Selection for increased spore efficacy by host genetic background in a wheat powdery mildew population

ABSTRACT A field experiment was designed to test the hypothesis of for increased reproductive ability on different host genetic backgrounds within a wheat powdery mildew population. Studies have suggested that, in host mixtures, such selection could increase the reproduction rate of simple patho-types that always develop on the same host genetic background, whereas complex pathotypes should not be affected because they infect different host genotypes. In our experiment, the Erysiphe graminis population reproduced for successive generations on cvs. Orkis and Etecho, either grown as pure stands or in a mixture. In an additional treatment, the host cultivar changed after each generation. Isolates were sampled in April and, after seven pathogen generations, in July. At the second sampling date and for pure stands only, mean spore efficacy was greater on the host from which isolates were sampled than on the other one. This was attributed to selection within the pathogen population for better spore efficacy on the host genetic background. This selection was independent of the virulence genes carried by the isolates. The possibility of a phenotypic plasticity effect was tested and rejected.     

Competition and interactions among stripe rust pathotypes in wheat-cultivar mixtures

A series of experiments was conducted with wheat stripe rust to analyse competition between simple and complex pathotypes in host mixtures. Two different pathotype combinations were tested, with different host components. Each combination included a complex (able to infect two host components) and two simple pathotypes. For one of the combinations, induced resistance was tested in a separate experiment as a possible interaction among pathotypes. Disease severity and pathotype frequencies were measured three times during the epidemic, on each host component grown in pure stands and in mixtures. In one of the experiments, pathotype frequencies were also measured within secondary foci. One of the complex pathotypes appeared to have a low fitness on one of the host components and did not significantly increase in frequency in host mixtures relative to pure stands. The average frequency of the other complex pathotype increased during the first epidemic cycles, but remained stable afterwards, below expected values. The results suggest that the development of complex pathotypes in host mixtures may be influenced by differential aggressiveness on the host components, by induced resistance and by random effects resulting from the formation of disease foci, and depends on pathogen autoinfection rate and dispersal mechanisms.     

The role of aggressiveness and competition in the selection of Phytophthora infestans populations

Selection within populations of Phytophthora infestans was investigated by comparing the aggressiveness of single‐lesion isolates on detached leaflets of four potato cultivars with differing levels of race‐nonspecific resistance to P. infestans. The isolates included 23 representative of Northern Ireland genotypes from the early 2000s, used to inoculate previously reported field trials on competitive selection (2003–2005), plus 12 isolates recovered from the 2003 trial. The cultivars were those planted in the previous trials: Atlantic (blight‐susceptible) and Santé, Milagro and Stirling (partially resistant). Very highly significant variation for latent period, infection frequency and lesion area was found between genotypes and cultivars; differences between genotypes were more marked on the more resistant cultivars, but no one genotype was the most aggressive across all. Detached leaflets were also inoculated with mixtures of isolates from each genotype group at three sporangial concentrations: differences in aggressiveness between genotypes were more apparent at lower concentrations and on the more resistant cultivars. Genotype groups that were the most aggressive on the more resistant cultivars tended to be those selected by the same cultivars in the field. A mixture of all isolates of all genotypes was used to inoculate detached leaflets of the same cultivars. With one exception, single spore isolates recovered from any one leaflet belonged to a single genotype, but different genotypes were recovered from different cultivars. Phytophthora infestans isolates from Northern Ireland showed significant variation for foliar aggressiveness, and pathogen genotypes exhibited differential aggressiveness to partially resistant cultivars and interacted competitively in genotype selection.     

Variation in aggressiveness is detected among Puccinia triticina isolates of the same pathotype and clonal lineage in the adult plant stage

Puccinia triticina reproduces asexually in France and thus individual genotype is the unit of selection. A strong link has been observed between genotype identities (as assessed by microsatellite markers) and pathotypes (pools of individuals with the same combination of qualitative virulence factors). Here, we tested whether differences in quantitative traits of aggressiveness could be detected within those clonal lineages by comparing isolates of identical pathotype and microsatellite profile. Pairs of isolates belonging to different pathotypes were compared for their latent period, lesion size and spore production capacity on adult plants under greenhouse conditions, with a high number of replicates. Isolates of the same pathotype showed remarkably similar values for the measured traits, except in three situations: differences were obtained within two pathotypes for latent period and within one pathotype for sporulation capacity. One of these differences was tested again and confirmed. This indicates that the average aggressiveness level of a leaf rust pathotype may increase without any change in its virulence factors or microsatellite profile.     

