, and is regarded as an important disease in many eastern (Bartošet al., 1992; Ittu et al., 1992; Manninger, 1992) and western (Forrer, 1988; Poinso & Ollivier, 1988; Casulli & Pasquini, 1992) European countries. Zadoks & Bouwman (1985) considered the disease a threat in north-western regions of Europe, as well as a permanent problem in southern and south-eastern regions. Estimates of yield losses of up to 10% (Manninger, 1992) and 25% (Poinso & Ollivier, 1988) caused by the disease were reported in wheat crops in Europe. From data collected between 1980 and 1983, Priestley & Bayles (1988) estimated that annual yield losses attributed to leaf rust in the most susceptible wheat cultivars in the UK totalled about £5 million.

Attempts to develop cultivars with genetic resistance were initiated in some European countries (Clifford & Harris, 1981; Bartošet al., 1992; Casulli & Pasquini, 1992). Pathogenicity surveys of wheat leaf rust have been ongoing in several European countries (Bartošet al., 1992; Casulli & Pasquini, 1992; Jones & Clifford, 1995), but differences in differential lines and in the techniques used have made comparisons between countries difficult. In view of this, a pathogenicity survey was conducted for wheat leaf rust in 1995, in which isolates were collected from most wheat-growing regions of western Europe. Isolates were also obtained from some neighbouring countries, to allow further comparisons.

Materials and methods

Collection and establishment of single-pustule isolates

Random isolates of Prt were collected by a jet spore sampler mounted to the roof of a car (Schwarzbach, 1979; Limpert et al., 1984). The car was driven through 31 wheat-producing areas of western Europe (Fig. 1) during June and July 1995. Spores sampled from the air settled on detached leaf segments of cultivar Kanzler (30 mm), maintained in Petri dishes containing water agar (6 g agar L⁻¹) supplemented with benzimidazole (35 mg L⁻¹) (WAB). Petri dishes were replaced at 50-km intervals. To simplify data presentation, results for some regions that were close to each other were combined.

Figure 1. Map of western Europe indicating the regions sampled of Puccinia recondita f.sp. tritici in 1995.

Individual pustules that developed on the leaf segments were subcultured by sterile needle transfer onto new leaf segments of cultivar Kanzler. Following inoculation and to induce infection, plates containing leaf segments were incubated at 100%RH, 17°C, in darkness for 18 h, after which they were incubated at 17°C with continuous light (20 μmol m⁻² s⁻¹). For longer-term storage (up to 6 weeks), plates were transferred to a refrigerator (c. 7°C) with low light (c. 10 μmol m⁻² s⁻¹) once pustules ruptured the epidermis of leaf segments.

Single-uredial isolates (60) were also established from infected leaves collected from two regions in which spore trapping was conducted (I3, Verona-Venedig, northern Italy; F9, Bourges-Nevers, central France), plus leaves collected in 1995 in Estonia, Finland, Hungary, Morocco, Poland and Siberia. In these cases, a bulk subculture was established directly from the infected leaves, from which two to three single-uredial isolates were established and applied to differential sets.

Growth and preparation of host material

Wheat seedlings used for maintaining and pathotyping isolates of Prt were grown in pots (70 mm diameter) containing peat moss, in a growth room at c. 20°C and under continuous light (c. 200 μmol m⁻² s⁻¹). Pots were covered with cellophane bags to prevent contamination with exogenous spores, and were watered by capillary action from soaked matting via a hole in the base of the pot.

All tests were conducted on detached segments of the primary leaves of 9–10-day-old seedlings, maintained in square plastic dishes (180 × 120 mm) containing WAB. The plastic dishes had 12 compartments (30 × 20 mm) and the lids had two small holes (c. 1-mm diameter) to provide exchange of gas whilst maintaining high humidity. Preparation of the agar, the plates, and cutting of leaf segments was performed as described previously ([9]Felsenstein et al., in press).

