Efficacy of Blackleg Major Resistance Genes in B. napus in Germany

Leptosphaeria maculans is one of the major pathogens of oilseed rape (B. napus). It causes blackleg disease, which accounts for significant yield losses worldwide. Using cultivars that harbor major resistance (R) genes is one of the most effective control methods. However, the efficacy of major R genes is related to the frequency of the corresponding avirulence (Avr) genes in a L. maculans population. In this paper, we report the Avr profiles of L. maculans populations and the ratio of its mating types in Northern and Central regions of Germany. Eleven Avr genes in five-hundred and seventy-four isolates were characterized either by applying cotyledon tests on a B. napus differential set or by amplifying avirulence gene-specific PCR markers. Fifty-two races were determined, among which the most dominant race was Avrlm6, -7, -11, AvrlepR1, -R2. Results showed that the resistance gene Rlm2 is 100% ineffective, some other major R genes such as Rlm1, Rlm3, Rlm4 and LepR3 are partially effective (with corresponding Avr frequencies ≤ 42%), while LepR1, LepR2, Rlm6, Rlm11 and Rlm7 can still provide relatively effective resistance in the German fields investigated (with corresponding Avr frequencies of 63–100%). Sexual reproduction is a factor that enhances the potential of L. maculans to evolve under selection pressure. Mating types of the L. maculans populations did not deviate from the ratio of 1:1 in the examined regions, indicating that sexual reproduction and ascospores play central roles in the L. maculans lifecycle. Overall, this study provides an important dataset for the establishment of a strategic plan to preserve the efficacies of major R genes in Germany by applying cultivar rotations of oilseed rape.


Introduction
Blackleg disease (phoma stem canker) is an economically important disease in many oilseed rape-growing areas in the world [1]. The causal agent of the disease is a two-species complex: Leptosphaeria maculans (anamorph = Plenodomus lingam) and Leptosphaeria biglobosa (anamorph = Plenodomus biglobosus) [2,3]. Both species coexist in the field. However, the ratio of L. maculans to L. biglobosa in a region is decisive for disease severity, so that L. maculans is mainly responsible for significant yield losses [4]. L. maculans was reported to be dominant in Germany and other western European countries [1].
Having a complex lifecycle with two reproduction systems and different dispersal mechanisms, L. maculans has a high evolutionary potential that allows it to readily adapt to new conditions, such as introducing new R genes in its host plant [5]. Generally, the fungus survives saprophytically on stubbles of a previous season by means of the sexual fruiting bodies, pseudothecia. Once mature, pseudothecia eject wind-borne ascospores, resulting in a primary infection of host plants. Ascospores germinate on cotyledons and young leaves. Hyphae grow biotrophically to form characteristic phoma lesions, mostly with asexual fruiting bodies, pycnidia. The latter contain pycnidiospores, which can be spread by rain splashes. Their spread may finally result in a secondary infection. The fungus grows Rlm13, LepR1, LepR2, LepR3 (BLMR1.1) and LepR4. Only three of them, Rlm2, Rlm9 and LepR3, were cloned [24,[34][35][36][37][38][39][40].
Sowing a specific commercial cultivar harboring a major R gene in a region over years results in high natural selection pressure. Hence, new L. maculans races evolve that can overcome the introduced major R gene. This is called a "boom and bust" cycle [5]. The amplitude of a "boom and bust" cycle of a major R gene differs among fungal phytopathogens. For L. maculans, several studies have documented the potential lifespan of a major R gene when intensively deployed under experimental conditions or at a commercial level. For example, Brun et al. reported that Rlm6 turned ineffective after three growing seasons in field experiments [41]. On a commercial level, breakdowns of major R genes have been observed within three to five years after their introduction into the market in many oilseed rape-growing countries, such as the efficacy loss of Rlm1 in France, Rlm3 in western Canada and "sylvestris"-derived resistance, namely Rlm1 and LepR3, in Australia [42][43][44].
