Read 2006_Stukenbrock_1.pdf text version

Molecular Ecology (2006)

doi: 10.1111/j.1365-294X.2006.02986.x

Blackwell Publishing Ltd

Global migration patterns in the fungal wheat pathogen Phaeosphaeria nodorum

E V A H . S T U K E N B R O C K , S Ø R E N B A N K E and B R U C E A . M c D O N A L D Institute of Integrative Biology, Plant Pathology, ETH Zurich, LFW, Universitätstrasse 2, CH-8092 Zurich, Switzerland

Abstract

The global migration patterns of the fungal wheat pathogen Phaeosphaeria nodorum were analysed using 12 microsatellite loci. Analysis of 693 isolates from nine populations indicated that the population structure of P. nodorum is characterized by high levels of genetic diversity and a low degree of subdivision between continents. To determine whether genetic similarity of populations was a result of recent divergence or extensive gene flow, the microsatellite data were analysed using an isolation-with-migration model. We found that the continental P. nodorum populations diverged recently, but that enough migration occurred to reduce population differentiation. The migration patterns of the pathogen indicate that immigrants originated mainly from populations in Europe, China and North America.

Keywords: directional gene flow, microsatellites, population divergence, population genetic structure, Septoria nodorum, sexual recombination, Stagonospora nodorum Received 3 December 2005; revision received 16 March 2006; accepted 30 March 2006

Introduction

Fungal plant pathogens exhibit a diverse range of life histories and dispersal mechanisms that have important consequences for disease dynamics and pathogen persistence. Genetic markers have been widely applied to study population dynamics and migration patterns of plant pathogens at spatial scales ranging from single fields to continents (Burdon et al. 1982; Burt et al. 1997; Engelbrecht et al. 2004; Banke & McDonald 2005). Long-distance dispersal of plant pathogens is a common phenomenon that may occur either naturally by air-dispersed spores or via human-mediated movement of infected plant material and seeds. By analysing the distribution of genetic diversity within and among populations, it is possible to identify centres of diversity and patterns of migration (Beerli & Felsenstein 2001). The centre of origin of a pathogen is likely to consist of populations with more genetic variability than recently founded populations (Templeton et al. 1995). Patterns of global migration from the centre of origin into new territories have been described for a few plant pathogens such as Phytophthora infestans from Mexico to North America and Europe, and

Correspondence: Eva H. Stukenbrock, Fax: +41-44-632-1572; E-mail: [email protected] © 2006 Blackwell Publishing Ltd

Mycosphaerella graminicola from the Middle East and Europe to `New World' continents (Fry et al. 1992; Banke et al. 2004). A newly founded population often shares alleles with the source population of immigrants. For the population geneticist, it may be difficult to differentiate the relative importance of gene flow and persistence of variation in both populations for maintaining shared polymorphisms (Hey et al. 2004). In the case of plant pathogens, this knowledge is needed to understand the processes underlying population divergence and evolution. An isolation-withmigration model was applied to genetic data to analyse gene flow between recently separated populations, e.g. of humans and fish (Nielsen & Wakeley 2001; Hey et al. 2004; Hey 2005), but to our knowledge the model has not yet been applied to studies of pathogens. Another factor that affects the ability of a pathogen to adapt to new environments is the mode of reproduction and mating system (McDonald & Linde 2002). The reproductive modes of fungal pathogens range from purely asexual to highly outcrossing (Milgroom 1996). Under asexual reproduction, combinations of alleles giving high fitness are kept together and selected clones carrying these allele combinations can increase to high frequencies. In the rice blast fungus Magnaporthe grisea, which is predominantly asexual (Babujee & Gnanamanickam 2000), this has

