• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mol Ecol. Author manuscript; available in PMC Jul 1, 2009.
Published in final edited form as:
PMCID: PMC2653454

Population structure and diversity in sexual and asexual populations of the pathogenic fungus Melampsora lini


Many pathogens undergo both sexual and asexual reproduction to varying degrees, yet the ecological, genetic and evolutionary consequences of different reproductive strategies remain poorly understood. Here we investigate the population genetic structure of wild populations of the plant pathogen Melampsora lini on its host Linum marginale, using amplified fragment length polymorphism (AFLP) markers, two genes underlying pathogen virulence, and phenotypic variation in virulence. In Australia, M. lini occurs as two genetically and geographically divergent lineages (AA and AB), one of which is completely asexual (AB), and the other able to reproduce both clonally and sexually (AA). To quantify the genetic and evolutionary consequences of these different life histories, we sampled five populations in each of two biogeographical regions. Analysis of AFLP data obtained for 275 isolates revealed largely disjunct geographical distributions for the two different lineages, low genetic diversity within lineages, and strong genetic structure among populations within each region. We also detected significant divergence among populations for both Avr genes and virulence phenotypes, although generally these values were lower than those obtained with AFLP markers. Furthermore, isolates belonging to lineage AA collectively harboured significantly higher genotypic and phenotypic diversity than lineage AB isolates. Together these results illustrate the important roles of reproductive modes and geographical structure in the generation and maintenance of virulence diversity in populations of M. lini.

Keywords: avirulence, co-evolution, host-pathogen, metapopulation, Red Queen, virulence


Host–parasite interactions are thought to be a major force driving co-evolutionary change and generating biological diversity (Thompson 2005). In host–pathogen systems, co-evolutionary dynamics are largely contingent upon interactions between host resistance and pathogen infection strategies. Genetic variation in these characters in both host and pathogen populations is common, and thought to be a crucial factor influencing disease dynamics in human (Cooke & Hill 2001), plant (Alexander et al. 1993) and animal (Bishop & Morris 2007) systems. However, the underlying micro-evolutionary processes that generate and maintain such diversity in natural host–pathogen interactions, and the epidemiological consequences of such variation, remain poorly understood (Thrall & Burdon 2003).

Pathogenic species exhibit extensive diversity in their modes of reproduction (i.e. relative level of sexual vs. asexual reproduction) (Kohn 1995; Milgroom 1996), with potentially profound consequences for levels of genetic variation within populations (Milgroom 1996), population growth rate (Heitman 2006), persistence within populations (Barrett et al. 2007), and evolutionary change (McDonald & Linde 2002). The role and significance of sexual reproduction in interactions between pathogens and their hosts in particular has been heavily debated (e.g. Keeling & Rand 1995; Fox et al. 1996; Lively et al. 2004; Salathé et al. 2006). For example, the Red Queen Hypothesis predicts that sexual reproduction in host species is an adaptation against rapidly evolving parasites because it enables hosts to respond quickly to the appearance of novel pathogen genotypes through recombination of resistance genes (Hamilton 1980; Clay & Kover 1996). By extension, the Red Queen Hypothesis also predicts that parasites infecting sexual hosts should be genetically more variable than those infecting asexual hosts because recombination of resistance genes will drive negative frequency-dependent selection on the corresponding virulence genes of parasites (Clay & Kover 1996; Ooi & Yahara 1999).

Spatial and environmental heterogeneity in the interactions between parasites and their hosts can have a major influence on co-evolutionary dynamics across landscapes (Thompson 2005). Many host species, particularly those in natural ecosystems, exist as interacting groups of small, geographically and genetically differentiated populations (Burdon 1992), subdividing pathogen populations into relatively small, discrete units. Such spatial and environmental heterogeneity among pathogen populations is likely to strongly influence spatial patterns of disease incidence and persistence (Thrall & Burdon 1997), with populations of many species shown to undergo frequent local extinctions and recolonizations (Antonovics et al. 1994; Ericson et al. 1999; Thrall et al. 2001a; Smith et al. 2003; Laine & Hanski 2006). Thus, advances in the understanding of broader evolutionary processes in pathogenic species are most likely to come from a metapopulation approach that accounts for populations structured into interconnected demes with ongoing local processes of extinction and recolonization.

Within such a metapopulation framework, variation in different micro-evolutionary forces among demes may act to generate and maintain resistance and virulence polymorphisms. Local host populations, particularly in natural ecosystems, often vary spatially in the identity and diversity of resistance genotypes present (Jarosz & Burdon 1991; Bevan et al. 1993; Laine & Hanski 2006), and local adaptation of pathogens to their hosts has been demonstrated as a strong driver of pathogen population genetic structure in a number of host–pathogen interactions (Greischar & Koskella 2007). Within populations, negative frequency-dependent selection, where pathogen (or host) genotypes have higher fitness when rare, is a common assumption in many models of host–parasite interactions, and may contribute to the maintenance of resistance and virulence polymorphisms within populations (Gillespie 1975; Clarke 1976; Anderson & May 1982). In addition, it is also likely that there is strong potential for nonselective factors, such as random genetic drift, founder events and selection on linked traits, to influence host and pathogen evolution (Parker 1991; Burdon & Thompson 1995; Salathé et al. 2005).

Comparative analyses of geographical variation in genes controlling pathogenic traits and neutral molecular markers can provide insight into the relative roles of natural selection and genetic drift in driving divergence among pathogen populations (Zhan et al. 2005). For example, when selection is important and favours different pathotypes or particular virulence genes in different populations, we would expect among-population variation in pathogenicity traits to exceed that of neutral markers. Conversely, when selection favours similar pathogenic traits or functional genes among populations, estimates of population differentiation calculated for the pathogenic traits are expected to be lower than estimates derived from neutral markers (Lewontin & Krakauer 1973; Latta 2004).

In this study, we focus on the fungal plant pathogen Melampsora lini. In Australia, M. lini infects Linum marginale, an endemic wild, herbaceous plant species. The interaction between M. lini and L. marginale follows a gene-for-gene model (Burdon 1994), and spatially structured variation in pathogen virulence has been demonstrated at scales ranging from individual populations to the entire range of the interaction in Australia (Burdon & Jarosz 1992; Burdon et al. 2002; Thrall et al. 2002). In Australia, isolates of M. lini have been shown to fall into two distinct lineages (termed AA and AB; Barrett et al. 2007) that differ in key genetic, reproductive and life history traits. Continental scale sampling and microsatellite genotyping shows that while lineage AA isolates have low overall genetic diversity and heterozygosity, lineage AB isolates maintain fixed heterozygosity at corresponding loci. Furthermore, while lineage AB isolates consistently have one allele in common with lineage AA isolates, the second allele is consistently singular to lineage AB. Similarly, sequence data show fixed heterozygosity in lineage AB isolates for ITS and β-tubulin loci, with one allele identical to the allele recovered from lineage AA isolates, and low nucleotide divergence between the A and B clades (< 2%; L.G. Barrett, P.H. Thrall, P.N. Dodds, M. van der Merwe, C.C. Linde and J.J. Burdon, under review). No intermediate genotypes have been detected to date. These results (Barrett et al. 2007) suggest that lineage AB is a fixed F1 cross between two genetically divergent lineages of M. lini (i.e. a hybrid between lineage AA and a yet to be identified lineage BB), and that sexual recombination (i.e. meiosis) is unlikely to occur in lineage AB. In addition, the apparent inability of lineage AB isolates to effectively endure as resting spores (due to the lack of a sexual cycle) appears to drive disjunct geographical distributions among the lineages (Barrett et al. 2007), although the strength of this separation at local scales remains unclear.