Setting a strategic regional plan to rotate cultivars harboring major R genes is essential to expand the efficacy longevity of major R genes. The reason is that using qualitative resistance is only reasonable as long as the corresponding Avr gene is dominant in the population. However, a reliable plan for major R gene rotation requires regular updating of the Avr profile of regional L. maculans populations [45]. Monitoring of Avr gene frequencies is thus crucial for practical recommendations for farmers and breeders.
The last study that investigated Avr gene frequencies in L. maculans populations in Germany was based on samples collected in the growing seasons of 2011 and 2012 [46]. Here, we aimed not only to update the Avr profile of L. maculans populations, but also to expand the range of the tested Avr genes by including Avrlm6, Avrlm11, AvrlepR1, AvrlepR2 and AvrlepR3, which have not been investigated in Germany so far. In addition, special attention was given to the change in Avrlm7 frequency, since Rlm7 has been known as the most effective commercialized major R gene in Germany in recent years. Additionally, this study aimed to check whether L. maculans population mating types deviate from the hypothesized 1:1 ratio in northern and central Germany.

Results
Six hundred isolates were collected during three seasons from 2017 to 2020 from seven regions in four provinces in northern and central Germany ( Figure 1). Using the primers ITS5/ITS4 targeting the internal transcribed spacer (ITS) regions, 26 isolates were identified as L. biglobosa while 574 isolates were assigned to L. maculans. In the pathogenicity test, the 26 L. biglobosa isolates caused only very small necrotic leaf spots with no pycnidia on the cotyledons. Additionally, in the susceptible check, cultivar Westar did not show more severe infection when inoculated with the L. biglobosa isolates. The 574 L. maculans isolates were further characterized to determine their races and their mating types.

Efficacy of Major Resistance Genes in German Fields
The 574 isolates assigned to L. maculans were used to monitor the Avr profile of the pathogen population and thus the R gene efficacies. Eleven avirulence genes were characterized (Avrlm1, Avrlm2, Avrlm3, Avrlm4, Avrlm6, Avrlm7, Avrlm9, Avrlm11, AvrlepR1, AvrlepR2 and AvrlepR3), either phenotypically by performing cotyledon tests on an oilseed rape differential set or genotypically by using Avr gene-specific PCR primers.
Results of the pathogenicity test provide evidence that all the tested isolates were virulent on differential lines that harbored Rlm2 and Rlm9 (Figure 2). This shows that Rlm2 is ineffective in the explored regions. However, the 100% virulence frequencies on the Rlm9 differentials were expected since the trap variety NK Bravour harbors Rlm9. In contrast, Figure 2 showed that AvrlepR1 proved to be the most abundant avirulence gene with 100% presence in all investigated regions, except in Peine, where two isolates virulent on the differential line Topas-LepR1 were detected, representing 2% of Peine's population and 0.3% of the whole isolate collection.  Virulence frequencies of L. maculans isolates originating from different fields tested on a B. napus differential set. Cotyledon tests were conducted with 574 isolates. Field sites and numbers of tested isolates: Hadmersleben, n = 88; Groß Helle, n = 83; Nienstädt, n = 97; Einbeck, n = 67; Sorüp = 60; Asendorf = 90, Peine, n = 89. Isolates were collected from 2017 to 2020.
There was a drastic increase in the frequency of avrlm7 isolates (i.e., isolates virulent on Rlm7) demonstrated in this study when compared with the results of our previous work on Rlm7 in Germany (2011-2012) [46]. A comparative analysis of both studies showed that the avrlm7 isolate's frequency increased within 5-7 years from 0.9% to 17.6% on average in all investigated regions. This study confirmed that Rlm7 has become less effective in Germany since 2011-2012, only a few years after its commercial release. According to Figure 2, the percentage of avrlm7 isolates ranged from 7% in Einbeck to 28% in Peine.