Diversity,spatial variation,and temporal dynamics of virulences in the German leaf rust (<Emphasis Type="Italic">Puccinia recondita</Emphasis> f. sp. <Emphasis Type="Italic">secalis</Emphasis>) population in winter rye

A large collection of German rye leaf rust isolates was analysed to characterize the diversity, spatial variation and temporal dynamics of virulences. Virulence-avirulence phenotypes (=pathotypes) were determined on 23 host differentials. We found 93 pathotypes among 177 single-uredinial isolates in 2000, 201 pathotypes among 437 isolates in 2001, and 125 pathotypes among 213 isolates in 2002. In total, the 827 analyzed isolates represented 317 pathotypes. Frequency of virulences on the individual differentials varied from 2% to 97%. Eight of the differentials showed a high resistance level with virulence frequencies <10%. Virulence complexity of the isolates ranged from 3 to 21 with a mean of nine. The percentages of highly virulent isolates (>14 virulences) increased from 4 to 15% during the sampling period. A high level of virulence diversity was observed within and between individual sampling sites with Simpson indices around 0.9. Evenness indices ranged from 0.88 to 0.92. Four of the five most frequent pathotypes were found in each year but their frequency never exceeded 10%. Isolates with unusual virulence combinations could be clearly separated by principal component analysis. Location-specific pathotype frequencies were revealed in each year, but the frequency patterns varied across years. On four fields a considerable increase of highly virulent pathotypes occurred within 6 weeks during the epidemic. The high diversity of pathotypes as well as the fast accumulation of highly virulent pathotypes favour the adaptation of the pathogen to race-specific host resistances. More durable resistance might be achievable by combining new effective race-specific resistances with adult-plant and/or race-non-specific quantitative resistances.     

Evaluation of cacao (Theobroma cacao) clones against seven Colombian isolates of Moniliophthora roreri from four pathogen genetic groups

Artificial pod inoculation was used to compare the relative aggressiveness of seven Colombian isolates of Moniliophthora roreri (the causal agent of moniliasis or frosty pod disease), representing four major genetic groupings of the pathogen in cacao (cocoa), when applied to five diverse cacao genotypes (ICS-1, ICS-95, TSH-565, SCC-61 and CAP-34) at La Suiza Experimental Farm, Santander Department, Colombia. The following variables were evaluated 9 weeks after inoculation of 2- to 3-month-old pods with spore suspensions (1·2 × 10⁵ spores mL⁻¹): (i) disease incidence (DI); (ii) external severity (ES); and (iii) internal severity (IS). IS was found to be of greatest value in classifying the reaction of the host genotype against M. roreri . Genetic variation reported between isolates and cacao genotypes was not matched by similar diversity in their aggressiveness. All isolates were generally highly aggressive against most cacao genotypes, with only two isolates showing reduced IS and ES reactions. There was considerable variation between clones in the IS and ES scores, but one cultivated clone (ICS-95) displayed a significant level of resistance against all seven isolates. This clone may be useful in cacao breeding initiatives for resistance to moniliasis of cacao.     

The threat of wild habitat to scab resistant apple cultivars

Evaluations of plant resistance to pathogens are rarely made using isolates from wild habitats, although the heterogeneity of such habitats may generate pathogen diversity which could be a source of new virulence in cultivated habitats. The aim of this study was to investigate whether scab resistance factors, identified and characterized in apples using isolates of Venturia inaequalis from a cultivated habitat, remained effective against isolates from a wild habitat. Three V. inaequalis core collections originating from the cultivated apple Malus × domestica and from two wild species, M. sieversii and M. sylvestris, were established to maximize pathogen diversity. For each core collection, 10 isolates were inoculated in mixtures onto 51 genotypes from an apple progeny segregating for two qualitative resistance genes and six quantitative resistance loci (QRL). On each apple genotype, isolates that contributed to the scab symptoms were identified within the mixture using microsatellite markers. The most frequently detected isolates were inoculated singly to compare their aggressiveness according to their host origin. The results showed that isolates from a wild habitat were able to infect the susceptible apple genotypes. However, these isolates were never more aggressive than isolates from the cultivated habitat on the resistance factors tested. It can therefore be concluded that the resistance factors used in this study, identified with V. inaequalis isolates from a cultivated habitat, remained effective against isolates from M. sylvestris and M. sieversii.     