We used a mixture of molecular and phenotypic approaches to explore the processes structuring genetic variation in M. lini across two distinct biogeographical regions (mountains and plains) in the state of New South Wales. These regions differ markedly in a range of factors likely to influence the geographical distributions of lineage AA and AB. The plains populations are comparatively hot and dry compared to the subalpine mountains populations; and the main growing seasons of the host are disjunct between regions (Burdon et al. 1999; Barrett et al. 2007). Furthermore, in the plains populations, where host plants survive over summer largely as underground rootstock, pathogens survive between epidemics exclusively in the dormant teliospore stage (during which meiosis takes place) on senescent host stems. In contrast, in the mountains populations, pathogens rarely form telia, and instead overwinter largely in the clonal urediospore stage on dormant green host shoots (Barrett et al. 2007). Given that lineage AB isolates seem unable to effectively form resting teliospores (Barrett et al. 2007), these observations lead to the expectation that lineage AA and AB will be largely geographically disjunct among the mountains and plains. This scenario thus offers an excellent opportunity to explore how genetic diversity and structure in a pathogenic species might be influenced by mode of reproduction. We used amplified fragment length polymorphism (AFLP) markers to assess the geographical distribution of lineages AA and AB, and combined AFLP markers with phenotypic virulence data and allelic data from two avirulence (Avr) loci to examine patterns of genetic and genotypic diversity within and between lineages AA and AB. We further use these markers to evaluate the relative importance of local selection, neutral drift and gene flow in generating divergence within lineages and among local pathogen populations through comparison of estimates of population genetic structure.

Materials and methods

Study sites

The pathogen populations used in this study occur in two geographically and environmentally distinct areas (Fig. 1). One of these areas lies within the western plains region of New South Wales. Of the five plains populations examined here, three occur in a cluster to the north and east (Garra, Molong, Larras Lee), with the other two occurring further to the west and south (Gundagai, Marrar) (Fig. 1). Together these populations are part of a formerly patchy, but continuous, distribution of Linum marginale across the western plains region of New South Wales (an area that has been extensively cleared for agriculture). The second region encompasses the Kiandra and Wild Horse Plains, in the northern part of the Kosciuzko National Park (the mountains populations). While the five mountains populations are all within 20 km of each other, the populations are spatially discrete, and earlier studies show populations in this region are quite distinct with regard to pathogen virulence and host resistance structure (e.g. Thrall et al. 2001b) (Fig. 1).

Fig. 1
Map with insert region showing locations of study regions and populations in New South Wales, Australia.

Phenological and epidemiological patterns in the LinumMelampsora interaction differ markedly between mountains and plains populations. In the mountains region, plants overwinter as underground rootstocks with a few short shoots protected from frost and snow by the surrounding vegetation. With the coming of spring, fresh shoots develop and plants flower in mid- to late-summer before dying back with the onset of autumn frosts. Disease epidemics reach their peak in late summer, and the pathogen overwinters as dormant (clonally reproducing) uredial infections on occasional green shoots (Jarosz & Burdon 1991; Barrett et al. 2007). In the plains populations, very hot dry summers and mild winters result in a virtual reversal of the patterns observed in the mountains. Epidemics start in the autumn and reach their peak in early spring. In the plains populations, the pathogen survives the summer drought as dormant teliospores (specialized resting spores which are a necessary precursor to the sexual cycle) which are produced in large quantities as shoots senesce (Barrett et al. 2007).

Sample collection

Rust samples were collected from mountains populations in the summer of 2004/2005, and from plains populations in the spring of 2005 (mid-epidemic in both regions). At each site, 21–33 single pustule-derived individuals were recovered (Table 1) from 40 samples haphazardly collected from different infected plants by rubbing cotton buds across sporulating uredia. In the laboratory, urediospores were inoculated onto the universally susceptible Linum usitatissimum cultivar Hoshangabad. Inoculated plants were left in a humid atmosphere overnight before being transferred to a glasshouse. Approximately 1 week later, single pustules were isolated and put through up to three cycles of increase on Hoshangabad, to ensure that each isolate consisted of a single genotype, and that sufficient urediospores were available for DNA extraction and pathotype analysis.

Table 1
Lineage composition, AFLP and Avr multilocus genotypic diversity and virulence phenotypic diversity for plains and mountains populations of Melampsora lini

Pathotype identification

We assessed the virulence of each of 275 isolates using the ‘standard differential set’ of 11 lines of L. marginale that has been used extensively in previous work assessing variation in virulence within M. lini populations (Burdon & Jarosz 1991; Jarosz & Burdon 1991; Burdon & Roberts 1995; Thrall et al. 2001b). Pathogenicity reactions were scored after 14 days, with the interaction between each pathogen isolate and individual differential host lines being classified as either virulent (1) or avirulent (0).

Individual pathotypes were thus defined by their unique combination of infectivity responses across the L. marginale differentials. The 10 pathogen populations were then analysed both in terms of the frequency and distribution of individual pathotypes.

DNA extraction

Prior to DNA extraction, approximately 100 mg of urediospores contained within an Eppendorf tube were allowed to germinate overnight in 100 μL of sterile water. Total genomic DNA was extracted from these urediniospores using a DNeasy plant mini kit (QIAGEN), following standard protocols.

AFLP analyses

Although microsatellite markers have been developed for M. lini, they proved too invariant for population genetic studies (Barrett & Brubaker 2006; Barrett et al. 2007). Here, we use an AFLP method adopted from Becerra Lopez-Lavalle & Brubaker (2007) with modifications as follows. We used eight EcoRI/MseI primer combinations in the selective amplification (GC/GT; ACC/GT; ACT/GG; AGA/GG; GC/GA; ACC/GC; ACT/GC; AGA/GT) to generate markers ranging from 50 to 500 bp in size. In each case, the EcoRI selective primer was labelled with one of the fluorescent dyes FAM, VIC, PET or NED. Selective polymerase chain reaction (PCR) products were separated on an ABI 3130 automated sequencer (PE Applied Biosystems) with a GeneScan LIZ 500 internal size standard. Electropherograms were subsequently analysed using GENEMAPPER version 4.0 (PE Applied Biosystems). The intensity of each individual peak was normalized on the basis of the total signal intensity and the peak was considered only if its intensity exceeded a fixed threshold. To test the repeatability of AFLP results, 10 individuals from each sample were completely replicated; no differences between the two sets of samples were observed. Only samples that could be scored unambiguously by eye were retained for further analysis. The presence and absence of AFLP fragments in each sample was analysed using the software GENEMAPPER. Fragments obtained using each primer combination were scored as either present or absent for each size-specific locus. AFLP fragments were treated as dominant marker loci with two states, presence (1) or absence (0).

AvrP123 and AvrP4 amplification and sequencing

The avirulence genes AvrP123 and AvrP4 were initially cloned and characterized in the interaction between M. lini and the cultivated flax species L. usitatissimum (Catanzariti et al. 2006), where multiple virulence polymorphisms are maintained via amino acid differences in the expressed Avr proteins (Dodds et al. 2006). AvrP4 encodes a 95 amino acid protein with a predicted 28 amino acid cleavable secretion signal peptide and AvrP123 encodes a 117 amino acid protein with a predicted 23 amino acid cleavable secretion signal peptide. For the interaction between M. lini and L. marginale, Barrett et al. (under review) demonstrated that both AvrP123 and AvrP4 are also recognized by L. marginale host plants, and that strong diversifying selection is acting at both these loci in Australian populations. This parallel study further demonstrated that lineage AB isolates are consistently heterozygous at these loci, and that haplotype diversity in lineage AB isolates at these loci is confined to alleles from the ‘A’ genome. One of these alleles (the ‘A’ allele), is very closely related to alleles recovered from lineage AA isolates. The second allele (the ‘B’ allele), which is highly divergent, displayed no variation.