Frequencies of compatible interactions on Rlm1, Rlm3, Rlm4 and LepR3 varied from region to region. Phenotypical data revealed that the functional Avrlm1 ranged from 2% to 19%, Avrlm3 from 3% to 26%, Avrlm4 from 3% to 13% and LepR3 from 16% to 42%. These results indicate that the referred R genes are still partially effective. Interestingly, although the alternating functionality supposed due to the masking effect between Avrlm7 and Avrlm3 can be recognized in Figure 2, 17 avrlm7 isolates out of 101 were virulent on the differential Topas-Rlm3, despite the absence of the functional Avrlm7.
LepR2 is also partially effective. However, with a frequency of functional AvrlepR2 ranging from 63% to 91%, LepR2 can be considered similarly effective as Rlm7 in some regions, and thus, the two can be used in rotation with each other to avoid or delay the further resistance breakdown of both. The results in Figure 2 emphasize the importance of considering the regional differences for L. maculans management strategies. For example, Figure 2 shows that, while LepR2 is more effective than Rlm7 in Peine, the opposite is true in Einbeck and Nienstädt.
Since we did not have access to B. napus lines harboring the major R genes Rlm6 and Rlm11, we used specific PCR primers to test for Avrlm6 and Avrlm11. To our knowledge, there are no reports from other research groups who have differential lines including Rlm6 and Rlm11 about masking effects of Avr genes that hinder the recognition of Avrlm6 or Avrlm11. Additionally, it was reported that deletion is the major mechanism of gaining virulence in these Avr genes [33,47]. Therefore, we assumed that the results of the PCR avirulence gene tests would likely match the results of the phenotyping by cotyledon tests. Figure 3 shows that the frequency of Avrlm6 ranged from 88% to 100%, while the frequency of Avrlm11 was 72% to 95%. Consequently, it can be concluded that Rlm6 and Rlm11 are relatively effective, and their efficacies are comparable to that of Rlm7. However, it is noteworthy that there are no reports stating the introduction of these two genes into commercial cultivars in Germany.

L. maculans Races in German Populations
In general, 52 races were described among the 574 L. maculans isolates collected from 7 sites in Germany. However, richness in races according to the Margalef index differed between regions. Table 1 illustrates that Nienstädt showed the highest race diversity with 25 races and a Margalef index of 5.25, whereas Sörup displayed the lowest race diversity with 14 races and a Margalef index of 3.18. The races presented in Table 2 are based on the phenotyping tests and the PCR tests of Avrlm6 and Avrlm11. Since LepR3 is assumed to interact with Avrlm1, it was not possible to ensure the presence of AvrlepR3 distinctly from Avrlm1 in isolates avirulent on both LepR3 and Rlm1. Therefore, these isolates were marked with asterisks to draw attention to the possible redundance. Some researchers hypothesize that Avrlm1 and AvrlepR3 are identical, and that AvrlepR3 is a hypothetical gene [48]. However, there were races in our tested collection that were virulent on Rlm1 and avirulent on LepR3. This indicates that this assumption may not be true. Table 2. Race spectrum of L. maculans populations from seven field sites in Germany collected from 2017 to 2020. Races are described based on phenotypic characterization of Avrlm1, Avrlm2, Avrlm3, Avrlm4, Avrlm7, AvrlepR1, AvrlepR2 and AvrlepR3. Avrlm6 and Avrlm11 were characterized based on specific PCR primers.

Percentage of Total Number of Isolates Collected Per Region
Asendorf Einbeck
The predominant race in all regions investigated was Avrlm 6, 7, 11, AvrLepR 1, 2, which represented 30.8% of the whole races in the tested German L. maculans population. The second most dominant race was Avrlm6, 7, 11, AvrlepR1, 2, 3, making up 11.8% of the population, and differing from the most dominant one by the presence of AvrlepR3. The third dominant race was Avrlm6, 7, 11, AvrlepR1, with a share of 11.5%, and the fourth ranking was Avrlm3, 6, 11, AvrlepR1, 2, at a rate of 7.1%. The four most dominant races together accounted for 61.3% of all races in the tested German population. An effective oilseed rape cultivar rotation should basically consider these most dominant races.