Virulence phenotypes in powdery mildew (Blumeria graminis) populations and resistance genes in triticale (x Triticosecale)

Triticale (xTriticosecale) is a new host for powdery mildew (Blumeria graminis) being not infected by this pathogen in Germany before 2001. To evaluate population structure of the pathogen and race-specific resistances in the host, 694 isolates were collected in 12 states of Germany in the years 2007 to 2009 on triticale. They were tested by a newly developed initial differential set of 20 triticale cultivars. Corresponding virulences were found for all differentials except for cultivar Grenado. In total, 272 different virulence phenotypes (=pathotypes) were detected. The virulence complexity of the isolates ranged from 6 to 19 of 20 possible virulences with a mean of 15. In all years, a high level of diversity of the powdery mildew populations was observed with Simpson indices in the range 0.95 to 0.97. The distribution of the pathotypes was even across Germany with an Evenness index in the range 0.82 to 0.88. A set of 19 isolates with different virulence pathotypes and 10 cultivars were selected to be used to identify race-specific resistances of triticale cultivars and breeding lines. Some cultivars susceptible to most of the isolates in seedling stage were moderately resistant in adult-plant stage. The high diversity and complexity of the pathotypes found in German powdery mildew populations as well as an increasing acreage of only a few dominating triticale cultivars accelerate the adaptation of the pathogen to race-specific host resistances suggesting restricted durability only. More durable resistance might be achieved by combining new effective race-specific (qualitative) resistance genes with race-nonspecific (quantitative) resistances effective in the adult-plant stage that are already available.     

Variation in pathogen aggressiveness within a metapopulation of the Cakile maritima–Alternaria brassicicola host–pathogen association

Variation in aggressiveness and its consequences for disease epidemiology were studied in the Cakile maritima–Alternaria brassicicola host–pathogen association. Variability in pathogen growth rates and spore production in vitro , as well as disease severity and lesion growth rate on C. maritima in glasshouse inoculation trials, were investigated . Substantial variation was found in growth rates among individual A. brassicicola isolates, as well as among pathogen populations. A significant trade-off also existed between growth and spore production, such that faster-growing isolates produced fewer spores per unit area. While there was little evidence for a link between growth in vitro and either disease severity or lesion development among fast- vs slow-growth isolate classes at the individual isolate level, the results suggest that variation in pathogen fitness components associated with aggressiveness may influence disease dynamics in nature. An analysis using an independent data set of disease prevalence in the associated host populations found a significant positive relationship between the average growth rate of pathogen populations in vitro and disease progress over the growing season in wild host populations. Trade-offs such as those demonstrated between growth rate and spore production may contribute to the maintenance of variation in quantitatively based host–pathogen interactions.     

Aggressiveness of eight <Emphasis Type="Italic">Didymella rabiei</Emphasis> isolates from domesticated and wild chickpea native to Turkey and Israel,a case study

Ascochyta blight, caused by Didymella rabiei, affects both domesticated chickpea and its congeneric wild relatives. The aim of this study was to compare the aggressiveness of D. rabiei isolates from wild and domesticated Cicer spp. in Turkey and Israel on wild and domesticated hosts from both countries. A total of eight isolates of D. rabiei sampled from C. pinnatifidum, C. judaicum and C. arietinum in Turkey and Israel was tested on two domesticated chickpea cultivars and two wild Cicer accessions from Turkey and Israel. Using cross-inoculation experiments, we compared pathogen aggressiveness across the different pathogen and host origin combinations. Two measures of aggressiveness were used, incubation period and relative area under the disease progress curve. The eight tested isolates infected all of the host plants, but were more aggressive on their original hosts with one exception; Turkish domesticated isolates were less aggressive on their domesticated host in comparison to the aggressiveness of Israeli domesticated isolates on Turkish domesticated chickpea. C. judaicum plants were highly resistant against all of the isolates from different origins except for their own isolates. Regardless of the country of origin, the wild isolates were highly aggressive on domesticated chickpea while the domesticated isolates were less aggressive on the wild hosts compared with the wild isolates. These results suggest that the aggressiveness pattern of D. rabiei on different hosts could have been shaped by adaptation to the distinct ecological niches of wild vs. domesticated chickpea.     