Due to the lack of variation in ‘B type’ alleles, primers and PCR protocols were designed to amplify only the ‘A type’ alleles in lineage AB isolates (Barrett et al. under review). For AvrP4, we amplified a region including 180 bp of 5′ flanking sequence, 285 bp of open reading frame, and 103 bp of 3′ flanking sequence (F-CATCAAAATCTAACCCGTAC and R-TTGTTCAGGATAGATAGTGC). PCR amplifications were performed on a Hybaid Express thermocycler under the following conditions: 95 °C for 3 min, 34 cycles at 94 °C for 30 s, 56 °C for 45 s, 72 °C for 90 s, followed by a 4 °C holding step. For AvrP123 we amplified a region including 119 bp of 5′ flanking sequence, 351 bp of open reading frame, and 128 bp of 3′ flanking sequence (F-ATTGTGAAC-CTTTTGAAGGAC and R-CGCCATGGTATTGTTCAGAC). PCR amplifications were performed as for AvrP4 except for a 58 °C annealing temperature. PCR products were cleaned using Multiscreen PCR96 filter plates (Millipore) and sequenced directly using the forward primers in conjunction with BigDye Terminator Cycle Sequencing Reaction Kit (PE Applied Biosystems). Sequencing products were resolved on an ABI 3130 automated sequencer. To assist with alignment of different sequences, alignments were initially performed on the amino acid sequence, using conserved cystine motifs as reference points and then back-translated to the original nucleotides. We analysed Avr sequence variation by classifying amino acid variants as allelic variants. As noted above, all analyses based on data collected for lineage AB isolates are based on the ‘A’ haplotypes only.

Data analyses

We conducted analyses at both the between- and within-lineage levels, using different approaches. Analyses of genotypic diversity across different levels, and overall genetic differentiation [i.e. analyses of molecular variance (AMOVA) and Nei’s genetic distance] were performed using all isolates. For comparative analyses of spatial genetic structure [i.e. θB and analysis of similarity (ANOSIM) comparisons] we treated the two lineages separately. Due to low sample size, for these analyses lineage AB isolates collected from the plains populations and the single lineage AA isolate collected from the mountains populations were excluded.

Genotypic diversity among lineages and populations was calculated based on the number of shared multilocus AFLP or Avr genotypes. Phenotypic diversity was calculated based on the shared number of virulence phenotypes. The genotypic or phenotypic identity of the different isolates was evaluated using the software GENOTYPE (Meirmans & Van Tienderen 2004). Genotypic diversity for lineages and populations was estimated via the number of unique multilocus genotypes, Nei’s diversity index (Nei 1987) (calculated using GENODIVE; Meirmans & van Tienderen 2004). A bootstrap resampling method as implemented in GENODIVE was used to test whether these indices were significantly different among lineages.

For all other lineage and population-level genetic analyses, we used a clone-corrected data set. This approach was taken so as to minimize the effects of clonal reproduction during epidemics on estimates of genetic diversity and population differentiation. We identified clones based on variation between isolates in AFLPs, virulence phenotypes, and avirulence gene genotypes. AFLP genetic diversity was evaluated as the percentage of polymorphic loci.

To test for random associations among AFLP markers, multilocus linkage disequilibrium was measured by calculating the Index of Association (IA; Maynard Smith et al. 1993), using the software MULTILOCUS version 2.2 (Agapow & Burt 2001). The observed IA for each population and each lineage was compared with an expected IA under random mating simulated through the reshuffling of data over 1000 permutations. IA has an expected value of zero if there is no association of alleles at unlinked loci as expected in a randomly mating population.

The degree of genetic relatedness (AFLP data) among all pairs of populations was evaluated using unweighted pair-group mean analysis (UPGMA) based on estimates of Nei’s genetic distance (1972) calculated using GENALEX version 6.0 (Peakall & Smouse 2005). The UPGMA tree was constructed using NT-SYSPC version 2.11 (Rohlf 1993). The genetic similarity among isolates (AFLP) within lineages was visualized using principal coordinate analysis (PCoA) based on Euclidean distances between AFLP multilocus genotypes using GENALEX 6.0.

To examine hierarchical partitioning of molecular variation among lineages and populations, we subjected AFLP to AMOVA using GENALEX 6.0. To test for genetic structure among populations within lineages, we used a Bayesian approach to estimate an FST analogue among populations within each lineage (denoted θB) for the AFLP data, and the two Avr loci, as implemented in the software Hickory (Holsinger et al. 2002). The data were run with the default parameters (burn-in = 50 000, number of samples = 250 000, thinning factor = 50). The data were analysed using the f-free model, which does not assume any prior knowledge regarding the degree of inbreeding within populations. Comparisons of θB between data sets were performed by calculating the difference between paired random samples of the posterior distribution of θB for each data set, and declaring the difference in θB significant if the 95% credible interval of this difference excluded zero. This approach allowed us to estimate θB for both AFLP and Avr gene loci, and to formally test hypotheses regarding variation in genetic structure among neutral and selected markers. For comparative purposes, we also estimated genetic structure within lineages with AMOVA, using the software GENALEX 6.0.

To estimate levels of geographical structure in the virulence phenotypic data, we used ANOSIM (Clarke 1993). We used this approach because the nature of the phenotypic virulence data precluded more conventional approaches to measuring population divergence (e.g. FST or QST, or their analogues). Each virulence phenotype is under direct genetic control, and is controlled by a limited but undetermined number of genes, each with major effects. Each individual response on each differential line is thus informative, but cannot be considered as an independent marker, because the phenotype is determined by multiple genes. At the same time, the data cannot be considered as a continuously distributed quantitative trait, because each phenotype is controlled by genes with major effects, and is quantified in a binary manner. Similar to AMOVA, ANOSIM can be used to assess similarity among individuals when grouped at different hierarchical levels. The method does not make assumptions about the nature of the data, and uses rank order of dissimilarity values among individual samples as its base. Under the null hypothesis of no differentiation, the test statistic, R, changes little when population labels are rearranged randomly. R values range between 1 (maximum separation among populations) and 0 (completely random grouping) (Clarke 1993). The associated test for significance was calculated over 1000 permutations. For comparative purposes, we also performed these analyses using the AFLP data set. ANOSIM was performed using the software PRIMER version 6 (PRIMER-E Ltd, Plymouth Marine Laboratory, Plymouth, UK).

To explore how much AFLP genetic variation between populations (within a lineage) was explained by geographical distance, matrices of pairwise genetic distance (Nei 1972) and geographical distance were subject to a Mantel test (1967) using GENALEX 6.0 (Peakall & Smouse 2005) with 10000 permutations.


Patterns of diversity among isolates

We recovered AFLP multilocus genotypes, virulence phenotypes and genotypes at the two Avr loci for 275 isolates of Melampsora lini, sampled from 10 natural populations across two geographical regions. Among the mountains populations, we recovered exclusively lineage AB isolates, with the exception of one lineage AA isolate in population N2. Among the plains populations, we recovered isolates from both lineages in all populations, although isolates belonging to lineage AA were predominant (population average: 87%; Table 1).

For the AFLP data set, a total of 208 AFLP markers were scored and retained for further analysis. Of these, 88 markers were polymorphic across the data set. Of the 208 fragments, 60 were unique to lineage AB, and four were unique to lineage AA. In total, 124 isolates were assigned to lineage AA, and 151 isolates to lineage AB. Genetic diversity as determined by AFLP was generally low, with only 12% and 14% of all scored AFLP markers being polymorphic within lineages AA and AB, respectively.

