The clonal dynamic in wild and agricultural plant pathogen populations

-

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

Abstract: The stability or change in clone frequencies during the disease cycle and from year to year is what I term the clonal dynamic. Among pathogenic fungi, the prevalence of efficient asexual reproduction affords the opportunity for invasive, epidemic, clonal colonization and spread. Clonality is probably most extreme in monoculture, although it could be expected to be important in wild plants and in transfers of adaptive pathogenic genotypes between wild and cultivated plants. The clonal dynamic was studied in Sclerotinia sclerotiorurn in two experiments, one on four Canadian field populations of cultivated canola and the other on two Norwegian populations of a wild perennial plant, Rarlurlculus ficaria. Additional samples were made from canola and other crops in Canada and Norway. Four major differences between the agricultural and wild populations were observed. First, in agricultural populations, DNA fingerprint (multilocus haplotype) and mycelial compatibility group were coupled; all individual members of a clone shared a unique fingerprint and all were mycelially compatible. In wild populations, DNA fingerprint and mycelial compatibility group were decoupled. Second, in agricultural populations fingerprint diversity was high, with 594 genotypes recovered from 2747 isolates, but frequently sampled clones were recovered from a wide geographical area repeatedly over a 3-year period; in wild populations fingerprint diversity was low, with 7 genotypes from 300 isolates, and highly localized. Third, in agricultural populations, no evidence of outcrossing and segregation was observed; in the wild populations, some sibling ascospores showed different mycelial compatibility reactions, indicating that crossing had occurred. Last, in agricultural populations, clones were randomly dispersed spatially, probably the result of immigration and mixing of inoculum in air; in the apparently isolated wild populations, strong spatial substructuring was indicated by the distribution of fingerprints, apparently the result of highly localized inbreeding. Clonality was therefore clearly detected in the cultivated plant populations but was difficult to distinguish from inbreeding in the wild populations. Key words: multilocus haplotype, clonality, asexual reproduction, population genetics.

Rksumk : L'auteur nomme dynamique clonale, la stabilitC ou le changement des frtquences des clones au cours du cycle de la maladie, d'une annte a l'autre. Parmi les champignons pathogenes, la prevalence d'une reproduction sexuelle efficace confire une possibilitt de colonisation et de dispersion invasive, CpidCmique et clonale. La formation de clones est probablement plus extr&me en monoculture, bien qu'elle puisse vraisemblablement &tre importante chez les plantes sauvages et dans les transferts de genotypes pathogenes adaptatifs entre les plantes sauvages et les plantes cultivtes. La dynamique clonale a t t t ttudite chez le Sclerotirlia sclerotiorurn dans deux expkriences, une sur quatre populations de canola cultivtes au champ, et l'autre sur deux populations norvkgiennes d'une plante ptrenne sauvage, le Ranunculus ficaria. Des Cchantillonnages suppltmentaires ont Ctt effectuts dans le canola et autres cultures au Canada et en Norvege. Les auteurs ont observe quatre differences majeures entre les populations sauvages et cultivCes. D'abord, dans les populations agricoles, le patron d'ADN (haplotype multilocus) et le groupe de compatibilitC myctlienne sont couples; tous les membres individuels d'un clone partagent un patron unique et tous les mycCliums sont compatibles. Dans les populations sauvages, le patron d'ADN et le groupe de compatibilitC sont dCcouplCs. Deuxiemement, dans les populations agricoles, la diversitt des patrons est grande, avec 594 gtnotypes recouvrts sur 2747 isolats, mais les clones Cchantillonnts frtquemment ont CtC obtenus a partir de vastes regions geographiques tres souvent au cours d'une ptriode de 3 ans; dans les populations sauvages, la diversite des patrons est faible, avec 7 genotypes obtenus partir de 300 isolats, qui sont fortement 1ocalisCs. Troisikmement, dans les populations agricoles, il n'y a pas de preuve de croisement et de stgrtgation; dans les populations sauvages, certaines ascospores soeurs montrent des rtactions de compatibilitC mycelienne differentes, ce qui indique la presence de croisements. Enfin, dans les populations agricoles, les clones sont disperses au hasard dans l'espace, resultant probablement de l'immigration et du mtlange de l'inoculum dans

Received August 18, 1994.

L.M. Kohn. Department of Botany, University of Toronto, Erindale College, Mississauga, ON L5L 1C6, Canada. Can. J. Bot. 73(Suppl. 1): S1231-S1240 (1995). Printed in Canada 1 Imprim6 au Canada

Can. J. Bot. Vol. 73 (Suppl. I ) , 1995 l'air; dans les populations sauvages apparemment isolees, la distribution des patrons indique une forte sous-structure spatiale, resultant apparemment d'autocroisement fortement localis&. On a donc pu deceler clairement la formation de clones dans les populations de plantes cultivtes mais il a CtC difficile de la distinguer de I'autocroisement dans les populations naturelles.

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

Mots clPs : haplotype multilocus, clonalitee, reproduction sexuelle, genetique des populations. [Traduit par la redaction]

Any successful fungal pathogen requires the means to acquire nutrition from a host and the opportunity to proliferate adaptive genotypes if it is to fulfil its evolutionary "motive" of reproductive continuity and retention of an exclusive niche or food source. Although pathogenesis is a common outcome of parasitism, it is possible that where fungi on wild plants confront agriculture, plant pathogenesis might evolve from forms of parasitism less costly to the host, such as endophytic or epiphytic association, from parasitism on another host, or from saprotrophy. Compared with the patchiness of wild plant populations, monoculture presents large, relatively homogeneous, seasonally replaced plant populations that could favor a change to pathogenesis. While initial colonization or recolonization may be by sexually produced meiospores, asexual reproduction is probably key to the invasive or epidemic spread of many fungal pathogens. Asexual propagules include rnitospores, sclerotia, and assorted melanized hyphal aggregates, such as bulbils and rhizomorphs. In fungi, clonality may result from mitosis in asexual reproduction, from selfed, homothallic sexual reproduction, and more trivially, from the mitotic division(s) following meiosis of the kind that results in eight (or more), rather than four ascospores. Clones can persist for at least as long as the epidemic or seasonal spread of the pathogen. In some fungi, clones are maintained over the longer term, measured year to year or across many host to host transfers. Individual members of a clone are dispersed principally by wind, water, or animal vectors, including man: over long distances. Pathogenic fungi exploit both their ability to reproduce asexually and the disseminability of their asexual propagules to rapidly colonize host populations as they become available. Asexual reproduction, with the outcome of clonality, also facilitates colonization of nonliving substrates by animal and plant pathogens with saprotrophic phases, as well as by ruderal opportunists, such as Penicillium and Aspergillus. Asexual or parthenogenetic reproduction has been associated with invasive insect groups entering areas unoccupied by the ancestral bisexual forms and with invasive perennial plant species, e.g., water hyacinth (Eichornia crassipes) with "its rapid powers of clonal multiplication and water dispersal of vegetative fragments" (Barrett and Richardson 1986). Not surprisingly, clonality has been reported in bacterial, protistan, and fungal pathogens of -humans (Lenski 1993; Maynard Smith et al. 1993; Pujol et al. 1993; Tibayrenc et al. 1991).