2 E . H . S T U K E N B R O C K , S . B A N K E and B . A . M c D O N A L D resulted in a clonal population structure with limited genetic variability characterized by a few clonal lineages distributed across large geographical areas (Ou 1980). At the other extreme, frequent sexual reproduction may lead to a high degree of genetic variability and limited clonality as demonstrated globally for field populations of the wheat pathogen Mycosphaerella graminicola (Chen & McDonald 1996; Zhan et al. 2003). The fungal pathogen Phaeosphaeria (syn. Stagonospora) nodorum causes Stagonospora nodorum leaf blotch and glume blotch diseases on wheat. The life cycle of this heterothallic ascomycete includes both an asexual and a sexual stage that exhibit different mechanisms of dispersal. Asexually produced pycnidiospores are dispersed over short distances by rain-splash while sexually produced ascospores are wind-dispersed and have the potential to be blown over considerable distances (Griffiths & Hann 1976; Brennan et al. 1985; Keller et al. 1997a, b; Arseniuk et al. 1998). Seed-borne dispersal is also believed to play an important role in moving P. nodorum over long distances (Shah et al. 1995; Bennett et al. 2005). The finding of high levels of genotypic diversity and linkage equilibrium have lead to contrasting hypotheses regarding the dispersal mechanisms and primary inoculum of the pathogen (Keller et al. 1997a, b; Bennett et al. 2005). Lack of population differentiation over considerable distances coupled with the finding of linkage equilibrium for restriction fragment length polymorphism (RFLP) loci suggested that airborne ascospores provide an important source of primary infection (McDonald et al. 1994; Keller et al. 1997a, b). This hypothesis was supported by the finding of equal distribution of the two mating types in field populations from around the world (Solomon et al. 2004; Sommerhalder et al. 2006). However, the sexual stage of P. nodorum has only rarely been seen in nature, and it has been suggested that primary infection via infected seed coupled with humanmediated dispersal of infected seed over long distances could be a major contributor to the observed patterns of genetic variability (Bennett et al. 2005). Little is known about P. nodorum population differentiation at a global scale. In an earlier study, populations from Oregon, Texas and Switzerland were compared using RFLP markers (Keller et al. 1997b). But it is possible that the population structure of these regions was not representative of other important wheat-growing regions in Asia, Australia and Africa. Considering the potential for longdistance dispersal via ascospores and transport of infected seeds, we hypothesized that high rates of migration had occurred across and among continents, leading to low levels of population differentiation on a global scale. It is most likely that migration among populations has not been symmetrical, but rather reflects historical movement of the host. Wheat was introduced into different continents at different times. From its centre of origin in the Fertile Crescent, wheat was first introduced into Europe and Asia during the spread of agriculture 7000­10 000 years ago (Salamini et al. 2002). It was only during the last 500 years that European colonists introduced wheat into the New World (Americas and Australia). We hypothesized that the pathogen populations closest to the centre of origin of the host would represent the oldest pathogen populations and would exhibit the highest levels of genetic diversity, while New World populations would be younger and exhibit lower levels of diversity. As previous studies suggested that sexual reproduction plays an important role in the life cycle of the pathogen (McDonald et al. 1994; Keller et al. 1997a), we hypothesized that frequent recombination makes a significant contribution to the structure of populations on all continents. To test these hypotheses, we here used a set of 12 polymorphic microsatellite loci to characterize the population genetic structure of P. nodorum on a global scale including nine populations from five different continents. The main objectives in the present study were (i) to determine the global genetic structure of P. nodorum and to identify a putative centre of diversity; (ii) to estimate the contribution and direction of gene flow to the observed population genetic structure; (iii) to determine the relative time of divergence between the different continental populations; and (iv) to infer the relative contributions of sexual and asexual reproduction to pathogen evolution.

Materials and methods Sample collection

A total of 693 Phaeosphaeria nodorum strains isolated from wheat were included in the analysis of genetic structure. Isolates originated from field populations of P. nodorum from six different regions on five continents: Australia (Narrogin, Western Australia), China (Fujian Province), South Africa (Southwestern Cape), Central and North America (Michoacan in Mexico and Oregon, Texas and New York in the USA), and Europe (Denmark and Switzerland) (Table 1). The populations from Texas and Oregon were described in previous studies using RFLP markers (McDonald et al. 1994, Keller et al. 1997a). Pathogen populations were collected from one or two wheat fields in each region. The Texas and Swiss populations included samples from two different fields. Sampling strategies differed among populations and therefore the geographical populations sometimes represented different spatial scales within each region. For example, the Swiss population was sampled from wheat fields that were 150 km apart, while the Texas population was sampled from fields that were 100 m apart. Three sampling methods were used. In four of the collections, a standardized six- or eight-site hierarchical transect was used (McDonald et al. 1994). For two collections, infected leaves were