Pathogenicity analyses revealed extensive variation across all isolates, with 62 different virulence patterns detected across the 11 host differential lines. The number of host lines overcome by individual isolates ranged between 3 and 10. Across all isolates, we recovered 4 and 5 clade ‘A’ amino acid variants from AvrP123 and AvrP4, respectively. The sequence data for each allele is deposited in GenBank (accession numbers for AvrP123: EU642492, EU642493, EU642494, EU642495; for AvrP4: EU642476, EU642478, EU642479, EU642482, EU642483). Isolates with identical AFLP multilocus genotypes frequently represented multiple virulence phenotypes, and isolates with the same virulence phenotype had different AFLP genotypes. Similarly, identical Avr genotypes harboured multiple virulence phenotypes, and vice versa (Table S1, Supplementary material).

Variation between lineages

The two lineages differed significantly in levels of genotypic diversity as determined by AFLP markers, Avr genes, and virulence phenotypes. In all three cases, lineage AA harboured significantly (P < 0.01) more genotypic variation than lineage AB (Table 2). Combining all of these marker systems, the total number of genotypes in lineage AA was 100 (of 124 isolates), compared to 61 (of 151) in lineage AB. For lineage AB, 18 of these isolates were recovered from plains populations, 16 of which were genotypically unique.

Table 2
Summary statistics describing genotypic and phenotypic variation in 124 and 151 isolates of Melampsora lini belonging to lineage AA and lineage AB, respectively

Overall, AMOVA revealed significant and high levels of structure in the data set (ΦPT = 0.94, P = 0.001), with 90% of the AFLP diversity attributed to differences among lineages, with the remainder among isolates within populations (6%), or between populations of the same lineage (4%). Consistent with these observations, the UPGMA phylogram based on the genetic distance among populations showed very strong separation between plains and mountains populations (Fig. 2).

Fig. 2
UPGMA phylogram based on pairwise estimates of Nei’s genetic distance (AFLP) among populations of Melampsora lini.

Principal coordinate analysis of the Euclidean distance among lineage AA and AB isolates highlights clearly contrasting AFLP genetic structure within the two lineages (Fig. 3). Lineage AA isolates, while showing structure at the population level, form a relatively continuous cloud of points across axes 1 and 2 (explaining 36.5% and 20.1% of the variation, respectively). In contrast, lineage AB isolates form a series of four relatively discrete clusters along PCoA axis 1 (axes 1 and 2 explain 56.1% and 17.1% of all variation, respectively).

Fig. 3
Principal coordinates plots based on Euclidean distances among multilocus AFLP genotypes for all clone corrected isolates from plains and mountains populations. (a) Lineage AA isolates (n = 100), axes explain 56.6% (36.5 + 20.1) of the variation. (b) ...

The IA statistic revealed significant levels of nonrandom association among AFLP markers for both lineages AA and AB. However, while significant for both lineages, IA was more than twice as high in lineage AB as in lineage AA. In conjunction with previous results showing fixed heterozygosity at microsatellite and Avr loci (Barrett et al. 2007), these elevated levels of linkage disequilibrium in lineage AB further support the suggestion that this lineage only reproduces clonally. Significant levels of IA and the Bayesian estimate of the fixation index (f = 0.7) for lineage AA suggest frequent inbreeding within this lineage. The frequency distribution of the different AFLP multilocus genotypes within lineage AB further suggest that clonal reproduction is dominant in this lineage, with two distinct AFLP genotypes representing more than 70% of all sampled isolates (Fig. 4b). In contrast, while identical AFLP genotypes were present within the lineage AA samples, no single genotype reached frequencies of higher than 10% across populations (Fig. 4a).

Fig. 4
Histograms of frequencies of different AFLP multilocus genotypes including more than one individual out of a total of 275 isolates of Melampsora lini. Lineage AA and AB isolates had no multilocus genotypes in common. (a) Lineage AA isolates. A total of ...

Similarly, virulence phenotypes show a distinctly uneven frequency distribution in lineage AB compared to lineage AA, and the two lineages also displayed distinctly different virulence structures in terms of the identity and frequency of individual pathotypes (Fig. 5a). The specific differences among the two lineages in their ability to overcome different resistance genes are indicated by striking differences in the frequency with which pathogen isolates from the two lineages were able to overcome individual host differential lines (Fig. 5b).

Fig. 5
Virulence profile of lineage AA and AB isolates. (a) Histograms of frequencies of different virulence phenotypes (including more than one individual) among 124 lineage AA and 151 lineage AB isolates of Melampsora lini as determined on 11 differential ...

Variation among populations

Within each lineage, among-population pathogen diversity as determined by all three marker systems varied widely (Table 1). Of the mountain populations (lineage AB), Kiandra was notable for having especially low levels of AFLP and virulence diversity, being dominated by a single clone corrected genotype. In contrast, diversities as indicated by virulence phenotypic diversity and Avr markers within population N2 were equal, or higher, than in several plains (lineage AA) populations. Within lineage AA, low levels of diversity were observed in the Larras Lee and Garra populations compared to other plains populations (Table 1).

Within lineages, estimates of divergence among populations revealed consistent patterns of genetic and phenotypic differentiation. Bayesian estimates of θB based on AFLP and avirulence gene data resulted in much lower deviance information criteria values for models allowing population structure than for models allowing no population structure (Table 3), providing evidence for population differentiation across all markers. For the AFLP data, estimates of population structure were high for both lineages (θB = 0.336 for AA, θB = 0.367 for AB). Strong population structure within lineages was also revealed for AvrP4, although the genetic structure among lineage AB populations was weaker (θB = 0.325 for AA, θB = 0.186 for AB). In contrast, estimates of population structure for AvrP123 were much lower, although still significant, for both lineages (θB = 0.112 for AA, θB = 0.046 for AB). Posterior comparisons of θB revealed significantly lower levels of population differentiation for Avr loci compared to AFLP data, with the exception of the estimate of θB for AvrP4 among lineage AA populations, which was not significantly different from estimates of population differentiation for AFLP markers (θB = 0.325 and 0.336, respectively) (Table 3). Analysis of structure in the phenotypic virulence data using analysis of similarity also revealed significant spatial structure within both lineages (lineage AA: R = 0.27; lineage AB: R = 0.12), indicating moderate and low levels of separation among populations, respectively. Within lineages, Mantel tests revealed no significant association between the degree of population differentiation as estimated by AFLP markers(ΦPT) and geographical distance (km) (lineage AA, P = 0.08; lineage AB, P = 0.08).

Table 3
Estimates of population structure (θB), standard deviations (SD) and credible intervals (CRI) from Bayesian modelling for AFLP fingerprints and two avirulence loci for two genetic lineages of Melampsora lini. Deviance information criteria calculations ...

Relationships between patterns of phenotypic and genetic divergence among populations were not concordant among lineages. For lineage AA, pairwise comparisons of among-population differentiation as determined by AFLP multilocus genotypes (ΦPT), Avr gene genotypes (ΦPT), and virulence phenotypes (R) were not significantly correlated (Mantel tests of matrix comparisons, P > 0.05). In contrast, for clonal lineage AB, comparisons of pairwise differentiation among the different markers using Mantel tests revealed a number of statistically significant relationships (for AFLP vs. virulence phenotypes, P = 0.01, R = 0.79; for AFLP vs. Avr multilocus genotypes, P = 0.01, R = 0.93; for virulence phenotypes vs. Avr multilocus genotypes P = 0.02, R = 0.8). AvrP123 was significantly correlated to both AFLP genotypes (P = 0.04, R = 0.88) and virulence phenotypes (P = 0.03, R = 0.75). There were no significant relationships between AvrP4 and either AFLP genotypes or virulence phenotypes (P > 0.3).