Detection of clonality Clonality at its most obvious and prevalent extreme may be detected directly by the repeated recovery of genotypes that can be identified with independent genetic markers and by

the repeated recovery of such genotypes over a wide geographical area or from year to year. Clonality can also be detected indirectly by testing for random mating. Under the assumption of random mating throughout an entire population, termed panmixis, each individual in a population has an equal chance of mating with every other individual. Milgroom (1995~)describes four tests, all based on the random association of alleles and independent loci that would be expected to result from random mating. An additional test uses a phylogenetic approach based on the assumption that in a purely clonal population, all individual trees inferred for different segments of DNA will be congruent. In populations in which genetic exchange and recombination occur, sequences are likely to have different patterns of descent and trees will be incongruent (Burt et al. 1994). Sample size is critical to these tests because too small a sample may lead to an erroneous result, most probably a failure to reject the null hypothesis of panmixis and to detect clonality. Measurements used in these tests may be from samples of fungal pathogen populations on crop hosts over two or more seasons (Brown and Wolfe 1990; Chen et al. 1994). To test for panmixis in mixed clonal and sexually mating populations on perennial hosts, samples made over 1 year may be censored to remove clonal (repeated) elements, thereby considering each genotype only once (Milgroom et al. 1992). The extent to which observed mating approximates an expected random pattern can depend on the scale of sampling, which may include undetected sibling species or what is clearly a single, genetic species collected over global areas, continents, local regions, field populations (crops), or in a census of single host plants or individual lesions. Even where mating is not random at some hierarchical level of sampling because of genetic division along host, geographic, or other lines, mating may still approximate a random pattern at lower hierarchical levels. An additional factor is spatial clustering of clonal genotypes at the lowest level of sarnpling, the census, which can disguise random mating at an intermediate or higher level of sampling (Milgroom 19956). Within subpopulations with smaller genetic neighborhoods than the larger population (i.e., with reduced gene flow), there may be reduced levels of heterozygosity and even fixation of certain genes. Consequently, there will be a smaller number of segregating markers in the subpopulation than in the large population. Unlike selfing, in this kind of inbreeding genetic exchange (crossing) and recombination are occurring among very closely related but not identical genotypes. Under these circumstances, clonal reproduction leaves an ambiguous signature; even though inbreeding and clonality are distinct modes of reproduction, in an inbreeding population it becomes difficult to distinguish clonality. A possible example of this scenario in field populations of Sclerotinia sclerotiorum on a wild plant is discussed below.

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

H I .4. Kohn

Even when clonality is indicated by direct observation or indirect tests, the combination of sexual and asexual reproductive strategies in many fungal species predicts a mixture of sexual mating and clonality. Using a modification of a test for random mating, the multilocus associations test of Brown et al. (1980), Maynard Smith et al. (1993) provided evidence that bacterial population structure, once thought to be predominantly clonal in many species, may actually range from panmictic to strictly clonal on all levels of hierarchical sampling. Some populations include both clonal and interbreeding, sexual elements; for example, populations of Neisseria meningitidis are effectively panmictic, but some epidemic genotypes may increase explosively by clonal means. Some species, such as Rhizobium meliloti, comprise genetically distinct groups that are possibly geographically or ecologically partitioned. These groups are fixed for certain alleles and appear to be distinct lineages; mating and recombination do not occur between groups. Within each group, however, there is random association between alleles at different loci that segregate in the population. The novelty in the analyses of Maynard Smith et al. (1993) was the exploratory use of subsamples, e.g., comparing tests of the entire sample with tests of the sample censored to include only one representative of each genotype, or censored to include only one representative of each group or cluster of genotypes. Small subsample size is a potential hazard to this approach.

Clonal evolution The vernacular and genetic applications of the term clone have led to an expectation of uniformity among individual members of a clone, of ramets of a genet. All members of a clone, a mitotic lineage, would theoretically be identical carbon copies in the absence of mutation at any point in the lineage. More plausibly, mutations occur in individuals resulting in tree-like patterns of evolutionary divergence from the ancestral cell line (see Fig. 1A of Maynard Smith et al. 1993). In other words, members of a clone may harbor variability, especially in evolutionarily older clonal lineages or in lineages with high mutation rates. Without recombination, a genotype can only produce offspring with fewer mutations than it carries by back mutation, an inefficient process at best. Clonal lineages may not be immortal. A synergistic interaction of accumulated deleterious mutations, leading to reduced population size and followed by fixation of deleterious mutations by random drift and final extinction, an extension of "Muller's ratchet," was invoked by Lynch and Gabriel (1990) in modeling "mutational melt down" in small populations. Their model predicts extinction of clonal lineages in lo4- lo5 generations. The potential for evolutionary diversification within clonal lineages should be considered when characterizing collections of isolates, especially with neutral markers that show a high degree of variability in fungi that reproduce exclusively or extensively by asexual means. In a recent study of Pyricularia grisea (teleomorph, Magnaporthe grisea) in genetically diverse, Colombian disease breeding nursery (40 pathotypes in a 30-ha area), Levy et al. (1993) identified six clonal lineages, each with a specific array of pathotypes and each infecting a limited number of rice cultivars. DNA fingerprints were highly related within each lineage, but patho-

types varied. This contrasted with results from a previous study on a less genetically diverse (eight pathotypes), U.S. sample from an archival culture collection in which each clonal lineage showed modal virulence characteristics, with one pathotype per lineage, that had apparently been stably maintained for the 30-year period over which the collection was assembled (Levy et al. 1991). Levy et al. (1993) speculated that the differences in pathotype polymorphism within clonal lineages in the two samples might have resulted from sampling bias or weak selection. Sampling bias in the U.S. collection could have favoured the fittest genotypes established under episodic, epidemic conditions rather than chronic conditions, or isolates with strong, definitive reactions on international rice differentials. Weak selection could have been the outcome of relatively little breeding for resistance in the U.S., resulting in less diversity in resistance genes in blast-resistant cultivars than in the blast disease breeding nursery in Colombia.