Diversity and Complexity of Erysiphe graminis f.sp. hordei Collected from Barley Cultivar Mixtures or Barley Plots Treated with a Resistance Elicitor

Powdery mildew populations were analysed to determine the effects of a resistance elicitor and cultivar mixtures on genetic complexity and diversity. Isolations were made from a range of spring barley monocultures and mixtures in a field trial, and characterised for virulence and RAPD profile. In a second trial, isolates were taken from a single mixture from untreated and resistance elicitor-treated areas and from the components of the mixture in monoculture. The mildew population was not only highly heterogeneous for virulence characteristics, but also proved heterogeneous within pathotypes for molecular markers, indicating the major impact of sexual recombination on population structure and the lack of clonal dominance. Various diversity measurements were compared and the value of dissimilarity measurement for revealing genetic distance within a population was highlighted. There was a trend towards increasing complexity as the season progressed, but there was no consistent relationship between cultivar or mixture, disease control treatment, fertiliser treatment, replicate or position in trial, and pathogen genotype. Whilst the resistance elicitor did reduce mildew by 78% in the first trial, and there was no interaction with fertiliser level in its expression, control was substantially less in the second trial. There were no differences between mildew isolates from elicitor and control treatments. It was felt that more effective and consistent resistance elicitors need to be developed before it can be stated that they are unlikely to be eroded by selecting resistant or adapted mildew genotypes.     

Genetic Diversity in South American Colletotrichum gloeosporioides Isolates from Stylosanthes guianensis, a Tropical Forage Legume

The degree of genetic diversity of 127 Colletotrichum gloeosporioides isolates from Stylosanthes guianensis genotypes in South America was measured at the molecular level by random amplified polymorphic DNA (RAPD) with nine arbitrary primers of 10 bases, and by restriction fragment length polymorphism (RFLP) with a non-LTR (long terminal repeats) retrotransposon DNA sequence. The RAPD products revealed scorable polymorphism among the isolates, and a total of 80 band positions were scored. Sixty-three of the 127 isolates were clustered into 13 distinct lineages usually correlating with geographic origin. Where isolates from various regions were clustered together, most had identical host genotype origin. The pathogen population sampled from Carimagua, Colombia, a long-time Stylosanthes breeding and selection site, with a savanna ecosystem, was highly diverse. A set of 12 S. guianensis genotype differentials was used to characterize pathogenic variability of 104 isolates and their virulence patterns were grouped into 57 pathotypes. However, when they were tested on four Australian differentials, they grouped into 11 pathotypes. As shown in previous studies, no strict correlations existed between genetic diversity measured by RAPD or RFLP, and pathotype defined by pathogenicity pattern on the differentials. Southern blot analysis of the 127 isolates revealed 23 hybridizing fragments, resulting in 41 fingerprint patterns among the 127 isolates. Relationships between RFLP and RAPD variables were examined using Spearman's Rank Correlation Coefficient, which showed that the two measures of genotypic variation are in agreement.     

Aggressiveness of Isolates of Phytophthora infestans from the Columbia Basin of Washington and Oregon

ABSTRACT The aggressiveness of 22 isolates of Phytophthora infestans collected from naturally infected potato plants in the Columbia Basin of Washington and Oregon was determined on detached potato leaflets at 18 degrees C in an incubator. Selected isolates were evaluated on whole plants in a greenhouse. Aggressiveness was measured by using the area under the lesion expansion curve (AULEC), incubation period, latent period, sporulation capacity, and lesion size on detached leaflets and the area under the disease progress curve and sporulation capacity on whole plants. The detached-leaflet assay was useful in that a large number of isolates were tested, several components of aggressiveness were studied, and significant differences were found among isolates. Significant variation for components of aggressiveness was found within and among isolates classified according to genotype. Significant interactions among isolates and cultivars were found for some components of aggressiveness, so results were pooled according to cultivar. On average, US-8 and US-11 isolates had higher AULEC scores, indicating aggressiveness higher than that of US-7, US-6, and US-1 genotypes. One US-8 genotype isolate had a higher standardized sporulation capacity than isolates of the other genotypes. US-6 genotype isolates were the least aggressive group, as indicated by low AULEC, sporulation capacity, and lesion size values. The replacement of the US-1 genotype by the US-8 genotype in the Columbia Basin may be partially explained by the increased aggressiveness of US-8 isolates. Additionally, potato growers may need to shorten intervals between fungicide applications and begin applications earlier.     