The geographical mosaic theory of co-evolution predicts that traits influencing the outcome of co-evolutionary interactions will differ among populations, thereby generating a selection mosaic that varies across landscapes (Thompson 1994, 2005). In this context, our results have important implications for understanding how spatial, environmental and life-history variation influence the population and regional genetic structure of Melampsora lini. Through the integration of AFLP, Avr gene and phenotypic virulence data, we demonstrate strong genetic and pathogenic divisions among populations of M. lini within two biogeographical regions, driven by key differences between pathogen lineages in genetic composition and mode of reproduction. Within this broad genetic and geographical framework, selection and genetic drift interact to generate genetic and pathogenic divergence among local populations.

The geographical distribution of lineage AA and AB

Analysis of molecular and phenotypic virulence data at the regional scale confirm the strong genetic separation between the M. lini lineages AA and AB previously suggested by geographically broad sampling (Barrett et al. 2007). This pattern and the dominance of lineage AA isolates in the plains environment is likely a consequence of life-history differences associated with the hybrid origin of lineage AB. The loss or lack of the sexual mode of reproduction effectively restricts lineage AB isolates to the clonal (urediospore) stage of the lifecycle (Barrett et al. 2007). In the hotter, drier, plains populations, the regular occurrence of summer drought results in the widespread death or senescence of the host, Linum marginale. Without green host tissue, M. lini is unable to survive in the uredial stage, meaning that telial formation is critical for long-term persistence. A severe summer drought in the year prior to sampling during which isolates in three monitored plains populations survived exclusively as telia, suggests that lineage AB isolates recovered in plains populations are likely to have dispersed from more mesic sites during that growing season. In some years, summer rains may result in year-round persistence of green host tissue, reducing the likelihood of local extinction of lineage AB isolates in plains populations.

In contrast to the plains environment, host phenology is essentially reversed in the mountains region, and the striking dominance of lineage AB isolates in the mountains populations is more difficult to explain. The presence of only a single lineage AA isolate in the five mountains populations implies strong barriers to the establishment of this lineage in this environment. Given that mountains hosts are not universally or even unusually resistant to lineage AA isolates (L. G. Barrett, CSIRO Plant Industry, unpublished manuscript), this distribution is unlikely to be driven by among-region differences in host resistance. Instead, it is possible that an increased tendency to form telia during warm weather, while crucial to survival in the plains, may affect other fitness parameters relating to disease spread such as spore production and transmission. Consistent with this idea, glasshouse experiments demonstrate that lineage AB isolates are significantly less likely to form telia under warm conditions than lineage AA isolates (Barrett et al. 2007). Lineage AB isolates may therefore be better at exploiting host plants during the warm summer months, which coincides with the main growing season for L. marginale in the mountains.

Genetic and pathogenic structure within lineages

It is generally expected that populations able to reproduce sexually will generate and maintain higher genotypic diversity than asexual populations. Indeed, a number of studies show greater phenotypic and genetic diversity in sexual compared to asexual populations of plant pathogens (Milgroom 1996). Here, we demonstrate that levels of genotypic diversity within lineages of M. lini are consistent with such expectations, with lineage AA harbouring significantly higher AFLP, Avr, and virulence diversity than lineage AB despite census populations being generally smaller and more isolated in the plains region where AA predominates (Burdon et al. 1999). The two lineages also differ strikingly in how variation is distributed among isolates. Under strict asexual reproduction, founding genotypes should diversify over time into discrete clonal lineages characterized by a complex of closely related genotypes. In contrast, the exchange of genetic information among individuals in sexually reproducing populations should result in a more homogenous population structure. Consistent with this expectation, PCoA of the AFLP data showed that genomic variation is distributed relatively evenly across isolates for lineage AA, while lineage AB isolates cluster into a few discrete genotypic groups.

For lineage AA isolates in the plains populations, the high diversity of genotypes in each population suggests that sexual reproduction is common, despite the strong potential for clonal multiplication during epidemics. Only at Larras Lee, and to a lesser extent, Garra, did we find evidence for clonal dynamics strongly influencing population structure. Generally, high genotypic diversity was maintained in plains populations despite evidence for significant levels of inbreeding. The low number of identical clones in these populations may reflect a high number of sexual recombinants as primary inoculum, or selection for different virulence phenotypes by individual host plants. While all populations were sampled during established phases of infection, these results represent only a snapshot of pathogen population structure during a single epidemic. Clearly, the clonal structure of populations may change throughout the growing season, with individual pathotypes having the potential to rapidly increase in frequency as epidemics progress (Jarosz & Burdon 1991; Burdon & Jarosz 1992).

Although the population structure detected for the mountains (lineage AB) populations was consistent with that expected under clonal reproduction, within-population diversity varied markedly, with some populations maintaining unexpectedly high levels of variation. For example, in population N2, 21 of the 28 isolates sampled were genetically unique, with AFLP and virulence diversity being higher than some of the sexually reproducing lineage AA populations. In contrast, we recovered only six genetically unique individuals from 33 sampled isolates in Kiandra, reflecting dominance by a single pathotype. Interestingly, this pathotype is also the most common pathotype across all mountains populations and has dominated Kiandra for many years (Burdon & Jarosz 1991; Thrall et al. 2001b; J. J. Burdon, CSIRO Plant Industry, unpublished data). Thus, while patterns of diversity in some populations suggest strong selection for variation in virulence, low levels of diversity in others agree with the general-purpose-genotype hypothesis (Lynch 1984) that predicts selection of a few asexual genotypes characterized by broad tolerances for fluctuating environmental or biotic conditions.

Although difficult to evaluate, given that the putative BB parent has not been detected, the presence of multiple shared Avr gene alleles between AA and AB isolates suggests that either multiple hybridization events, multiple origin of the same alleles, or somatic exchange of genetic information, must have occurred at some stage. Genetically discrete isolates within clusters of closely related AFLP genotypes, and isolates with identical AFLP multilocus genotypes, but different virulence phenotypes, are most likely to have been generated via stepwise patterns of mutation, following establishment of a founding genotype. Stepwise mutation has been demonstrated as an important mechanism for generating diversity in populations of asexual pathogen species (Jimenez-Gasco et al. 2004; Fisher et al. 2005; Hovmøller & Justetson 2007); however, horizontal gene exchange may also be important in driving divergence between otherwise very closely related isolates. Thus, although lineage AB Avr genotypes and AFLP multilocus genotypes on average are strongly linked, there were several examples where lineage AB isolates with identical AFLP genotypes carried different Avr gene alleles. While the generation of these genotypes via multiple hybridization events between identical AFLP genotypes cannot be discounted, horizontal gene exchange via hyphal fusion among genetically different individuals seems equally parsimonious. The contribution of such non-sexual genetic exchange to generating pathogen diversity has been clearly demonstrated in several species (Spiers & Hopcroft 1994; Burdon & Silk 1997; Friesen et al. 2006).