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

GLOBAL MIGRATION OF PHAEOSPHAERIA NODORUM 3

Table 1 Phaeosphaeria nodorum populations included in the analysis Continental region Africa Asia Australia Europe Latin America North America Population South Africa China Australia Denmark Switzerland Mexico New York Oregon Texas Location Southwestern Cape Fujian Province Narrogin Abed Bern, Winterthur Michoacan Ithaca Hyslop Overton Year 1995 2001 2001 1994 1999 1993 1991 1993 1992 Sampling strategy Random Transects Hierarchical Random Hierarchical Random Transects Hierarchical Hierarchical Collectors P. Crous R. Wu B.A. McDonald & R. Loughman M. Rasmussen B.A. McDonald & V. Michel L. Gilchrist G. Bergstrom M. Schmidt B.A. McDonald & L. Nelson

collected at 1- to 2-m intervals along two or more transects. Three of the collections were made from infected leaves chosen at random from different parts of a field.

(Zhan et al. 2003) and by Stoddart & Taylor's measure (1988) as described by McDonald et al. (1994). The significance of differences in genotypic diversity was calculated using a t-test (Chen et al. 1994). Distribution of gene diversity. Distribution of gene diversity was estimated using a hierarchical analysis of molecular variance (amova) in arlequin version 2.000 by calculating the sum of squared size differences (RST) (http:/ /lgb.unige.ch/ arlequin) (Weir & Cockerham 1984; Michalakis & Excoffier 1996). This model was also used to calculate genetic distances between microsatellite loci. RST values are based on a stepwise mutation model (SMM) developed for microsatellite markers (Kimura & Ohta 1978). Microsatellite size polymorphism arises by slipped-strand mispairing during DNA replication so that the same microsatellite allele may arise multiple times. The SMM accounts for this possible size homoplasy in contrast to traditional variance models such as FST based on infinite allele models. Allelic richness. To test if genetic diversity within P. nodorum populations differed on a global scale we used estimates of allelic richness calculated by the program fstat (http:/ / www.unil.ch/izea/softwares/fstat.html) (Goudet 2001). Allelic richness was corrected for unequal sample sizes by rarefaction to a uniform sample size of n as described by El Mousadik & Petit (1996). The New York population was removed from this analysis due to small sample size. Switzerland and Denmark were pooled to form a single European population due to the small Danish sample size and the lack of subdivision between the two populations (see Table 4). Allelic richness was calculated as: rn = [1 - ((N - Ni)/n)/(N/n)] where Ni represents the number of occurrences of the ith haplotype among N sampled individuals within a population, while n is the smallest sample size, in this case 24 (clone-corrected population from Mexico). Each population

Data collection

DNA extraction was performed as described by McDonald et al. (1994). Twelve polymorphic EST-derived satellite loci (Stukenbrock et al. 2005) were used to characterize the isolates. This set of loci included 11 microsatellites (SNOD1, SNOD3, SNOD5, SNOD11, SNOD15, SNOD16, SNOD17, SNOD21, SNOD22, SNOD23 and SNOD26) and one minisatellite repeat (SNOD8). Multiplex polymerase chain reaction (PCR) conditions were carried out as described previously using fluorescent labelled primers (Stukenbrock et al. 2005). Allele assignments were on an ABI 3100 sequencer using the manufacturer's instructions with the program genescan version 3.7 from Applied Biosystems. A 12-digit numeric identifying code was formulated for each isolate by joining together the numbers identifying the alleles present for each locus. This numeric code represented the multilocus genotype for each isolate.