The clonal dynamic The carry-over and dispersal of clones, as well as the stability or change in clone frequencies during the disease cycle, is what I term the clonal dynamic. Since any fungus that reproduces asexually will produce clones, an obvious question is whether these clones are transient or stable. In agricultural populations of Sclerotinia sclerotiorum, discussed in more detail below, some clones are carried over from year to year, recovered at high frequency in population samples, and widely dispersed geographically. The stability of these clones is probably due to a combination of asexual reproduction and predominantly, though I suspect not exclusively, selfed, homothallic sexual reproduction plus efficient dispersal over time. With or without strong selection, clones carried over from year to year could amplify and become well dispersed. New clones or genotypes would tend to be sampled at lower frequencies and have more local dispersal. In heterothallic fungi, at least those pathogenic on annual plants, the available evidence suggests that clones are seasonal, perhaps epidemic, and are replaced by recombinant genotypes when sex occurs. This phenomenon was observed in a 3-year study of Mycosphaerella graminicola in which recurring alleles but not clones were found from year to year, indicative of a sexual founding population each year (Chen et al. 1994). In the oomycete Phytophthora infestans, epidemics in agriculture are caused by explosive increases in asexually produced sporangia that can be dispersed for tens of kilomeires (Fry et al. 1992); such agricultural populations are ephemeral, with potentially limited carry-over from season to season or year to year. However, epidemiological and genetic evidence indicate that after migrating from Central Mexico to Europe, a single clone of P. infestans caused the Irish potato famine; today this clonal lineage is dispersed worldwide (Goodwin et al. 1994). Drenth et al. (1994 and personal communication) reported that although asexual clonal genotypes of P. infestans cause most of the seasonal, epidemic spread of late blight of potato in Europe, these clones are to some extent replaced each season by recombinant genotypes when sex occurs, probably at least partly owing to the ability of the sexual oospores with their durable walls to perennate in soil compared with the asexual sporan-

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

Can. J. Bot. Vol. 73 (Suppl. I ) , 1995

gia. In S. sclerotiorum it is the asexual propagules, the sclerotia, that resist predation, desiccation, and freezing to carry clonal genotypes over in soil for several years. In contrast, the structure of one population of Cryphonectria parasitic~,a pathogen that forms perennial disease lesions on a perennial plant, was determined to be a mixture of clonality and panmixis, with an estimate of the clonal fraction of the population of approximately 15% (Milgroom et al. 1992; see also Milgroom 19956). The spatial distribution of clones can indicate the scale of clonality or whether inoculum sources are immigrant or resident within the field population. A critical feature of analyses for spatial distribution is the comparison of observed average distances between individuals of the same clone, made from samples of adequate size, to expected average values from randomized data permutations (for a review, see Milgroom 1995~).In S. sclerotiorum, discussed below, immigration of clonal genotypes and intermixing with resident genotypes is indicated by the spatial mixing at all scales of sampling and the wide geographical dispersal of clones. Another strategy is the immigration of genetically diverse, sexually produced founding populations, such as ascospores produced via random mating. These founding populations initially colonize plants in the field, then increase over the season by asexual reproduction resulting in fine-scaled, clustered patterns of individuals belonging to the same clone. This has been observed in Mycosphaerella graminicola (McDonald and Martinez 1990). During the disease cycle, fluctuations in immigration of clones, selection for certain clones, or chance (especially under some conditions, for example in small populations) could result in changes in clone frequencies during the season. Such a change within the disease cycle is described below for Sclerotinia sclerotiorum. The recent history of the Dutch elm disease Dandemic has shown ex~losiveincreases at epidemic fronts, putatively under strong, episodic selection, of possibly clonal, vegetative compatibility "super groups," which may maintain reproductive isolation over time by maintaining the low frequency of one mating type or by other factors reducing outbreeding ability (Bates et al. 1993; Brasier 1987, 1991; Mitchell and Brasier 1994).

Clonality and changes in pathogenic specificity If rapid increases of adaptive, clonal genotypes give fungi with asexual reproduction the opportunity invasively to exploit a host or resource when it becomes available, then why not also the opportunity to invade a new host or resource? Although much work is in progress on the spread of pathogens in systems such as barley powdery mildew (Wolfe et al. 1992), Dutch elm disease, late blight of potato, rice blast, and the spread of introduced pathogens, evidence is still circumstantial or incomplete for the movement of pathogens from wild to agricultural hosts. This work will require a combination of phylogenetic and population approaches. Colletrotrichum graminicola (teleomorph is Glomerella graminicola) infects Zea mays (corn, maize), Sorghum bicolor (grain sorghum, milo, sweet sorghum, sorghum), and S. halepense (Johnson grass, Egyptian millet, allepo grass), as well as other cereals and legumes (Farr