Higher overall diversity in plains compared to mountains populations could significantly impact on host–pathogen co-evolutionary interactions at the regional scale. As suggested by the Red Queen Hypothesis, where hosts and pathogens are naturally coevolving, high levels of diversity in pathogen populations may promote higher diversity through sexual reproduction in corresponding host populations (Haldane 1949). Thus, the results presented here are particularly intriguing when interpreted in light of host resistance variation and mating system in plains and mountains environments (Burdon et al. 1999; Thrall et al. 2001b). In the mountains populations (lineage AB), hosts are almost completely selfing. In contrast, in the plains environment, where the majority of isolates reproduce sexually, hosts show significant levels of outcrossing. These differences in host mating system are reflected in significant variation among regions in the diversity of resistance phenotypes, with host populations in the plains region showing consistently higher values. Furthermore, strong differentiation between pathogen lineages in virulence responses to individual host differential lines, and in the identity of associated virulence phenotypes between lineages, corresponds to the observed divergence between mountains and plains host populations in the identity of resistance phenotypes (Burdon et al. 1999).

Genetic structuring among populations

Analysis of the AFLP data demonstrates relatively strong genetic divergence between populations of both lineages. This structure is maintained despite the fact that M. lini produces large numbers of urediospores that have the potential to disperse aerially over long distances. Indeed, the presence of lineage AB isolates in all plains populations at low frequencies, and the lack of evidence for any isolation by distance effects, suggests that long-distance spore dispersal is both stochastic and relatively frequent. Despite this, levels of among-population genetic differentiation are considerably higher than those reported for many wind-dispersed, foliar plant pathogens of cultivated crops, such as Mycosphaeralla graminicola (GST = 0.05; Linde et al. 2002), Phaeosphaeria nodorum (θ = 0.05; Pimentel et al. 2000), and Tapesia yallundae (θ = −0.008, Douhan et al. 2002). It is important to note that these agricultural examples largely represent pathogens of high-density crops, which can survive saprophytically to varying degrees.

Strong bottlenecks during epidemic troughs and the stochastic nature of recolonization processes are likely to be key factors driving divergence among populations (Brown & Hovmoller 2002). Given that M. lini is a biotrophic foliar pathogen, and occurs in small, fragmented and genetically diverse host populations, these kinds of stochastic demographic dynamics are common, with annual boom-and-bust epidemic dynamics likely promoting frequent local extinction, stochastic recolonization and subsequent genetic drift (Wade & McCauley 1988; Burdon & Jarosz 1992;Ingvarson et al. 1997; Pannell & Charlesworth 1999; Haag et al. 2006). Heterogeneity in terms of host availability determined by resistance structure and environmental conditions further limit opportunities for the establishment and persistence of pathogen migrants which are likely to be accentuated by clonal reproduction (lineage AB), and inbreeding during the sexual phase of the life-cycle (lineage AA).

Previous studies (largely within the mountains) suggest that selection for virulence in the L. marginaleM. lini interaction can strongly influence local patterns of genetic variation in M. lini. Considerable variation for host resistance exists within and among host populations (Burdon & Jarosz 1991; Jarosz & Burdon 1991) and comprehensive cross inoculation trials have demonstrated strong local adaptation of the pathogen to host populations (Thrall et al. 2002). Furthermore, pathogen virulence and host resistance are highly correlated, such that broadly virulent pathogens occurred more frequently in highly resistant host populations, whereas avirulent pathogens dominate susceptible host populations (Thrall & Burdon 2003). Consistent with expectations arising from these observations, we found significant structuring between populations, in both lineage AA and AB, for virulence phenotypes, and both Avr gene loci.

Direct comparisons of population structure revealed through phenotypic characters or genes under selection with that shown by neutral genetic markers can provide insight into the relative roles of drift and selection in driving genetic divergence among populations (Merilä & Crnokrak 2001). However, comparisons between different marker systems used to make inferences about local adaptation require that markers are comparable and unbiased. One concern is that the virulence phenotypes, although under direct genetic control, may underestimate the true level of divergence among populations relative to AFLP markers, particularly if either homoplasy or epistasis are common. Furthermore, the use of a standard set of host differential lines may also not provide a complete estimate of the diversity of pathogen populations when infecting their local hosts. Despite these caveats, our AFLP estimates of genetic structure provide a useful baseline against which to compare the structure of variation in phenotypic virulence characters, and Avr gene allele frequencies.

Most studies comparing phenotypic traits (Merilä & Crnokrak 2001; McKay & Latta 2002) or genes under selection (Conway et al. 2001; Abdel-Muhsin et al. 2003; Anderson et al. 2005) with neutral markers demonstrate higher levels of differentiation among populations for phenotypic traits than for neutral marker genes. Such findings are generally interpreted as evidence for local selection driving the divergence of specific traits or genes. In contrast, in the present study, data from two Avr loci and for virulence phenotypes showed levels of spatial genetic structure either significantly less, or similar to, structure determined by the AFLP markers. Both AvrP123 ( θAAB=0.05,θABB=0.11) and AvrP4 ( θAAB=0.32,θABB=0.19) are significantly differentiated among populations; however, with the exception of estimates of θB for AvrP4 for lineage AA, these values are significantly lower than estimates for the AFLP markers. Similarly, global R statistics describing divergence among populations for AFLP markers and virulence phenotypes indicate that separation among populations was more than twice as strong for AFLP than for virulence markers.

The strong population genetic structure in the AFLP data compared to the Avr genes and virulence phenotypes could be a consequence of selection for particular virulence genes or phenotypes acting to constrain among-population divergence in virulence structure (i.e. selection for the same virulences in different populations). Thus, it is possible that selection imposed by host resistance structure is preventing population divergence beyond that seen for AFLP markers. Given the potential for very large annual fluctuations in demographic population size in M. lini, such a scenario is not implausible. However, even where selection is acting to limit divergence among populations, drift may also have the potential to influence the pathogenic identity of local populations, particularly where individual virulence phenotypes of Avr gene alleles might be under only weak selection, or selectively neutral in local host populations (Salathé et al. 2005).

To evaluate the relative influence of drift vs. local selection in generating divergence among populations, we examined pairwise patterns of divergence among populations for the different genetic and phenotypic characters. In the sexual lineage AA, we found no significant relationships among populations for the degree of divergence as determined by any of the various genetic and phenotypic characters. These results demonstrate the potential for different evolutionary forces to act on these markers independently, and suggest that local patterns of selection contribute to phenotypic and Avr gene divergence among lineage AA populations. In contrast, for lineage AB, we found highly significant correlations between all phenotypes and genotypes at the population level. However, comparisons among populations for lineage AB are almost certainly complicated by a lack of sexual reproduction. Given that very strong linkages among all genetic elements in lineage AB are likely, unravelling the relative roles of drift and selection is highly problematic.


The pattern of variation in genetic and pathogenic structure of M. lini provides strong evidence that geographical structure across both regional and local scales is an important component in the generation and maintenance of virulence diversity in this system. The spatial distribution of the variation is consistent with evolutionary dynamics acting to structure diversity at two distinct spatial scales. At the regional scale, patterns of environmental heterogeneity interact with host and pathogen life-history traits to create two broad areas (mountains vs. plains) with distinct genetic and demographic dynamics. Importantly, correlated patterns of pathogen virulence diversity and host resistance diversity (Burdon et al. 1999) indicate contrasting rates of co-evolutionary response among the two regions, suggesting the possibility of a scenario where distinct co-evolutionary hotspots and coldspots (Thompson 2005) are generated through corresponding modes of host and pathogen reproduction. Within these regions local metapopulation processes act to structure the distribution of variation, driven by local patterns of selection and neutral genetic drift. A major goal for future studies will be to quantify the influence of these patterns on disease epidemiology, host mating system and host resistance.