Differential test sets

The differential set used comprised leaf segments (30 mm) of Thatcher and the lines near-isogenic to Thatcher but carrying resistance genes Lr1, Lr2a, Lr2b, Lr2c, Lr3a, Lr3bg, Lr3ka, Lr9, Lr10, Lr11, Lr13, Lr15, Lr17, Lr19, Lr21, Lr23, Lr25, Lr26, Lr29 and Lr30. The Thatcher near-isogenic series was developed by Dr P. L. Dyck, Agriculture Canada. The additional cultivars Thew (Lr20), Agent (Lr24), Gatcher (Lr27 + Lr31), CS 2A/2M (Lr28) and Harrier (LrH; McIntosh et al., 1995) were included in the set because Thatcher derivatives with these genes were not available. The identity and purity of all differential lines was checked in Australia before tests were conducted, using Australasian isolates of Prt of known pathogenicity.

Previous studies (Felsensteinet al., in press) established good correlations between the expression of genes Lr1, Lr2a, Lr2b, Lr2c, Lr3a, Lr3bg, Lr3ka, Lr9, Lr15, Lr17, Lr19, Lr20, Lr23, Lr24, Lr25, Lr26, Lr27 + Lr31, Lr28, Lr30 and LrH in intact seedlings and detached leaf segments. As post-inoculation temperature was important for the reliable expression of some genes (Felsensteinet al., in press), leaf segments were incubated at either 17°C (Lr1, Lr2a, Lr2b, Lr3a, Lr3bg, Lr3ka, Lr9, Lr11, Lr15, Lr20, Lr25, Lr26, Lr28) or 23°C (Lr2c, Lr10, Lr13, Lr17, Lr19, Lr21, Lr23, Lr24, Lr27 + Lr31, Lr29, Lr30 and LrH) following inoculation and infection. The expression of genes Lr10, Lr11, Lr21 and Lr29 in leaf segments was found by Felsensteinet al. (in press) to be unreliable, and so tests on these were repeated several times.

Inoculation of differential lines

Single-uredial isolates were applied to differential sets using a settling tower, which was placed over the plates with leaf segments (180 × 120 × 250 mm high). Urediospores were sucked into an eye dropper pipette using a rubber teat, which was then removed. The wider end of the pipette was placed through a hole (7 mm diameter) in the top of the settling tower, and spores were blown into the tower by air supplied by a 10-ml syringe connected to the narrow end of the pipette by plastic tubing. Infection was induced as previously described.

Disease assessment and pathotype nomenclature

Plants were assessed for disease response at 9 days (23°C) or 12 days (17°C) after inoculation using the 0–4 infection type scale (McIntosh et al., 1995).

The pathotypes identified in the present study were given UN numbers (Basile, 1957), based on pathogenicity on differentials with the genes Lr1, Lr2a, Lr2c and Lr3a.

Results

Isolates collected by spore trapping

Isolates of Prt were recovered from all 31 regions sampled. A total of 850 single-uredial isolates was successfully established and 53 virulence/avirulence combinations (pathotypes) were identified on the differential lines. The pathotypes grouped into seven UN races, and were also numbered from 1 to 53 to account for pathogenic variability on the remaining Thatcher lines and cultivars (Table 1).

Table 1.  Pathotypes, Unified Numeration (UN) codes, and number of isolates detected of all pathotypes of Puccinia recondita f.sp. tritici identified in regions of western Europe in 1995 ^a  Tested for pathogenicity on Thatcher, lines near-isogenic to Thatcher carrying resistance genes Lr1, Lr2a, Lr2b, Lr2c, Lr3a, Lr3bg, Lr3ka, Lr9, Lr10, Lr11, Lr15, Lr17, Lr19, Lr21, Lr23, Lr25, Lr26, Lr29 and Lr30, and Thew (Lr20), Agent (Lr24), Gatcher (Lr27+Lr31), CS 2A/2M (Lr28) and Harrier (LrH; McIntosh et al., 1995).^b Unified Numeration number as determined from Basile (1957).