Isolates of the L. maculans population in the explored regions had between two and eight functional avirulence genes ( Figure 4). Among them, 43% had five different Avr genes, 22% had four Avr genes and 20% had six Avr genes. Only a few isolates (1%) had two or eight Avr genes.

L. maculans Mating Type Ratio in German Fields
Mating types were defined using multiplex PCR in 562 isolates ( Figure 5). Both idiomorphs of the pathogen existed in each region and Fisher's exact test proved no significant departure from the 1:1 ratio of the two mating types (Table 3). This emphasizes the importance of the annual sexual reproduction in the lifecycle of L. maculans in the explored regions. In fact, primary infection in Germany depends on the spread of ascospores, a factor that enhances the ability of the fungus to adapt rapidly to new qualitative major R genes implemented in newly commercialized cultivars. . Agarose gel electrophoresis of the PCR-amplified mating type loci of L. maculans using a multiplex PCR system described by Cozijnsen and Howlett [11]. Three primers were used. The use of a common primer in combination with the MAT1-1 locus-specific primer results in an amplicon of 686 bp, while the use of it in combination with the MAT1-2 locus-specific primer affords an amplicon of 443 bp. Table 3. Proportion of mating types of L. maculans in an isolate collection from different regions in central and northern Germany. Multiplex PCR was used to define the idiomorphs of the pathogen. Fisher's exact test showed no significant deviation from the 1:1 ratio of the mating types (p = 0.05).

Site
No

Discussion
This is the first study identifying the frequencies of AvrlepR1, AvrlepR2, AvrlepR3, Avrlm6 and Avrlm11 in the L. maculans population in German fields. Besides, it has updated the Avr profiles of Avrlm1, Avrlm2, Avrlm3, Avrlm4 and Avrlm7, which have not been investigated in Germany since 2012 [46]. Cotyledon tests showed that 100% of the isolates were virulent on differential lines harboring Rlm2. The absence of Avrlm2 was also described in northern Germany in isolates sampled in 2011-2012 [46]. Similar results have previously been found in France in 2000-2001 [49]. However, this situation is different on other continents. For example, in western Canada, a survey on samples collected in 2012-2014 showed that Avrlm2 reached 80% [50]. All our sampled isolates were virulent on the Rlm9 differential. This can be explained by the use of the Brassica napus trap cultivar NK-Bravour that carries the Rlm9 R gene. This led to a preselection of avrlm9-harboring isolates, whereas Avrlm9-harboring isolates were counter-selected.
Epistatic effects, as a mechanism for evading recognition, were reported by Avrlm7 toward Avrlm3 [26] and Avrlm9 [27]. Indeed, the results of our study supported that the presence of the functional Avrlm7 masked the recognition of Avrlm3. Investigations of epistatic mechanisms of Avrlm7 toward Avrlm3 and Avrlm9 revealed that this suppression was caused by neither stopping the expression of Avrlm3 and Avrlm9, nor a physical interaction of the Avr effector proteins of Avrlm3, Avrlm9 and Avrlm4-7 [27,51]. Using protein structure approaches, it was demonstrated that the three effector proteins of these genes belong to a new family of effectors, called Leptosphaeria AviRulence-Supressing effectors (LARS). LARS are structurally analogue effectors that differ in their amino acid identities, although they share common targets in the plants [52].
In our study, several compatible interactions on the Rlm3 differential lines were detected among isolates lacking the functional Avrlm7. This indicates that although Avrlm7 masks the recognition of Avrlm3, its absence does not mean that the pathogen has no other means to avoid recognition. A variety of virulence-gain mechanisms in plant pathogens have been described, such as deletion of the Avr gene, point mutations that allow the pathogen to avoid recognition despite the presence of the Avr gene, amino acid substitution and masking the Avr protein through another Avr protein [26,47,53]. Gene silencing of Avrlm3 proved that this Avr gene is crucial in L. maculans pathogenicity and has an important effect on its lifecycle in B. napus [51]. Thus, deletion of the gene can be ruled out by elucidating the virulence of avrlm7 isolates on Rlm3. Plissonneau et al. explained the virulence of avrlm7 isolates toward Rlm3 by the high allelic polymorphism of Avrlm3, which allows a high level of possible protein isoforms. Therefore, it can be speculated that the alternative mechanism used by the pathogen when the epistatic effect disappears is to substitute an amino acid in the effector protein to allow a so-called "camouflage" type of escaping recognition [51]. Setting strategies for R gene management based on restoring the efficacy of Rlm9 and Rlm3 by losing the functional Avrlm7 could mislead farmers into reducing caution in Rlm7 deployment.