The gene diversity pattern of Diplocarpon rosae populations is shaped by the age,diversity and fungicide treatment of their host populations

Black spot disease caused by the hemibiotrophic ascomycete Diplocarpon rosae is the most devastating disease of field‐grown roses. Although resistance to black spot is an important trait for rose breeding, little information on the diversity of the pathogen is currently available. To date, a number of single‐spore isolates have been characterized based on a set of test genotypes of the host. In this study, six polymorphic simple sequence repeat (SSR) markers for D. rosae were developed and their potential application to single‐spore isolates collected from a wide range of geographic locations within and outside Europe discussed. Populations of the fungus were then analysed in different German rose collections/populations. Gene diversity was highest in older rose collections managed without fungicide application, and lowest in the two‐ to three‐year‐old testing sites of German rose breeders. Additional analyses of a global collection of single samples and populations revealed no unique alleles from any of the locations, indicating that global trading of rose varieties has led to an admixture of the pathogen.     

Aggressiveness of Mycosphaerella graminicola Isolates from Susceptible and Partially Resistant Wheat Cultivars

ABSTRACT The selective effect of quantitative host resistance on pathogen aggressiveness is poorly understood. Because two previous experiments with a small number of bread wheat cultivars and isolates of Mycosphaerella graminicola had indicated that more susceptible hosts selected for more aggressive isolates, we conducted a larger experiment to test that hypothesis. In each of 2 years, six cultivars differing in their levels of partial resistance were planted in field plots, and isolates were collected from each cultivar early and late in the growing season. The isolates were inoculated as populations bulked by cultivar of origin, field replicate, and collection date on seedlings of the same six cultivars in the greenhouse. The selective impact of a cultivar on aggressiveness was measured as the difference in aggressiveness between early and late isolates from that cultivar. Regression of those differences on disease severity in the field yielded significance values of 0.0531 and 0.0037 for the 2 years, with moderately resistant cultivars selecting for more aggressive isolates. In a related experiment, the protectant fungicide chlorothalonil was applied to plots of two susceptible cultivars to retard epidemic development. When tested in the greenhouse, isolates of M. graminicola from those plots were significantly more aggressive than isolates from the same cultivars unprotected by fungicide.     

Evolution of a pathogen population in host mixtures: simple race–complex race competition

The evolution of the genetic structure of a pathogen population was studied in a varietal mixture with an epidemic simulator based on the model EPIMUL. The pathogen population was composed of simple races able to develop on only one component of the mixture and a complex race which developed on all mixture components. The effects on the simple race–complex race competition of a cost of virulence, of density dependence and of differential adaptation were studied. The selection for simple or complex races in the pathogen population did not depend on initial race frequencies. For a given multiplication rate, complex race frequency increased faster when the spore dispersal gradient was shallow, when distribution of initial disease was generalized, when amount of initial disease was reduced and when the number of mixture components was increased. This was attributed to a better efficacy of the mixture in controlling simple races, resulting in a higher relative fitness of the complex race. For measured values of density dependence or differential adaptation effects, the complex race was at a higher frequency after a mean number of pathogen cycles between 2.5 and 5. The effect of the cost of virulence was stronger and, in certain situations, could result in selection for simple races. In the conditions of our simulations and with the effects tested, stabilization of the pathogen population in host mixtures was unlikely to occur. However, more information is needed concerning the rate at which complex races could evolve and how quickly mixture resistance could be eroded.     

Characterization of Sclerotinia minor populations in Texas peanut fields

In the summer of 2004 an epidemic of sclerotinia blight of peanut, a disease caused by Sclerotinia minor, occurred in Texas in fields where the disease was never previously detected. The disease was observed on many plants within one of the fields (>3000 disease foci), although most foci were <1 m. It is hypothesized that these observations were inconsistent with the recent introduction of a monocyclic pathogen, even if disease developed under conducive environmental conditions. The pattern of disease is most suggestive of the presence of foliar (ascospore) infections, although air temperature was above the known limits for apothecia development if the pathogen had arrived in the field in 2004 peanut seed. To further examine this epidemic, 232 isolates were collected, across a variety of spatial scales spanning this field and other Texas peanut fields, and evaluated for aggressiveness, fungicide sensitivity and genotypic diversity. There was wide variation among isolates for the phenotypic characteristics measured, but there was no evidence that a genotypically unique, highly aggressive, and fungicide resistant isolate had been introduced or evolved. The predominant genotype, TX1, which contained 154 isolates, was found in every county and field population.     