All isolates produced low infection types for the genes Lr9, Lr19, Lr21, Lr24, Lr25 and Lr29. One group of pathotypes (typified by pt 17 and including pts 4, 14, 15 and 17) gave a distinctly higher infection type (;++C) on the differential line with Lr9 than did all other pathotypes (0;=). Isolates were polymorphic for pathogenicity to the remaining 19 genes.

With the exception of only two pathotypes (pts 18 and 33), which comprised about 1% of the isolates examined, all pathotypes were either avirulent for the genes Lr3a, Lr3bg, Lr3ka and Lr30, or virulent for these genes (Table 1). Pathotypes virulent for these four genes could be divided further into three groups (pts 19–47, 48–51 and 52–53) on the basis of pathogenicity for Lr1, Lr2a and Lr2b.

Four pathotypes (pts 6, 17, 36 and 39) were predominant, representing approximately 64% (18, 20, 10 and 18% respectively) of the isolates analysed. These pathotypes were isolated from most of the regions sampled (Table 2), which indicated that they were widespread throughout western Europe in 1995. In contrast, pt 3, which was very common in southern France (regions F1 and F2; 35% of isolates), was not found in any other region, and pts 52 and 53 were restricted to Austria (Table 2). Although widespread, pts 6, 17, 36 and 39 were not evenly distributed throughout the regions. For example, pt 17 was common in regions sampled in Belgium, northern France and the UK, and pt 39 was common in regions sampled in central and southern France and in northern Italy (Table 2).

Table 2.  Frequency (%) of representative pathotypes of Puccinia recondita f.sp. tritici identified in regions of western Europe in 1995 ^a  Pathotype numbers as given in Table 1.^b Tested for pathogenicity on Thatcher, lines near-isogenic to Thatcher carrying resistance genes Lr1, Lr2a, Lr2b, Lr2c, Lr3a, Lr3bg, Lr3ka, Lr9, Lr10, Lr11, Lr15, Lr17, Lr19, Lr21, Lr23, Lr25, Lr26, Lr29 and Lr30, and Thew (Lr20), Agent (Lr24), Gatcher (Lr27+Lr31), CS 2A/2M (Lr28) and Harrier (LrH; McIntosh et al., 1995).

Isolates established from field-collected infected leaves

Thirty-three and 28 isolates were established from infected leaves collected in regions I3 and F9, respectively. All pathotypes identified were also isolated from the respective regions by spore trapping, and all were among the most common in the regions as determined by spore trapping (Tables 2 and 3).

Table 3.  Frequency (%) of pathotypes established from infected leaves collected from northern Italy (I3, Verona-Venedig) and central France (F9, Bourges-Nevers) in 1995 ^a  Pathotype numbers as given in Table 1.

All of the isolates established from infected leaves collected in Estonia, Finland, Hungary, Morocco, Poland and Siberia had virulence/avirulence formulae similar to those of pathotypes isolated from spore trapping in western Europe (Table 4). Several of the pathotypes common in western Europe were also found in other regions. For example, pt 36, which was detected in all but three of the regions sampled by spore trapping, was also isolated from infected leaves collected from Estonia, Hungary and Poland (Table 4). All isolates obtained from Siberia resembled pts 52 and 53 in possessing virulence for Lr1, all three Lr2 alleles, and all three Lr3 alleles.

Table 4.  Pathotypes of Puccinia recondita f.sp. tritici identified from infected wheat leaves collected from several regions in 1995 ^a  Tested for pathogenicity on Thatcher, lines near-isogenic to Thatcher carrying resistance genes Lr1, Lr2a, Lr2b, Lr2c, Lr3a, Lr3bg, Lr3ka, Lr9, Lr10, Lr11, Lr15, Lr17, Lr19, Lr21, Lr23, Lr25, Lr26, Lr29 and Lr30, and Thew (Lr20), Agent (Lr24), Gatcher (Lr27+Lr31), CS 2A/2M (Lr28) and Harrier (LrH; McIntosh et al., 1995).^b Pathotype numbers as given to isolates collected from western Europe in 1995 (see Table 1).