Our results showed that major R genes Rlm1, Rlm3, Rlm4 and LepR3 are not able to provide sufficient resistance against phoma stem canker in German fields. In the explored regions, Avrlm4 isolate frequencies ranged from 3% to 13%. This is quite close to the rate described in a large-scale survey throughout France in 2000-2001 (0-19.5%) [49]. Avrlm4harboring isolates turn virulent when glycine at position 120 in the Avr protein is substituted with arginin [54], while Avrlm1 and AvrlepR3, which are suspected to be alleles of the gene Avrlm1-L3, turn virulent mainly due to deletion of the whole gene [55,56]. Our results showed that Avrlm1 frequency ranged from 2% to 19%. A rapid adaptation of the L. maculans population toward Rlm1 was noticed in France, where the rate of Avrlm1-harboring isolates made up 83% of the population in 1997-1998 and decreased dramatically to less than 13% in 1999-2000 [42]. Similarly, in Canada, Avrlm1 frequency did not exceed 5% [45]. In Australia, the efficacy of Rlm1 resistance in cultivar Surpass 400 notably decreased within three years after commercial release [23], as well as the efficacy of LepR3. This is explained by the fact that Avrlm1 has dual specificity and can trigger both R genes: Rlm1 and LepR3 [24]. It can be concluded that wherever Rlm1 efficacy is broken, LepR3's lifespan is shorter than that of other R genes, in which the corresponding Avr gene does not have dual specificity.
We identified a dramatic increase in the frequency of avrlm7 isolates compared with the 2011-2012 season. The frequency of avrlm7 isolates increased within five to seven years from 0.9% to 17.6% in fields located in central and northern Germany, and the highest presence of them was in Peine, where the frequency reached 28%. Winter and Koopmann stated that Rlm7 was the only still effective major R gene used in commercial cultivars in Germany [46]. The potential speed of Rlm7 breakdown was experimentally tested by applying maximum pressure on the L. maculans population through sowing cultivars with Rlm7 for four years, without applying crop rotation or ploughing in the crop debris. Under these experimental conditions, the results revealed that avrlm7 isolates frequency increased from 0 to 36% within 4 years [54]. In 2006, a study at the European level showed that the Avrlm7 frequency was 100% in France, Germany, Sweden and Poland [57]. However, the intense deployment of Rlm7 resulted in a rapid reduction of Avrlm7 not only in Germany, but also in France. There, Rlm7 was introduced commercially in 2004. By 2013, the frequency of Avrlm7 isolates reached an average of 19.5% and a maximum of 45%, depending on the region [58]. In contrast, Rlm7 is still effective in Canada, with an Avrlm7 frequency exceeding 98% in 2018-2019 [45]. Hence, the present study calls for a more cautious deployment of Rlm7, especially in Europe.
Some major R genes in oilseed rape were derived from related Brassica species. For instance, while Rlm11, LepR1, LepR2 and LepR3 were introduced into B. napus from B. rapa [33][34][35], Rlm6 originated from B. juncea [59]. Our survey provided strong evidence that LepR1 is the most effective major R gene in Germany, with 100% frequency of the isolates harboring AvrlepR1 in all regions investigated, except in Peine, where 2% of the isolates were virulent on LepR1. Although the proportion of the virulent isolates toward LepR1 was very low, and such isolates were found only in one region, the former observation of the rapid decline in efficacy of Rlm7 within 5-7 years should prompt to take measures to preserve the efficacy of LepR1.