et al. 1989). Vaillancourt and Hanau (1992) determined that isolates from maize and sorghum are morphologically similar but genetically distinct (i.e., not interfertile) and concluded that they are sibling species. Work in progress comparing populations (N > 50) from Johnson grass to those from sorghum shows significant genetic differences. Among six anonymous RFLP loci, only one allele was shared between sorghum and Johnson grass (Rosewich et al., personal communication). The differences in C. graminicola on these hosts suggest, among other possibilities, the divergence of reproductively isolated lineages, perhaps originally clonal, within each of which crossing now occurs. If plants have basic resistance to pathogens and other stresses, one outcome of the acquisition of pathogenicity factors in a fungus in response to selective pressure could be plants that tolerate infection with no loss of fitness (Heath 1991). A possible route from endophytic symbiosis or weak parasitism on a wild plant to necrotrophic pathogenesis on an agricultural host is- worth consideration. hat a genetic change as simple as a mutation at a single locus can effect such a change in an agricultural pathogen, although in this case from pathogen to endophyte, was demonstrated by Freeman and Rodriquez (1993). A nonpathogenic mutant of Colletotrichum magna, a pathogen that combines biotrophic and necrotro~hicelements in infection and establishment Drocesses, grew through cucurbit tissues as an endophyte but maintained wild-type levels of in vitro sporulation and infection characteristics as well as the wild-type host range. One might extrapolate from this and speculate that simple genetic change could effect a transfer from a wild host to a new, closely related agricultural host. A candidate for testing this hypothesis is ~ e ~ t o s ~ h a emaculans ria (anamorph, ~ h o m a lingam), causal agent of black leg of brassicas, particularly oilseed crops. After infection by airborne ascospores or splash-dispersed conidia, L. maculans establishes a symptomless, biotrophic, endophytic relationship within host cells. Disease is initiated when the fungus begins to induce cell death and acquire nutrition saprotrophically, forming lesions and cankers (Williams 1992). Two subgroups have been distinguished within the species, one weakly and the other strongly pathogenic on Brassica (Plummer et al. 1994; Williams 1992). Other specialized groups of isolates from the wild crucifers Thlaspi, Sisymbrium, and Descurania have been reported. The 7'hlaspi group, which can be found in 7'hlaspi growing as weeds in Brassica production areas, is known to be nonaggressive on brassicas (Williams 1992). Based on DNA sequence diversity, the three groups appear to be divergent and have been considered to represent distinct species (Morales et al. 1993; Williams 1992). Determination of population structure in these subgroups based on samples from wild and cultivated hosts and comparison with phylogenies might elucidate what seem like multiple jumps from wild to agricultural hosts and from endophytism to necrotrophic pathogenesis. Studies of genetic diversity tend to reveal subgroupings. When are these subgroups species? They may be species if they are reproductively isolated, i.e., they do not interbreed in nature, have unique phenotypic characters, ideally more than one kind of character, and are distinguished by congruent phylogenies based on different character sets. They may also represent divergent lineages within populations. If -

-

H I .4. Kohn

evolutionary processes are to be understood, sampling of field populations to test theories about the relationships of divergent groups will be much more informative than sampling isolates of taxa. This is the approach that has been taken in studies of Sclerotinia sclerotiorum on agricultural plant hosts and one wild host, Ranunculus jicaria.

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

The clonal dynamic in S. sclerotiorum on agricultural crops Sclerotinia sclerotiorum has been reported as a pathogen of over 360 species of wild and cultivated plants (Purdy 1979), and in fact, probably can infect most broad-leaved plants. Diseases caused by S. sclerotiorum are most prevalent in the families Solanaceae, Cruciferae, Umbelliferae, Compositae, Chenopodiaceae, and Leguminosae (Willetts and Wong 1980). Cereals are generally outside of the host range of this species, although there have been rare reports on plants in the Poaceae, for example, sorghum (Farr et al. 1989). The species is distributedworldiide but is more commonly encountered in temperate climates. Sclerotinia sclerotiorum has a haploid, multinucleate, hyphal somatic state that differentiatesinto sclerotia, which are compact masses of hyphae enveloped by a rind of cells with hard, melanized walls. Sclerotinia sclerotiorum does not produce disseminative conidia, although it produces microconidia that do not germinate (Kohn 1979) and presumably may function as spermatia. Sclerotia function as asexual propagules, capable of perennation in soil for documented periods of 4-5 years in soil (Adams and Ayers 1979). After physiological conditioning, sclerotia germinate to form stalked, cup-shaped, sexual fruit bodies, termed apothecia. Meiospores, the ascospores, are produced in asci in the apothecium and are discharged into the air. This species is homothallic and, on the basis of in vitro fruiting studies, seems to preferentially self-fertilize. No evidence of recombinant genotypes has been observed to date in screening of sibling monosporous isolates from fieldcollected apothecia (Kohli et al. 1992, 1995). Sclerotinia sclerotiorum is a necrotrophic parasite with a brief, initial biotrophic phase (Lumsden 1979; Tariq and Jeffries 1985, 1987). Diseases caused by S. sclerotiorum are either initiated on aerial plant parts by ascospores, as in stem rot of canola (Brassica napus or B. rapa), or around the stem-root interface by soil-borne sclerotia, as in basal stem rot and wilt of sunflower (Helianthus annuus). We have sampled S. sclerotiorum from canola crops, which I will term field populations, for a total of 4 years in Canada (Kohn et al. 1991: Kohli et al. 1992. 1995). Field populations of S. sclerotiorum have three basic characteristics. First, they are clonal. Second, each field is infected or infested by several clones. Clonality was detected by means of four presumably unrelated genetic criteria: mycelial compatibility grouping, DNA fingerprinting, a size dimorphism in the mitochondria1 small subunit rRNA gene which is due to the presence or absence of a group I intron, and RFLPs in Southern hybridizations with a probe containing the mitochondrial 24s rRNA gene from Neurospora crassa (Kohn et al. 1991; Carbone et al. 1995). Clones are identified by DNA fingerprinting with pLK44.20, a probe containing a repeated, dispersed element of nuclear DNA from S. sclerotiorum, and by mycelial compatibility testing. Individ-

Fig. 1. Histogram of frequencies of the 30 most abundant clones in the field studies in 1991 (Meadow Lake, Sask.) and in 1992 (Olds, Alta.). Counts include all isolates in the main sample, excluding the subsample of intensively sampled plots. In addition to the 30 clones shown, there were 2 clones isolated six times, 4 clones isolated five times, 6 clones isolated four times, 7 clones isolated three times, 45 clones isolated twice, and 443 clones isolated only once each for a total of 537 clones in the main sample.