Diagnostics,genetic diversity and pathogenic variation of ascochyta blight of cool season food and feed legumes

Molecular diagnostic techniques have been developed to differentiate the Ascochyta pathogens that infect cool season food and feed legumes, as well as to improve the sensitivity of detecting latent infection in plant tissues. A seed sampling technique was developed to detect a 1% level of infection by Ascochyta rabiei in commercial chickpea seed. The Ascochyta pathogens were shown to be genetically diverse in countries where the pathogen and host have coexisted for a long time. However, where the pathogen was recently introduced, such as A. rabiei to Australia, the level of diversity remained relatively low, even as the pathogen spread to all chickpea-growing areas. Pathogenic variability of A. rabiei and Ascochyta pinodes pathogens in chickpea and field pea respectively, appears to be quantitative, where measures of disease severity were based on aggressiveness (quantitative level of infection) rather than on true qualitative virulence. In contrast, qualitative differences in pathogenicity in lentil and faba bean genotypes indicated the existence of pathotypes of Ascochyta lentis and Ascochyta fabae. Therefore, reports of pathotype discrimination based on quantitative differences in pathogenicity in a set of specific genotypes is questionable for several of the ascochyta-legume pathosystems such as A. rabiei and A. pinodes. This is not surprising since host resistance to these pathogens has been reported to be mainly quantitative, making it difficult for the pathogen to overcome specific resistance genes and form pathotypes. For robust pathogenicity assessment, there needs to be consistency in selection of differential host genotypes, screening conditions and disease evaluation techniques for each of the Ascochyta sp. in legume-growing countries throughout the world. Nevertheless, knowledge of pathotype diversity and aggressiveness within populations is important in the selection of resistant genotypes.     

Distribution of pathotypes with regard to host cultivars in French wheat leaf rust populations

ABSTRACT Isolates of wheat leaf rust collected from durum and bread wheat cultivars in France during 1999-2002 were analyzed for virulence on 18 Thatcher lines with single genes for leaf rust resistance (Lr genes). Sampling focused on the five most widely grown bread wheat cultivars (two susceptible and three resistant) to allow statistical comparison of diversity indexes between the cultivars. Leaf rust populations from durum and bread wheats were different. The diversity of the bread wheat leaf rust pathotypes, as measured by the Shannon index, ranged from 2.43 to 2.76 over the 4 years. Diversity for wheat leaf rust resistance was limited in the host since we postulated only seven seedling resistance genes in the 35 cultivars most widely grown during 1999-2002. Leaf rust populations were strongly differentiated for virulence within bread wheat cultivars, and diversity was higher on those that were resistant, mainly due to a more even distribution of virulence phenotypes than on susceptible cultivars. The pathogen population on the susceptible cv. Soissons was largely dominated by a single pathotype (073100), whereas all other pathotypes virulent on cv. Soissons either decreased in frequency or remained at a low frequency during the period studied. Several pathotypes including the most complex one were found only on resistant cultivars, even though most of them were virulent on the susceptible cv. Soissons. Specific interactions were necessary, but not always sufficient, to account for pathotype distribution and frequencies on the cultivars, suggesting that selection for virulence to host resistance genes is balanced by other selective forces including selection for aggressiveness.     

Structure and pathogenic variability in Ascochyta rabiei populations on chickpea in the Canadian prairies

A population study of Ascochyta rabiei from the Canadian prairies was conducted to assess pathogenicity among isolates with the objectives to investigate the existence of a race or pathotype structure and to evaluate whether there had been a shift to higher aggressiveness between 1998 and 2002. Ninety-nine isolates collected in 1998, 2001 and 2002 were inoculated onto seven differential chickpea genotypes. Significant isolate × differential interactions occurred, but accounted for a small proportion of the total variability. It was found that very few interaction effects between all combinations of differentials and isolates were significant and frequency distributions of disease severity of isolates tested on the differentials revealed continuous distributions. These results suggest that no genotype-specific relationship existed between A. rabiei and its host and that Canadian populations of the pathogen cannot be objectively classified into races or pathotypes. Isolates from 2001 and 2002 caused significantly more disease than isolates from 1998, suggesting that disease epidemics encountered since 1999 were in part caused by a shift in the population to higher aggressiveness.