Data analyses

Gene diversity. Isolates represented by the same multilocus haplotype were treated as clones. For analyses of genetic diversity in P. nodorum populations we used a clonecorrected data set in which only one clone of each multilocus haplotype was included. The number of alleles and gene diversity for each locus and for each population across all loci was calculated using the program popgene32 (http:/ /www.ualberta.ca/fyeh/index.htm) (Nei 1973; Yeh et al. 1999). Genotypic diversity was quantified as the clonal fraction of each population, calculated using: 1 - [(number of different genotypes)/(total number of samples)]

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

4 E . H . S T U K E N B R O C K , S . B A N K E and B . A . M c D O N A L D sample was resampled using 24 random individuals to obtain the distribution of allelic richness among 100 resampled data sets. Resampling was performed using the program resampling stats for Excel (http:/ /www.resample.com/ content/software/excel) (Simon 1997). Gene flow and divergence time. The three North American populations were pooled into one population and the two European populations into another to examine migration among continents. The relative amounts of gene flow occurring between populations on different continents were estimated with the program im (http:/ /lifesci.rutgers.edu/ heylab/HeylabSoftware.htm# IM) (Nielsen & Wakeley 2001; Hey 2005). im applies a Bayesian likelihood analysis to fit an isolation-with-migration model to the data, allowing the estimation of many of the processes that occur during the division of one population into two. Likelihood estimates of the possible genealogies of the data set were calculated using a Markov chain Monte Carlo approach and a given mutation model assuming neutrality and free recombination among loci. Different mutation models can be applied for analysis of the data; in this case we used SMM (Kimura & Ohta 1978). The model estimated mutation rates for each locus separately and incorporated six basic demographic parameters, including (theta) of the ancestral and the two descendant populations scaled by mutation rate (for haploid organisms estimated as = 2Neµ where Ne = effective population size of either the two descendant populations or the ancestral population); directional gene flow rates per gene copy per generation: m1 from population one to two, and m2 from population two to one (m = m/µ where m = genes moved from one population to the other); and the time t since population divergence from an ancestral panmictic population (t = tµ where t = generations since population splitting). Ten loci were selected for the im analysis. SN11 was excluded due to a low degree of gene diversity and missing data at this locus in the Swiss population, and SN26 was excluded because the locus did not follow a stepwise mutation model. A total of 15 pairwise population analyses were conducted for the six regional populations: Australia, China, South Africa, Mexico, North America and Europe. Each of the 15 analyses was conducted using at least four independent runs with a total run length of 500 000 steps, and a burn-in period of 100 000 updates. Scalar for maximum migration rates, values and maximum time of population splitting were set at 50, 50 and 100, respectively, in all pairwise comparisons. Marginal histograms were compared between all independent runs to ensure consistent estimation of population parameters. Additionally, lower and upper bounds of the estimated 90% highest posterior density (HPD) intervals were calculated for each parameter. Linkage disequilibrium. Estimates of linkage disequilibrium in populations of P. nodorum were based on the index of association (IA; Maynard Smith et al. 1993). Calculation of the IA is based on the variance of pairwise distances between individuals (i.e. the number of loci at which they differ). The observed data set is used to create a scrambled data set resembling the one of a population with free recombination. The IA of the observed data is compared to the distribution of IA for 1000 resampled recombining data sets. IA is zero for a population with free recombination and the analysis tests whether the observed data set is within the distribution of IAs for the artificially recombined data sets. IA was calculated using the software multilocus provided by A. Burt (http:/ /www.bio.ic.ac. uk/evolve/ software/multilocus/). Linkage disequilibrium was additionally analysed using the multilocus association test described by Brown et al. (1980) implemented in the popgene32 software (Yeh et al. 1999). In this analysis, multilocus structure is evaluated by comparing observed and expected numbers of heterozygous loci of all pairwise haplotype combinations. This analysis tests the distribution of K, defined as the number of loci that are different when two random individuals are compared, and accounts for the degree of polymorphism at each locus and the association of all alleles among loci over all individuals (Brown et al. 1980). New York and Denmark were excluded from these analyses due to the low sample sizes in these populations.

Results Population gene and genotypic diversity

Twelve microsatellite loci were amplified from 693 isolates of Phaeosphaeria nodorum representing nine different regional populations from five continents. The number of alleles at each locus ranged from 2 to 28 (Table 2). Average gene diversity across all populations ranged from 0.19 (SNOD23) to 0.90 (SNOD26) with an average of 0.60 (Table 2). Gene and genotypic diversity for the nine populations are summarized in Table 3. The clonal fraction ranged from 0.33 in the Mexican population to 0.02 in the Swiss population. The degree of genotypic diversity, estimated by Stoddart and Taylor's measure, ranged from 33% (Mexico) to 86% (Australia) of its theoretical maximum, but the differences in genotypic diversities were not significant. Average estimates of Nei's (1973) gene diversity ranged from 0.44 (Mexico) to 0.57 (Texas), while total gene diversity across all populations was 0.58 (Table 3). An analysis of molecular variance (amova) was used to describe the hierarchical distribution of population subdivision at regional and continental scales. Analyses of the distribution of gene diversity within and among P. nodorum