Clones

ual members of some clones can be recovered from widely separated geographic locations repeatedly over several years. Of the 659 genotypes identified so far from 2876 isolates, one clone was frequently sampled in all 4 years of sampling, over a distance of ca. 2000 krn from central Ontario to northern Alberta, and nine clones were frequently sampled over a 3-year period in western Canada. However, while some clones have wide geographical dispersal and are recovered in high frequencies over several years, many clones or genotypes occur locally and at low frequency. The clonal dynamic, specifically clonal dispersal and spatial mixing within the disease cycle, was investigated in a sample of two fields in Saskatchewan in 1991 and two fields in Alberta in 1992 (Kohli et al. 1995). Each field was sampled in a grid of 128 quadrats (50 x 50 m2) plus 4 intensive quadrats each sampled in a diagonal transect. To detect any changes in clone frequency or spatial dispersal of clones during the disease cycle, fields were sampled twice, first from ascospore inoculum on canola petals in July and second from stem lesions in August. From a total of 2747 isolates, 594 unique genotypes were identified. In each field, a small number of clones represented the majority of the sample (Fig. l), with a large number of clones or genotypes sampled once or twice. Based on comparisons of clone frequencies with X2 tests, profiles of clone frequencies were significantly different between 1991 and 1992; profiles were not significantly different between the two fields in 1991 but significantly different between the two fields in 1992, indicating some local population substructure. Profiles of clone frequencies between petal and lesion samples were not significantly different in each of the two fields sampled in 1991 but significantly different in each of the two fields sampled in 1992, consistent either with selection or with continued immigration of some clones after the July sampling. Two new clones, not detected in the 1989, 1990, or 1991 samples, were reco-

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

Can. J. Bot. Vol. 73 (Suppl. I ) , 1995 vered at high frequency in the 1992 sample. Randomization tests indicated that individual members of clones were randomly intermixed, not spatially aggregated, on both canola petals and stems with disease lesions. Also, individual members of a clone on lesions were not closer to individuals of the same clone from petals than could be predicted by chance. These data give rise to several hypotheses. First, there is a set of clones that occur at high frequency in colonization of canola crops year after year that is common to the Canadian "canola belt," where canola has been grown in rotation for a long time (20-40 years). This clonal population is the pool from which field populations are recolonized. These clones are mixed both locally, in field populations, and more widely over this geographical area. Whether high frequency among some clones is determined by selection or chance cannot be determined without further testing. However, data from greenhouse experiments suggest that some clones are more aggressive on canola than others and preliminary analysis of in vitro fruiting experiments indicates that clones may differ in fecundity, measured as the number of apothecia produced per unit sclerotia (data not shown). Second, the lack of spatial aggregation of clones could result from colonization of a canola-crop by a mixture of immigrant and resident ascospore inoculum, rather than by resident inoculum alone. Finally, the large number of clones or genotypes sampled at low frequency plus the observation of two novel clones sampled at high frequency in the 1992 sample suggest that new genotypes evolve by sexual exchange and recombination and a new clone can build in local frequency before dispersing more widely. Although our studies so far have dealt with clones as they are, not with the evolution of new clones, we favor the hypothesis that some outbreeding and recombination is occurring, perhaps too infrequently to be detected by our field samples of apothecia. It is possible that recombination can only occur within clonal lineages (for example, Fig. lB of Maynard Smith et al. 1993). Sampling of S.sclerotiorum on sunflower over 2 vears indicates that clone frequency profiles are comparable to those observed on canola (Dhillon et al., unpublished data; Kohli et al. 1995). Preliminary analysis of isolates from a weed sample made in and around the Alberta canola fields in 1992 revealed both canola genotypes and unique genotypes, as would be expected in comparable samples from crops. Did S. sclerotiorum evolve on a wild or a cultivated plant? To address this question, one must consider both the host range of S. sclerotiorum and of its sister species in Sclerotinia. Of the three described species that are pathogens of cultivated plants, only S. sclerotiorum has a host range including wild and cultivated plants in many families; S. trifoliorum is limited to wild and forage legumes and S. minor, to peanuts, lettuce, and sunflower. These Sclerotinia species are probably very closely related, but are well defined by a variety of morphological and biochemical characters (Kohn et al. 1988). In a recent study of sequence divergence in the internal transcribed spacer 1 of the nuclear ribosomal RNA genome, Carbone and Kohn (1993) postulate a recent, common ancestry for these three species, as well as Sclerotium cepivorum, an asexual species with a host range limited to Allium spp. and other liliaceous hosts (Farr et al. 1989). Was speciation in this group preceeded by radiation from one or -

-

-

-

more wild hosts to cultivated hosts? Even with phylogenetic study of extant material, reconstruction of such an event may be hampered by an inappropriate sample. If contemporary populations of Sclerotinia sclerotiorum on cultivated hosts and weeds in crops of these hosts are compared with populations of S. sclerotiorum on wild hosts, we can at least determine whether there is genetic exchange between these wild and cultivated populations. Comparable studies could be done with S. trifoliorum and possibly with S. minor and Sclerotium cepivorum. If wild and cultivated populations are genetically distinct, we may be able to assess their relationship.

Population structure in S. sclerotiorum on a wild, perennial host Sclerotiniaficariae was described by Rehm (1893) as "probably a parasite of roots of Ranunculus ficaria," a common woodland species in Europe, easily distinguished from another sclerotiniaceous parasite of Ranunculus, S. tuberosa, by its smaller apothecia. The description makes a sharp distinction in ascus and ascospore size between S. ficariae and S. sclerotiorum (as S. libertiana) which Rehm redescribed and listed as occurring "on roots of Brassica Rapa, Beta, Raphanus, Foeniculum, Phaseolus vulgaris, Petunia, and Zinnia elegans in damp places." After examining the type specimen, collected from R. ficaria in the Berlin Botanical Garden, I synonymized this species under S. sclerotiorum. In later studies on RFLPs among Sclerotinia species, I confirmed this synonymy (Kohn et al. 1988). More recently, the internal transcribed spacer of one of the S. ficariae isolates was sequenced and found to be identical to those of agricultural isolates of S. sclerotiorum and S. minor (Carbone and Kohn 1993). Also, a group I intron in the mitochondria1 small subunit rRNA gene was sequenced in S. sclerotiorum and then identified in isolates of S. sclerotiorum, including R. ficaria isolates, S. minor, and S. trifoliorum (Carbone et al. 1995). In 1993, I sampled two sites where Sclerotinia sclerotiorum produces abundant apothecia under dense canopies of Ranunculus ficaria in Norway. Ranunculus ficaria is a perennial plant and spreads clonally by rhizomes, offering a seasonally recurring, stable host population; both sites are known to have persisted for several years. Apothecia are produced for about 2 weeks when R. ficaria is in bloom. Like Rehm, I suspect that S. sclerotiorum infects the rhizomes of R. ficaria. But because no symptoms of disease can be discerned and the fungus has not grown out of excised parts of the few plants sampled to date, further proof of pathogenicity is needed. It is unlikely that a host other than R. ficaria is infected since the plant so dominates the sites, but endophytism or localized, latent infection followed by late-season saprophytism are some possible alternatives to disease. The objective of the study was to compare wild field populations of S. sclerotiorum with populations on cultivated hosts. Two questions were addressed. First, would the two criteria detecting clonality in cultivated populations in North America , DNA fingerprinting with pLK44.20 and mycelial compatibility, be associated in wild populations? Second, if wild populations are clonal, are clones spatially mixed, as in cultivated populations, or locally aggregated? One of the