Discussion

The predominance of four pathotypes of Prt throughout the survey region provides strong evidence of clonal lineages and long-distance aerial migration of this pathogen within Europe. This is further supported by the detection of some of these pathotypes in countries such as Finland, Poland and Hungary. Hermansen et al. (1976) presented evidence of long-distance transport of viable spores of Piccinia hordei and Prt in Europe, and results of pathogenicity surveys with other wind-dispersed cereal pathogens, such as the mildew pathogens of barley (Erysiphe graminis f.sp. hordei; Limpert, 1987; Limpert et al., 1990) and wheat (E. graminis f.sp. tritici; Felsenstein et al., 1987), and the wheat stripe rust pathogen (Puccinia striiformis f.sp. tritici; Zadoks, 1965), have provided evidence of considerable long-distance dispersal of spores within Europe. These results clearly demonstrate the importance of studying populations of wind-dispersed plant pathogens in Europe on a continental scale. Migration and long-distance dispersal are also known to be important in the epidemiology of cereal rusts in other parts of the world, such as Australasia (Luig, 1985; Park et al., 1995), India (Nagarajan & Joshi, 1985) and North America (Roelfs, 1985).

The aecial stage of Prt has been observed on alternate hosts in Portugal (Freitas, 1957), southern Italy (Casulli, 1988) and Morocco (Ezzahiri et al., 1992), although the contribution of these hosts to the epidemiology and to the pathogenic variability of Prt in Europe as a whole is not known. Asexual reproduction is obviously important in Europe given the predominance of four pathotypes in the survey region. Ongoing surveys in the former Czechoslovakia (Bartošet al., 1992), Hungary (Manninger, 1992), Italy (Pasquini & Zitelli, 1984) and in the UK (Jones & Clifford, 1995) have indicated the recurrence of specific pathotypes over a series of years, again implicating the uredial stage in propagation and overseasoning of the pathogen.

Given the extent of pathogenic variability in Prt in western Europe, and the various differentials used in different studies, caution is required in comparing published results. Similarities were none the less apparent between some of the pathotypes identified in the present study and those reported by other workers. For example, pts 13 and 17 conformed to International Standard Race (ISR) 14 (Johnston & Browder, 1966), with added virulence for Lr26. Bartošet al. (1992) reported the occurrence of an isolate of ISR 14 with virulence for Lr26 (14SaBa), which was present in the former Czechoslovakia since the early 1970s. In the present study, pathotype 17 predominated in the north of France, Belgium and the UK, where it comprised 78% of isolates examined. Jones & Clifford (1995) reported the predominance of one pathotype (Wheat Brown Rust Virulence formula 1,2,6) in the UK for the years 1989–1995 and, as far as can be determined, this pathotype appears to be the same as pt 17 reported here. As with pathotype 17, the majority of isolates of WBR1,2,6 found in the UK in recent years have produced a higher infection type on the differential line with Lr9 (E. R. L. Jones, 1996, personal communication, Institute of Grassland and Environmental Research, Aberystwyth, UK), further supporting the similarity of these isolates. This pathotype is virulent for Lr26, and its increase in frequency from 1989 to 1992 was attributed to the popularity of cultivars with Lr26 in the UK during this period (Jones & Clifford, 1992). It therefore seems that pt 17, which occurred widely in western Europe in 1995, may be been common in recent times in the UK and the former Czechoslovakia.