In France, AvrlepR1, AvrlepR2, Avrlm6 and Avrlm11 frequencies in the L. maculans population were shown to still be high enough to control the disease in the field [60]. This is in line with our findings in Germany. A significant constraint in the targeted use of R genes in the field is a lack of knowledge about their presence in several commercial cultivars registered in Germany. LepR2, which is believed to be the same gene or an allelic form of RlmS, was reported in cultivars recently registered in France and Germany [61]. Considering the frequency of avrlepR2 in Germany of 9-37%, as shown in our study, an annual monitoring seems necessary in regions where cultivars harboring LepR2 are grown. This may help to avoid a situation such as that in Canada, where AvrlepR1, AvrlepR2 and AvrlepR3 occur already at low frequencies of ca. 19%, 5% and 5-28%, respectively [45].
The major mechanism of virulence gain of Avrlm6 is a deletion of the gene [47]. Van de Wouw et al. studied isolates collected in Australia over two decades (from 1987 to 2017) and observed fluctuating frequencies of Avrlm6 from 21% to 80% depending on the year. This behavior appeared to be independent from the commercial introduction of Rlm6 in Australia in 2010 [62]. Hence, the study speculated that Avrlm6 frequency might also be influenced by the intensive deployment of other major R genes in Australia [62]. In France, Rlm6 was not introduced commercially after the French National Institute for Agricultural Research (INRA) decided to keep it as a research tool [41]. As a result, Avrlm6 frequency was 100% in France [49], similar to Canada, where Avrlm6 frequency exceeds 98% [45]. In our study, a fluctuation in frequency between 88% and 100% depending on the region was observed, however the use of Rlm6 in Germany is not clear.
Avrlm11 has been reported to occur in high frequencies in many regions of the world. For instance, its frequency was >95% in France and 100% in Canada [33,45]. In our study, the rate of isolates harboring Avrlm11 ranged from 66% to 95%. Despite the high frequency of Avrlm11, it was found to be located on a dispensable mini-chromosome of L. maculans that was occasionally lost by meiosis. Thus, the loss of Avrlm11 is related to the loss of this mini-chromosome in L. maculans. [33]. In spite of its dispensability, this chromosome turned out to be influential for the viability of ascospores, and thus, its loss has fitness costs [33].
The race structure in our study was described based on the phenotypic analysis of eight Avr genes and the PCR assays for Avrlm6 and Avrlm11. Assuming each of the 10 tested Avr genes has a minimum of 2 alleles, the theoretical number of combinations would be 1024 (i.e., 2 10 ). This shows the theoretical potential of the pathogen to break resistance. In our study, we were able to determine 52 races, the majority of which had a complexity of five Avr genes. Race complexity depicts the range of the available effective major R genes for setting management strategies for resistant cultivar rotation in a region [45]. In general, we found, according to our isolate collection, that the major R genes that provide sufficient resistance in German oilseed rape fields are LepR1, LepR2, Rlm6, Rlm11 and Rlm7, while 61.3% of the isolates can equally infect cultivars harboring Rlm1, Rlm2 and Rlm4. This is consistent with our previous investigation in Germany [46]. Combining quantitative resistance with race-specific resistance can be important to expand the durability of major R genes. It was demonstrated that Rlm6 in a cultivar with a quantitative resistance background preserved its efficacy two years longer than in a cultivar with a susceptible background [41].
Pathogens with a high evolutionary potential are expected to overcome genetic resistance more rapidly [5]. It is believed that there are two important factors that enhance the evolutionary potential of L. maculans: first, the mixed reproduction systems, and second, the dispersal mechanisms. The sexual reproduction is a milestone for producing variations in the population, while the wind dispersal of spores spread the evolving races effectively over several kilometers [54,63]. However, in seasons when the environmental conditions do not favor sexual reproduction, asexual reproduction would be dominant, and thus, the ratio of idiomorphs might change. For example, in some Canadian regions, samples collected in 2010 and 2015 showed that the mating type MAT1-2 was significantly more dominant than MAT1-1 [64]. In our study, we showed that the idiomorphs of L. maculans in Germany did not significantly deviate from the ratio 1:1. This is consistent with the ratio observed in Overall, we comprehensively compared the effectiveness of major R genes in different countries. We found similarities in the R gene efficacies throughout Germany's neighboring countries, but increasingly significant differences the further those countries were from our study region. This may be related to the fact that the pathogen mainly spreads through wind-borne ascospores that can fly up to several kilometers and stay viable for six weeks [1]. We therefore recommend the consideration of geographical factors for the successful management of blackleg disease.