H I .4. Kohn

Fig. 2. Southern hybridizations of BatnHI-digested DNAs of Sclerotiriia sclerotiorum. Lanes 1 - 18 are from Norway, from canola (lanes 1 - 13), from potato (lanes 14- 17), and from the Vestfold site with Ranunculusficaria (lane 18, solid

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

arrow). Lane 19 (open arrow) is from isolate LMK 21 1, from Canadian canola, used as a standard in DNA fingerprinting of S. sclerotiorurn. Horizontal bars indicate groups of isolates collected from a single field. Norwegian agricultural isolates from canola and potato that shared the same unique fingerprint were also mycelially compatible.

- - - -

Fig. 3. Expected relationship between DNA fingerprints and mycelial compatibility in agricultural field populations of S. sclerotiorutn in North America and Norway. Apo denotes apothecium and Spore denotes sibling monosporous isolate from the apothecium with the same identifying number. DNA fingerprint types are designated as X, Y, and Z. Solid squares denote compatible mycelial interactions and open circles denote incompatible interactions, 7 days after inoculation.

X

Table 1. DNA fingerprint type for two to five sibling, monosporous isolates from apothecia of Sclerotinia sclerotiorutn collected with Ranurzculusficaria in 1993 at sites on Sandvika and Vestfold transects. No.

Sandvika

1

No. of apothecia"

DNA fingerprint

2 3 4 5 6 Vestfold

z

t+

x

Site

Y

1

2 3 "No variation in fingerprints was observed among monosporous isolates from the same apothecium.

sites, in Sandvika, a suburban area outside of Oslo well away from regions of agricultural production, was located in a wooded park and was ca. 380 m long. The other site, in Vestfold, an agricultural area where some canola is grown, was

I

Apoi. Spore3 Apo2. Spore1

located in a hedgerow between fields in cereal -pasture rotation and was ca. 200 m long. At each location, 6 apothecia were sampled from a 1-m2 quadrat with quadrats spaced approximately 40-70 m apart. I sampled a total of five sibling monosporous isolates from each of 36 apothecia at six sites at Sandvika and each of 18 apothecia at three sites at Vestfold. A sample of sclerotia of S. sclerotiorum from canola and potato in Norway was also acquired. Isolates from R. $curia showed remarkable phenotypic variability compared with the generally uniform appearance of agricultural isolates from both Canada and Norway; variability was observed in growth rate, pigmentation, and amount of aerial mycelium. DNA fingerprints with pLK44.20 (for methodology, see Kohn et al. 1991; Kohli et al. 1995) were much more complex, with many more fragments hybridizing to the probe DNA than were observed in fingerprints of DNAs from Canadian and Norwegian agricultural isolates (Fig. 2). Mean numbers of hybridizing fragments were 21 and 12, for wild and agricultural isolates, respectively. From this sample of isolates, five unique DNA fingerprints from Sandvika and two unique fingerprints from Vestfold were identified (Table 1). Expected results from DNA fingerprinting and mycelial compatibility testing of S. sclerotiorurn from agricultural field populations in North America and Norway are shown in Fig. 3. Observed results in a subsample of S. sclerotiorum isolates from the two wild field populations of R. $curia are shown in Fig. 4. The structure of the two wild Ranunculus populations differs from that of agricultural populations examined to date in four major ways. First, in the wild popu-

Can. J . Bot. Vol. 73 (Suppl. I ) , 1995

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

Fig. 4. Observed relationship between DNA fingerprints and

mycelial compatibility in two Norwegian field populations of S. sclerotiorum on the wild plant Ranunculus ficaria. In the isolate numbering system, S1-A3-I denotes monosporous isolate 1 from apothecium 3 (A3) at site one (Sl) in the transect. (a) The Sandvika site; (b) the Vestfold site. Solid squares denote compatible mycelial interactions and open circles denote incompatible interactions, 7 days after inoculation. Dashes denote missing data. Segregation was observed in S1-A3, S3-A4, S6-A2 (a) and WI-A3 and W3-A3 (b); for each of these apothecia, the sibling monosporous isolates were not identical in their compatibility reactions.

(a)

Sandvika transect

Site S1

H

Site S3

Site S6 240 m

140 m

D A ---g z

B

A

H

C

;