Supplementary Material


Supplementary material:

The following supplementary material is available for this article:

Table S1. AFLP multilocus genotypes, avirulence gene genotypes* and virulence phenotypes for isolates of Melampsora lini belonging to lineage AA or AB, collected from 10 populations across two regions. New AFLP mulitilocus genotypes and virulence phenotypes were assigned each time a unique fingerprint was encountered in the order listed below. *For lineage AB isolates we report only the ‘A’ allele. All lineage AB isolates are assumed heterozygous for a second invariant ‘B’ allele

This material is available as part of the online article from: http://www.blackwell-synergy.com/doi/abs/10.1111/j.1365-294X.2008.03843.x (This link will take you to the article abstract).

Please note: Blackwell Publishing are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.


We thank Caritta Eliasson for technical support, and Peter Dodds for providing assistance and expertise with the Avr gene component of this study. Three anonymous reviewers provided numerous constructive comments that significantly improved the quality of this manuscript.


  • Abdel-Muhsin AA, Mackinnon MJ, Awadalla P, et al. Local differentiation in Plasmodium falciparum drug resistance genes in Sudan. Parasitology. 2003;126:391–400. [PubMed]
  • Agapow PM, Burt A. Indices of multilocus linkage disequilibrium. Molecular Ecology Notes. 2001;1:101–102.
  • Alexander HM, Antonovics J, Kelly AW. Genotypic variation in plant disease resistance — physiological resistance in relation to field disease transmission. Journal of Ecology. 1993;81:325–333.
  • Anderson RM, May RM. Coevolution of hosts and parasites. Parasitology. 1982;85:411–426. [PubMed]
  • Anderson TJC, Nair S, Sudimack D, et al. Geographical distribution of selected and putatively neutral SNPs in Southeast Asian malaria parasites. Molecular Biology and Evolution. 2005;22:2362–2374. [PubMed]
  • Antonovics J, Thrall PH, Jarosz AM, Stratton D. Ecological genetics of metapopulations: the Silene-Ustilago plant–pathogen system. In: Real L, editor. Ecological Genetics. Princeton University Press; Princeton, New Jersey: 1994. pp. 146–170.
  • Barrett LG, Brubaker CL. Isolation and characterization of microsatellite loci from the rust pathogen, Melampsora lini. Molecular Ecology Notes. 2006;6:930–932.
  • Barrett LG, Thrall PH, Burdon JJ. Evolutionary diversification through hybridization in a wild host–pathogen interaction. Evolution. 2007;61:1613–1621. [PubMed]
  • Becerra Lopez-Lavalle A, Brubaker CL. Frequency and fidelity of alien chromosome transmission in Gossypium hexaploid bridging populations. Genome. 2007;50:479–491. [PubMed]
  • Bevan JR, Crute IR, Clarke DD. Resistance to Erisyphe fischeri in two populations of Senecio vulgaris. Plant Pathology. 1993;42:636–646.
  • Bishop SC, Morris CA. Genetics of disease resistance in sheep and goats. Small Ruminant Research. 2007;70:48–59.
  • Brown JKM, Hovmoller MS. Aerial dispersal of pathogens on the global and continental scales and its impact on plant disease. Science. 2002;297:537–541. [PubMed]
  • Burdon JJ. Host population subdivision and the genetic structure of natural pathogen populations. Advances in Plant Pathology. 1992;8:81–94.
  • Burdon JJ. The distribution and origin of genes for race-specific resistance to Melampsora lini in Linum marginale. Evolution. 1994;48:1564–1575.
  • Burdon JJ, Jarosz AM. Host–pathogen interactions in natural populations of Linum marginale and Melampsora lini. I. Patterns of resistance and racial variation in a large host population. Evolution. 1991;45:205–217.
  • Burdon JJ, Jarosz AM. Temporal variation in the racial structure of flax rust (Melampsora lini) populations growing on natural stands of wild flax (Linum marginale): local versus metapopulation dynamics. Plant Pathology. 1992;41:165–179.
  • Burdon JJ, Roberts JK. The population genetic structure of the rust fungus Melampsora lini as revealed by pathogenicity, isozyme and RFLP markers. Plant Pathology. 1995;44:270–278.
  • Burdon JJ, Silk J. Sources and patterns of diversity in plant pathogenic fungi. Phytopathology. 1997;87:664–669. [PubMed]
  • Burdon JJ, Thompson JN. Changed patterns of resistance in a population of Linum marginale attacked by the rust pathogen Melampsora lini. Journal of Ecology. 1995;83:199–206.
  • Burdon JJ, Thrall PH, Brown AHD. Resistance and virulence structure in two Linum marginaleMelampsora lini host–pathogen metapopulations with different mating systems. Evolution. 1999;53:704–716.
  • Burdon JJ, Thrall PH, Lawrence GJ. Coevolutionary patterns in the Linum marginaleMelampsora lini association at a continental scale. Canadian Journal of Botany. 2002;80:288–296.
  • Catanzariti AM, Dodds PN, Lawrence GJ, Ayliffe MA, Ellis JG. Haustorially expressed secreted proteins from flax rust are highly enriched for avirulence elicitors. Plant Cell. 2006;18:243–256. [PMC free article] [PubMed]
  • Clarke B. The ecological genetics of host-parasite relationships. In: Taylor ER, Muller R, editors. Genetic Aspects of Host-Parasite Relationships. Blackwell Scientific; Oxford, UK: 1976. pp. 87–103.
  • Clarke KR. Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology. 1993;18:117–143.
  • Clay K, Kover PX. The Red Queen Hypothesis and plant/pathogen interactions. Annual Review of Phytopathology. 1996;34:29–50. [PubMed]
  • Conway DJ, Machado RLD, Singh B, et al. Extreme geographical fixation of variation in the Plasmodium falciparum gamete surface protein gene Pfs48/45 compared with microsatellite loci. Molecular and Biochemical Parasitology. 2001;115:145–156. [PubMed]
  • Cooke GS, Hill AVS. Genetics of susceptibitlity to human infectious disease. Nature Reviews Genetics. 2001;2:967–977. [PubMed]
  • Dodds PN, Lawrence GJ, Catanzariti AM, et al. Direct protein interaction underlies gene-for-gene specificity and coevolution of the flax resistance genes and flax rust avirulence genes. Proceedings of the National Academy of Sciences, USA. 2006;103:8888–8893. [PMC free article] [PubMed]
  • Douhan GW, Peever TL, Murray TD. Multilocus population structure of Tapesia yallundae in Washington State. Molecular Ecology. 2002;11:2229–2239. [PubMed]
  • Ericson L, Burdon JJ, Muller WJ. Spatial and temporal dynamics of epidemics of the rust fungus Uromyces valerianae on populations of its host Valeriana salina. Journal of Ecology. 1999;87:649–658.
  • Fisher MC, Hanage WP, de Hoog S, et al. Low effective dispersal of asexual genotypes in heterogeneous landscapes by the endemic pathogen Penicillium marneffei. PLoS Pathology. 2005;1:159–165. [PMC free article] [PubMed]
  • Fox JA, Dybdahl MF, Jokela J, Lively CM. Genetic structure of coexisting sexual and clonal subpopulations in a freshwater snail (Potamopyrgus antipodarum) Evolution. 1996;50:1541–1548.
  • Friesen TL, Stukenbrock EH, Liu ZH, et al. Emergence of a new disease as a result of interspecific virulence gene transfer. Nature Genetics. 2006;38:953–956. [PubMed]
  • Gillespie JH. Natural selection for resistance to epidemics. Ecology. 1975;56:493–495.