Pathotype 52, which was isolated only in Austria, conformed to race 77 with added virulence for Lr26. Race 77 was important in eastern Europe during the 1960s and 1970s, when it was one of the most commonly isolated races in Czechoslovakia (Bartošet al., 1992) and Hungary (Manninger, 1992), and it also predominated in Romania from 1973 to 1980 (Ittu et al., 1992). Despite the predominance of race 77 in eastern Europe, it did not become frequent in western regions of Europe (Zadoks & Bouwman, 1985), and its predominance in eastern regions was attributed to the widespread cultivation of wheats possessing Lr3a and Lr26 both singly and in combination (Bartošet al., 1992). All isolates from Siberia also conformed to standard race 77.

Nearly all pathotypes detected could be placed into two groups on the basis of virulence or avirulence for the genes Lr3a, Lr3bg, Lr3ka and Lr30. Although regional differences were apparent in the distribution of individual pathotypes, there was no pattern in the distribution of the two groups of pathotypes apart from a reduced frequency of those with virulence for alleles at the Lr3 locus and Lr30 in northern France, Belgium and the UK. Reasons for the apparent association between pathogenecity for these four genes in Europe are not known. In an international survey on the pathogenicity of Prt, Huerta-Espino (1992) found virulence for Lr30 in 185 out of a total of 1477 isolates, with virulence occurring in the Americas, Africa, China, the Indian subcontinent, Israel, Turkey and Europe. Of the 185 isolates with virulence for Lr30, 176 also possessed virulence for the genes Lr3a and Lr3ka. This association may be due partly to the tight genetic linkage reported for the pathogenicity loci corresponding with host genes Lr3ka and Lr30 (Samborski & Dyck, 1976; Kolmer, 1992). It must be noted, however, that many pathotypes that possess virulence for one or more of the Lr3 alleles are avirulent for Lr30 (Huerta-Espino, 1992; McIntosh et al., 1995).

Although gene Lr13 was described as conferring adult plant resistance (Dyck et al., 1966), it can at times be detected in seedling tests (Pretorius et al., 1984). Caution is needed if seedling tests alone are used (Park et al., 1995). It was possible to confirm avirulence for Lr13 in pt 3 but, in the absence of adult plant tests, the pathogenicity of all other isolates for this gene must be considered as uncertain. Pt 3 was a distinctive pathotype commonly found in the south of France but not elsewhere. An isolate conforming to the same pathotype with the same pathogenicity was identified from Morocco, suggesting that country as a possible origin. Circumstantial evidence was cited that supports exchange of inoculum between North Africa and Europe for wheat stem (Puccinia graminis f.sp. tritici) and leaf rusts (Zadoks, 1965).

Previous studies of pathogenicity of Prt in Europe have reported little or no virulence for the genes Lr9, Lr19, Lr24, Lr25 and Lr29 (Bartošet al., 1992; Casulli & Pasquini, 1992; Manninger, 1992; von Kröcher et al., 1992), consistent with the results of the present study. Given that there are problems with an association of yellow flour pigmentation with Lr19 (Knott, 1989), and that there is evidence of reduced agronomic performance in lines with Lr25 (McIntosh et al., 1995), genes Lr9, Lr24 and Lr29 are probably of greatest potential use to wheat breeders in Europe. In future surveys, determination of the pathogenicity of isolates with respect to adult plant genes (Lr12, Lr13, Lr22a, Lr37) would be valuable in determining their potential in resistance breeding.

Acknowledgements

This work was conducted while the first author was the recipient of an Alexander von Humboldt Research Fellowship, which is gratefully acknowledged. The work was also supported by the Grains Research and Development Corporation, Australia. The authors would also like to thank Professor G. Wenzel and Professor F. J. Zeller for support and Ms Ildikó Bellovics for valuable technical assistance during the course of the study. Isolates of Prt from other regions were kindly supplied by Dr B. Ezzahiri (Morocco), Dr M. Jalli (Finland), Dr K. Manninger (Hungary), Dr E. Paderina (Siberia), Dr H. Peusha (Estonia) and Dr A. Strzembicka (Poland).