" " " $ Z $ $ $ &

~

2

~

, , , , , ,

(b) Site W1

H

Vestfold transect 130 m

Site W2

70 m

Site W3

H

~

lations, fingerprints and mycelial compatibility reactions are decoupled; in agricultural populations in North America, Norway, and probably elsewhere, each clone has a unique fingerprint and all individual members of a clone are mycelially compatible with all members of the same clone but incompatible with members of other clones. Second, fingerprint diversity is low and highly localized within the two Ranunculus populations; five apothecia from an agricultural field would likely represent 4 or 5 fingerprints. Third, there in~ sibling ascospores is evidence of s o m e ~ o ~ t c r o s s because from one apothecium may show different mycelial compatibility reactions, indicative of crossing and segregation; segregation has not been observed among sibling monosporous isolates from field-collected apothecia or from in vitro grown apothecia in attempted crosses between clones. Last, within what is apparently an isolated and inbred population, strong spatial substructuring is evident in the distribution of fingerprints; in agricultural populations examined to date, clones have been randomly dispersed spatially. I believe? that the Sandvika and Vestfold populations are each isolated and inbred, and that within these populations only a small number of alleles determining mycelial compatibility and DNA fingerprint are actually segregating and h o s t are fixed. I speculate that identical fingerprint or mycelial compatibility types can arise convergently, in independent crosses within these inbreeding, spatial s u b g r o u p s . ~ ~ o n s e quently, clonality in these inbred groups cannot be detected with the markers presently available. It will be difficult to distinguish between inbreeding and clonality without an exhaustive study of additional markers as well as exploratory analysis. What is suggested by the data on the wild Ranunculus populations is highly local, site-specific, near fixation of fingerprint and of compatibility o r incompatibility indicative of spatial structuring and substructuring of these populations quite unlike the spatial mixing observed in agricultural populations of S. sclerotiorum. That agricultural practices probably tend to mix and homogenize soil-borne pathogen populations is supported both by our observations in agricultural populations of S. sclerotiorum and also by the report of Gordon et al. (1992) of random distribution of mitochondria1 haplotypes of Fusarium oxysporum in cultivated soil compared with highly aggregated distribution in native soil. Within each of the two Ranunculus populations, the patterns of congruency and incongruency between fingerprint and mycelial compatibility imply that sexual crossing may be occurring within, but not necessarily between, spatially substructured lineages (see Fig. 1B of Maynard Smith et al. 1993). -

-

-

Perspectives on the role of clonality in plant pathogen populations

-

What does the comparison of wild field populations with agricultural field populations of S. sclerotiorum tell us about the role of clonality in the evolution of plant -pathogen interactions? Mainly that clonality, year to year carry-over of clones, spatial mixing via immigration, and possibly selection pressure on clone frequencies may all be associated with monoculture, agricultural practices, and the movement of agricultural materials, such as seed and machinery. This

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

H I .4. Kohn

leads to some speculative predictions that will b e tested with neutral markers in fungal systems in coming years. In wild pathogen populations of fungi with sexual reproduction that is not exclusively through selfing, some gene flow and interbreeding is expected on a wide geographical scale, but inbreeding could b e significant o n the local population scale. Among pathogen populations on wild perennial plants, in the presence of both sexual and asexual forms of reproduction, clonality would probably be detected at the most local, census level, but could be significant at local o r wider population levels in epidemics. Wild annual plants are probably colonized by a mixture of clonal pathogen genotypes and genotypes arising from sexual reproduction, perhaps in patterns approximating metapopulations, with seasonal, epidemic spread by asexual, clonal genotypes. Invasion of agricultural hosts by pathogen genotypes from wild plant populations, however, may be expected to follow epidemic, clonal patterns as single adaptive genotypes arise and proliferate through selection o r serendipity within the wild population.

References Adams, P.B., and Ayers, W.A. 1979. Ecology of Sclerotinia species. Phytopathology , 69: 896 - 899. Barrett, S.C.H., and Richardson, B.J. 1986. Genetic attributes of invading species. In Ecology of biological invasions. Edited by R.H. Groves and J.J. Burdon. Cambridge University Press, Cambridge, U.K. Bates, M.R., Buck, K.W., and Brasier, C.M. 1993. Molecular relationships between Ophiostorna ulrni and the NAN and EAN races of Ophiostoma novo-rrlmi determined by restriction fragment length. Mycol. Res. 97: 449-455. Brasier, C.M. 1987. The dynamics of fungal speciation. hr Evolutionary biology of the fungi. Edited by A.D.M. Rayner, C.M. Brasier, and D. Moore. Cambridge University Press, Cambridge, U.K. Brasier, C.M. 1991. Ophiostorna novo-ulmi sp. nov., causative agent of current dutch elm disease pandemics. Mycopathologia, 115: 151- 162. Brown, A.H.D., Feldman, M.W., and Nevo, E. 1980. Multilocus structure of natural populations of Hordeurn spontaneurn. Genetics, 96: 523-536. Brown, J.K.M., and Wolfe, M.S. 1990. Structure and evolution of a population of Erysiphe graminis f.sp. hordei. Plant Pathol. 39: 376-390. Burt, A,, Carter, D.A., Taylor, J.W., and White, T.J. 1994. Genetic epidemiology of the human pathogen Coccidioides immitis (Ascomycota). Fifth International Mycological Congress, August 14 -2 1, Vancouver, B.C. p. 26. (Abstr.) Carbone, I., and Kohn, L.M. 1993. Ribosomal DNA sequence divergence within internal transcribed spacer 1 of the Sclerotiniaceae. Mycologia, 85: 415 -427. Carbone, I., Anderson, J.B., and Kohn, L.M. 1995. A group I intron in the mitochondrial small subunit ribosomal RNA gene of Sclerotinia sclerotiorum. Curr. Genet. In press. Chen, R.-S., Boeger, J.M., and McDonald, B.A. 1994. Genetic stability in a population of a plant pathogenic fungus over time. Mol. Ecol. 3: 209-218. Drenth, A,, Tas, I.C.Q., and Govers, F. 1994. DNA fingerprinting uncovers a new sexually reproducing population of Phytophthora infestans in the Netherlands. Eur. J. Plant Pathol. 100: 97- 107. Farr, D.F., Bills, G.F., Chamuris, G.P., and Rossman, A.Y. 1989. Fungi on plants and plant products in the United