  • Greischar MA, Koskella B. A synthesis of experimental work on parasite local adaptation. Ecology Letters. 2007;10:418–434. [PubMed]
  • Haag CR, Riek M, Hottinger JW, Pajunen VI, Ebert D. Founder events as determinants of within-island and among-island genetic structure of Daphnia metapopulations. Heredity. 2006;96:150–158. [PubMed]
  • Haldane JBS. Disease and evolution. Ricerca Scientifica. 1949;19:68–76.
  • Hamilton WD. Sex versus non-sex versus parasite. Oikos. 1980;35:282–290.
  • Heitman J. Sexual reproduction and the evolution of microbial pathogens. Current Biology. 2006;16:R711–R725. [PubMed]
  • Holsinger KE, Lewis PO, Dey DK. A Bayesian approach to inferring population structure from dominant markers. Molecular Ecology. 2002;11:1157–1164. [PubMed]
  • Hovmøller MS, Justetson AF. Rates of evolution of avirulence phenotypes and DNA markers in a northwest European population of Puccinia striiformis f. sp tritici. Molecular Ecology. 2007;16:4637–4647. [PubMed]
  • Ingvarsson PK, Olsson K, Ericson L. Extinction-reconization dynamics in the mycophagous beetle Phalacrus substriatus. Evolution. 1997;51:187–195.
  • Jarosz AM, Burdon JJ. Host–pathogen interactions in natural populations of Linum marginale and Melampsora lini. II. Local and regional variation in patterns of resistance and racial structure. Evolution. 1991;45:1618–1627.
  • Jimenez-Gasco MD, Milgroom MG, Jimenez-Diaz RM. Stepwise evolution of races in Fusarium oxysporum f. sp. ciceris inferred from fingerprinting with repetitive DNA sequences. Phytopathology. 2004;94:228–235. [PubMed]
  • Keeling MJ, Rand DA. A spatial mechanism for the evolution and maintenance of sexual reproduction. Oikos. 1995;74:414–424.
  • Kohn LM. The clonal dynamic in wild and agricultural plant populations. Canadian Journal of Botany. 1995;73:S1231–S1240.
  • Laine AL, Hanski I. Large-scale spatial dynamics of a specialist plant pathogen in a fragmented landscape. Journal of Ecology. 2006;94:217–226.
  • Latta RG. Gene flow, adaptive population divergence and comparative population structure across loci. New Phytologist. 2004;161:51–58.
  • Lewontin RC, Krakauer J. Distribution of gene frequency as a test of the theory of the selective neutrality of polymorphisms. Genetics. 1973;74:175–195. [PMC free article] [PubMed]
  • Linde CC, Zhan J, McDonald BA. Population structure of Mycosphaerella graminicola: from lesions to continents. Phytopathology. 2002;92:946–955. [PubMed]
  • Lively CM, Dybdahl MF, Jokela J, Osnas EE, Delph LF. Host sex and local adaptation by parasites in a snail–trematode interaction. American Naturalist. 2004;164:S6–S18. [PubMed]
  • Lynch M. Destabilizing hybridization, general-purpose geneotypes and geographic parthenogenesis. Quarterly Review of Biology. 1984;59:257–290.
  • Mantel N. The detection of disease clustering and a generalised regression approach. Cancer Research. 1967;27:209–220. [PubMed]
  • Maynard Smith J, Smith NH, O’Rourke M, Spratt B. How clonal are bacteria? Proceedings of the National Academy of Sciences, USA. 1993;90:4384–4388. [PMC free article] [PubMed]
  • McDonald BA, Linde C. Pathogen population genetics, evolutionary potential, and durable resistance. Annual Review of Phytopathology. 2002;40:349–379. [PubMed]
  • McKay JK, Latta RG. Adaptive population divergence: markers, QTL, and traits. Trends in Ecology & Evolution. 2002;17:285–291.
  • Meirmans PG, Van Tienderen PH. genotype and genodive: two programs for the analysis of genetic diversity of asexual organisms. Molecular Ecology Notes. 2004;4:792–794.
  • Merilä J, Crnokrak P. Comparison of genetic differentiation at marker loci and quantitative traits. Journal of Evolutionary Biology. 2001;14:892–903.
  • Milgroom MG. Recombination and the multilocus structure of fungal populations. Annual Review of Phytopathology. 1996;34:457–477. [PubMed]
  • Nei M. Genetic distance between populations. American Naturalist. 1972;106:283–292.
  • Nei M. Molecular Evolutionary Genetics. Columbia University Press; New York: 1987.
  • Ooi K, Yahara T. Genetic variation of geminiviruses: comparison between sexual and asexual host plant populations. Molecular Ecology. 1999;8:89–97.
  • Pannell JR, Charlesworth B. Neutral genetic diversity in a metapopulation with recurrent local extinction and recolonization. Evolution. 1999;53:664–676.
  • Parker MA. Nonadaptive evolution and disease resistance in an annual legume. Evolution. 1991;45:1209–1217.
  • Peakall R, Smouse PE. GenAlEx v6: Genetic Analysis in Excel. Population Genetic Software for Teaching and Research. Australian National University; Canberra, ACT: 2005. [PMC free article] [PubMed]
  • Pimentel G, Peever TL, Carris LM. Genetic variation among natural populations of Tilletia controversa and T. bromi. Phytopathology. 2000;90:376–383. [PubMed]
  • Rohlf FJ. NTSYS-Pc: Numerical Taxonomy and Multivariate Analysis System, version 2.11. Applied Biostatistics; Setauket, New York: 1993.
  • Salathé M, Salathé R, Schmid-Hempel P, Bonhoeffer S. Mutation accumulation in space and the maintenance of sexual reproduction. Ecology Letters. 2006;9:941–946. [PubMed]
  • Salathé M, Scherer A, Bonhoeffer S. Neutral drift and polymorphism in gene-for-gene systems. Ecology Letters. 2005;8:925–932.
  • Smith DL, Ericson L, Burdon JJ. Epidemiological patterns at multiple spatial scales: an 11-year study of a Triphragmium ulmariaeFilipendula ulmaria metapopulation. Journal of Ecology. 2003;91:890–903.
  • Spiers AG, Hopcroft DH. Comparitive studies of the poplar rusts Melampsora medusae, M. larici-populina and their interspecific hybrid M. medusae-populina. Mycological Research. 1994;98:889–903.
  • Thompson JN. The Coevolutionary Process. University of Chicago Press; Chicago, Illinois: 1994.
  • Thompson JN. The Geographic Mosaic of Coevolution. University of Chicago Press; Chicago, Illinois: 2005.
  • Thrall PH, Burdon JJ. Host-pathogen dynamics in a meta-population context: the ecological and evolutionary consequences of being spatial. Journal of Ecology. 1997;85:743–753.
  • Thrall PH, Burdon JJ. Evolution of virulence in a plant host-pathogen metapopulation. Science. 2003;299:1735–1737. [PubMed]
  • Thrall PH, Burdon JJ, Bock CH. Short-term epidemic dynamics in the Cakile maritimaAlternaria brassicicola host–pathogen association. Journal of Ecology. 2001a;89:723–735.
  • Thrall PH, Burdon JJ, Young A. Variation in resistance and virulence among demes of a plant host–pathogen metapopulation. Journal of Ecology. 2001b;89:736–748.
  • Thrall PH, Burdon JJ, Bever JD. Local adaptation in the Linum marginaleMelampsora lini host–pathogen interaction. Evolution. 2002;56:1340–1351. [PubMed]
  • Wade MJ, McCauley DE. Extinction and recolonization: their effects on genetic differentiation of local populations. Evolution. 1988;42:995–1005.
  • Zhan J, Linde CC, Jürgens T, Merz U, Steinebrunner F, McDonald BA. Variation for neutral markers is correlated with variation for quantitative traits in the plant pathogenic fungus Mycosphaerella graminicola. Molecular Ecology. 2005;14:2683–2693. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...