Sampling and Isolation of Leptosphaeria maculans
To explore the race spectrum of Leptosphaeria maculans in Germany, samples were collected from plots sown with the winter oilseed rape cultivar NK-Bravour (Syngenta Seeds GmbH, Bad Salzuflen, Germany). Plants grown in these plots were used as L. maculans trap plants, since this cultivar only harbors the major R gene Rlm9 (H. Uphoff and M. Gundemann, Syngenta Seeds, pers. comm) [66]. The field trials were established in cooperation with breeding companies in seven regions of four provinces in Germany. Four fields were located in Lower Saxony in Einbeck (KWS), Nienstädt (Bayer CropScience), Peine (Limagrain) and Asendorf (DSV), one field in Saxony Anhalt in Hadmersleben (Syngenta), one field in Mecklenburg Western Pomerania in Groß Helle (NPZ) and one field in Schleswig Holstein in Sörup (BASF). Samples were collected in seasons 2017/2018, 2018/2019 and 2019/2020 (Table 4). Leaf samples with characteristic phoma lesions were collected in autumn at growth stage BBCH 18 and in spring at growth stage BBCH 30 [67]. Samples were dried and stored at 4 • C until they were used for isolation. Isolation was conducted by incubating a dried leaf segment with a lesion in a humid chamber at room temperature to induce conidiation. Single pycnidium isolates were prepared by plating spores on synthetic Nutrient-Poor Agar (SNA) medium amended with 200 ppm streptomycin under sterile conditions. Petri dishes were incubated under NUV light at 20 • C for six days. Afterwards, a mycelial plug was transferred to a V8-juice medium supplemented with 200 ppm streptomycin and incubated for 10-14 days under the same conditions. Subsequently, spore suspensions were prepared and adjusted to a density of 1 × 10 7 spores/mL using a hemocytometer. Spore suspensions were stored at −20 • C and thawed just before they were used for inoculation.
To characterize L. maculans isolates, cotyledon tests were applied using a differential set of B. napus genotypes harboring the major R genes Rlm1, Rlm2, Rlm3, Rlm4, Rlm7, Rlm9, LepR1, LepR2 and LepR3. The cultivars Westar and Topas DH16516 had no major R genes and were used as susceptible controls (Table 5).
Seeds were sown in trays filled with potting soil and covered with transparent plastic plates to keep a high level of humidity. After three days, when seeds germinated, the covers were taken off. On the sixth day, seedlings were transplanted in multi-pot propagation trays with a 3:3:1 mixture of garden soil, compost and sand. On day seven, seedlings were inoculated with 10 µL of spore suspension placed on each lobe of the cotyledons after injuring it with a needle. Seedlings were then put under controlled conditions of 16:8 h light (day/night) at 20 • C. For each isolate and differential line, eight seedlings were inoculated. Symptoms were evaluated 14 days post-inoculation according to the IMASCORE rating scale, where class one shows typical hypersensitive reactions and class six reflects tissue collapse with sporulation. Classes one to three are considered as incompatible reactions while classes four to six are noted as compatible ones [72].