States. APS Press, St. Paul, Minn. Freeman, S., and Rodriguez, R.J. 1993. Genetic conversion of a fungal plant pathogen to nonpathogenic, endophytic mutualist. Science (Washington, D.C.), 260: 75 -78. Fry, W.E., Goodwin, S.B., Matuszak, J.M., Spielman, L.J., and Milgroom, M.G. 1992. Population genetics and intercontinental migrations of Phytophthora infestans. Annu. Rev. Phytopathol. 30: 107 - 129. Goodwin, S.B., Cohen, B.A, and Fry, W.E. 1994. Panglobal distribution of a single clonal lineage of the Irish potato famine fungus. Proc. Natl. Acad. Sci. U.S.A. 91: 11 591 - 11 595. Gordon, T.R., Okamoto, D., and Milgroom, M. 1992. The structure and interrelationship of fungal populations in native and cultivated soils. Mol. Ecol. 1: 24 1-249. Heath, M.C. 1991. Evolution of resistance to fungal parasitism in natural ecosystems. New Phytol. 119: 331 -343. Kohli, Y., Morrall, R.A.A., Anderson, J.B., and Kohn, L.M. 1992. Local and trans-Canadian clonal distribution of Sclerotinia sclerotiorum on canola. Phytopathology, 82: 875 -880. Kohli, Y., Brunner, L.J., Yoell, H., Milgroom, M.G., Anderson, J.B., Morrall, R.A.A., and Kohn, L.M. 1995. Clonal dispersal and spatial mixing in populations of the plant pathogenic fungus, Sclerotinia sclerotiorr~m.Mol. Ecol. In press. Kohn, L.M. 1979. A monographic revision of the genus Sclerotinia. Mycotaxon, 9: 365-444. Kohn, L.M., Petsche, D.M., Bailey, S.R., Novak, L.A., and Anderson, J.B. 1988. Restriction fragment length polymorphisms in nuclear and mitochondrial DNA of Sclerotinia species. Phytopathology, 78: 1047- 1051. Kohn, L.M., Carbone, I., Stasovski, E., Royer, J., and Anderson, J.B. 1991. Mycelial incompatibility and molecular markers identify genetic variability in field populations of Sclerotinia sclerotiorum. Phytopathology, 81: 480-485. Lenski, R.E. 1993. Assessing the genetic structure of microbial populations. Proc. Natl. Acad. Sci. U.S.A. 90: 4334-4336. Levy, M., Romao, J., Marchetti, M.A., and Hamer, J.E. 1991. DNA fingerprinting with a dispersed repeated sequence resolves pathotype diversity in the rice blast fungus. Plant Cell, 3: 95 - 102. Levy, M., Correa-Victoria, F.J., Zeigler, R., Xu, S., and Hamer, J.E. 1993. Genetic diversity of the rice blast fungus in a disease nursery in Columbia. Phytopathology, 83: 1427- 1433. Lumsden, R.D. 1979. Histology and physiology of pathogenesis in plant diseases caused by Sclerotinia species. Phy topathology , 69: 890 - 896. Lynch, M., and Gabriel, W. 1990. Mutation load and the survival of small populations. Evolution, 44: 1725 - 1737. Maynard Smith, J., Smith, N.H., O'Rourke, M., and Spratt, B.G. 1993. How clonal are bacteria? Proc. Natl. Acad. Sci. U.S.A. 90: 4384-4388. McDonald, B.A., and Martinez, J.P. 1990. DNA restriction fragment length polymorphisms among Mycosphaerella graminicola (anamorph Septoria tritici) isolates collected from a single wheat field. Phytopathology, 80: 1368- 1373. Milgroom, M.G. 1995a. Analysis of population structure in fungal plant pathogens. In Disease analysis through genetics and molecular biology: interdisciplinary bridges to improved sorghum and millet crops. Edited by J.F. Leslie and R.A. Frederiksen. Iowa State University Press, Ames, Iowa. In press. Milgroom, M.G. 19956. Population biology of the chestnut blight fungus, Cryphonectria parasitica. Can. J. Bot. 73(Suppl. 1). This issue.

Can. J. Bot. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

Can. J. Bot. Vol. 73 (Suppl. I), 1995 Milgroom, M.G., Lipari, S.E., and Powell, W.A. 1992. DNA fingerprinting and analysis of population structure in the chestnut blight fungus, Cryphonectria parasitica. Genetics, 131: 297 -306. Mitchell, A.G., and Brasier, C.M. 1994. Contrasting structure of European and North American populations of Ophiostoma ulmi. Mycol. Res. 98: 576-582. Morales, V.M., Pelcher, L.E., and Taylor, J.L. 1993. Comparison of the 5.8s rDNA and internal transcribed spacer sequences of isolates of Leptosphaeria maculans from different pathogenicity groups. Curr. Genet. 23: 490 -495. Plummer, K.M., Dunse, K., and Howlett, B.J. 1994. Non-aggressive strains of the blackleg fungus, Leptosphaeria maculans, are present in Australia and can be distinguished from aggressive strains by molecular analysis. Aust. J. Bot. 42: 1-8. Pujol, C., Reynes, J., Renaud, F., Raymond, M., Tibayrenc, M., Ayala, F.J., Janbon, F., MalliC, M., and Bastide, J.-M. 1993. The yeast Candida albicans has a clonal mode of reproduction in a population of infected human immunodeficiency virus-positive patients. Proc. Natl. Acad. Sci. U.S.A. 90: 9456-9459. Purdy, L.H. 1979. Sclerotinia sclerotiorum: history, diseases and symptomology, host range, geographic distribution, and impact. Phytopathology, 69: 875 -880. Rehm, H. 1893. Ascomyceten: Hysteriaceen und Discomyceten. In Kryptogamen-Flora von Deutschland, Oesterreich und der

Schweiz 11, 1(3), Lief. 40. Edited by L. Rabenhorst. E. Kummer, Leipzig. pp. 785-849. Tariq, N.-N., and Jeffries, P. 1985. Ultrastructure of penetration of Phaseolus spp. by Sclerotinia sclerotiorum. Can. J. Bot. 64: 2909-2915. Tariq, V.-N., and Jeffries, P. 1987. Cytochemical localization of lipolytic enzyme activity during penetration of host tissues by Sclerotinia sclerotiorum. Physiol. Mol. Plant Pathol. 30: 77-91. Tibayrenc, M., Kjellberg, F., Arnaud, J., Oury, B., Breniere, S.F., Darde, M., and Ayala, F.J. 1991. Are eukaryotic microorganisms clonal or sexual? A population genetics vantage. Proc. Natl. Acad. Sci. U.S.A. 88: 5129-5133. Vaillancourt, L.J., and Hanau, R.M. 1992. Genetic and morphological comparisons of Glomerella (Colletotrichum) isolates from maize and sorghum. Exp. Mycol. 16: 219-229. Willens, H.J., and Wong, J.A.-L. 1980. The biology of Sclerotinia sclerotiorum, S. trijolioru.n, and S. minor with emphasis on specific nomenclature. Bot. Rev. 46: 101 - 165. Williams, P.H. 1992. Biology of Leptosphaeria maculans. Can. J. Plant Pathol. 14: 30-35. Wolfe, M.S., Brandle, U., Koller, B., Limpert, E., McDermott, J.M., Miiller, K., and Schaffner, D. 1992. Barley mildew in Europe: population biology and host resistance. Euphytica, 63: 125- 139.