DNA Extraction and PCR Assays
DNA templates were prepared using a simple boiling DNA extraction method. Shortly, 100 µL of spore suspension (10 7 spores/mL) was centrifuged at 16,000× g for 10 min. Afterwards, the supernatant was discarded carefully to keep the pellet. Fifty microliters of Tris (10 mM) was added to the pellet and homogenized by vortexing. The suspension was exposed to an ultrasound for 5 s and then transferred to a water bath at 98 • C for 10 min. Subsequently, the tubes were put on ice for 10 min. Before applying the PCR tests, the quality and quantity of the extracts were checked using agarose gel electrophoresis (0.8% (w/v)) supplied with Midori Green (NIPPON Genetics Europe GmbH). Five µL of DNA was mixed with two µL of the loading dye (100 mM EDTA, 50% (v/v) glycerol and 0.025% (w/v) bromophenol-blue) and run in TBE buffer at 3 V/cm for 60 min. DNA quantity was afterwards compared with a standard Lambda phage DNA of concentrations ranging from 150 to 35.5 ng. The concept of releasing DNA by boiling the cells was also used by Adwan [73].
To distinguish L. maculans and L. biglobosa, we amplified the internal transcribed spacer (ITS) regions of the pathogen using ITS4 and ITS5 primers ( Table 6). The steps of the 35 PCR thermal cycles are described in Table 7. Avrlm6 and Avrlm11 were amplified using the primer pairs listed in Table 6 in 40 and 30 thermal cycles, respectively. PCR profiles are shown in Table 7. In general, the end volume of a PCR reaction was 25 µL, of which 7 µL was DNA extract. The mixture contained 1 µM of each primer pair, except for the mating type primers, where 0.67 µM of each of the three primers were used. The concentration of the nucleoside triphosphates (dNTPs) was 0.2 mM, and 1 unit of the FastGene Taq DNA polymerase (NIPPON Genetics Europe GmbH) was added. The reaction was conducted in a buffer containing 1.5 mM of MgCl 2 . PCR reactions were conducted in a T Professional Basic Gradient thermal cycler (Biometra, Göttingen, Germany).
To visualize the PCR products, 5 µL was mixed with 2 µL of the loading dye described above. The mixture was loaded on 1% agarose gel supplied with Midori Green. Electrophoresis was run at 3 V/cm for 60 min for all PCR products, except for the mating type tests, where the electrophoresis was run for at least two hours. Mating types of L. maculans were defined using a multiplex PCR, as described by Cozijnsen and Howlett [11]. An isolate of L. maculans has a single mating locus (MAT locus) and one of the two loci MAT1-1 or MAT1-2. The common primer for the MAT locus in combination with the MAT1-1-specific primer resulted in an amplicon of 686 bp, while the common primer in combination with the MAT1-2-specific primer afforded an amplicon of 443 bp.

Statistical Analysis
For phenotypic characterization of the L. maculans population, the numbers of compatible (susceptible) reactions observed in the cotyledon tests were represented relative to the number of isolates collected per region. Similarly, Avrlm6 and Avrlm11 that produced bands in PCR tests were represented as the proportion of the total number of isolates analyzed per region. The Margalef index (D Mg ) was calculated using the following formula: where S is the number of races per region, and N is the number of L. maculans isolates in the corresponding region.
Fisher's exact test χ 2 was applied using the software R [76] to test the null hypothesis that the mating types of L. maculans in German fields do not deviate from the ratio 1:1 at a confidence level of 95%.

Conclusions
In this study, the L. maculans population in German oilseed rape fields was characterized by their races. The resulting Avr profile of the pathogen provides an important basis for evaluating the state of efficacy of major R genes used in commercial cultivars. Such knowledge is a prerequisite for establishing a management strategy with the aim of protecting R genes from a premature decline in efficacy and for expanding their durability in practical use. The rapid breakdown of major R genes identified in this study strongly implies the need for combining resistant cultivars with other control measures. Considering the long breeding cycles of 7-10 years required to have a new major R gene established in a registered new cultivar, it is important to combine quantitative resistance with major R genes in order to achieve more durable and sustainable disease control. The similarities observed in Avr profiles of L. maculans between France and Germany suggest that such management strategies can be valuable on a broader geographical scale. In addition, different Avr profiles existing in European and Canadian oilseed rape fields imply that transfer of races between continents by seed trade must be avoided. Hence, we emphasized the importance of considering L. maculans in the phytosanitary measures, ensuring international clean seed pathways.