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

GLOBAL MIGRATION OF PHAEOSPHAERIA NODORUM 5

Table 2 The number of alleles and measures of gene diversity in Phaeosphaeria nodorum populations Coding/ noncoding DNA NC C NC NC NC C NC C NC NC NC No data New South Total Australia China Denmark Mexico York Oregon Texas Africa Switzerland no. of Gene (N = 71)* (N = 97) (N = 9) (N = 24) (N = 16) (N = 51) (N = 53) (N = 46) (N = 284) alleles diversity 3 4 7 3 2 3 4 7 4 5 3 8 10 5 11 5 2 5 4 14 7 8 4 12 3 1 6 2 2 2 4 3 4 3 1 6 4 3 7 3 1 3 3 5 5 3 1 10 8 2 10 5 2 5 5 7 6 7 2 9 6 3 10 3 1 3 4 7 7 6 2 11 8 2 10 5 2 5 6 8 6 7 3 9 3 4 4 3 2 4 4 7 4 5 2 8 19 7 14 2 2 8 6 15 10 10 6 18 29 10 17 8 2 10 10 21 12 13 10 20 0.47 0.30 0.83 0.44 0.50 0.52 0.68 0.77 0.77 0.68 0.19 0.90

Locus number SNOD1 SNOD3 SNOD5 SNOD8 SNOD11 SNOD15 SNOD16 SNOD17 SNOD21 SNOD22 SNOD23 SNOD26

Size range 253­439 278­318 409­472 325­425 231­ 236 156­174 188­208 92­158 191­ 228 225­267 294­402 188­268

*N, sample size for each population. Calculated according to Nei (1973).

Table 3 Overall measures of gene and genotypic diversity and estimates of linkage disequilibrium in populations of Phaeosphaeria nodorum Total no. of isolates 73 101 10 291 31 18 56 58 55 693 Clone corrected 71 97 9 284 24 16 51 53 46 651 Clonal fraction 0.03 0.04 0.10 0.02 0.23 0.11 0.09 0.09 0.16 0.06 Average gene diversity (h) 0.47 ± 0.22 0.54 ± 0.24 0.47 ± 0.29 0.55 ± 0.25 0.44 ± 0.28 10.50 ± 0.27 0.47 ± 0.3 0.57 ± 0.23 0.55 ± 0.16 0.58 ± 0.22 Allelic richness 3.15 ± 022 4.15 ± 0.31

Population Australia China Europe Denmark Switzerland Mexico North America New York Oregon Texas South Africa All populations

|/| max* 86% 68% 83% 33% 50%

IA 0.04 0.12 ND 0.03 0.24 ND 0.09 0.08 0.63

P 0.37 0.09 ND 0.12 0.08 ND 0.28 0.81 0.001

2 Sk

L2§ 2.58 2.71 ND 2.32 3.09 ND 2.55 2.73 2.29

2.25 2.55 ND 2.25 4.64 ND 2.53 2.73 4.25

4.44 ± 0.39 3.03 ± 0.23 ND 3.65 ± 0.28 4.16 ± 0.33 3.44 ± 0.2 4.81 ± 0.46

41%

*|/| max: genotypic diversity calculated according to Stoddart & Taylor (1988). IA, the index of association measure of multilocus linkage disequilibrium (Maynard Smith et al. 1993). 2 Sk, observed variance of the number of heterozygous comparisons calculated using Brown's multilocus association (Brown et al. 1980). 2 §L2, upper 95% confidence limit of Sk . Values in bold show statistically significant values.

populations showed that only 10% of the total gene diversity was distributed among the five continents while no subdivision was found between populations within European and North American continents (0%). The main source of variation, 90%, was found within the nine regional populations.

levels of population differentiation (RST < 0.16) for all pairwise population comparisons except South Africa. The five highest RST values were found between South Africa and the other populations, except Switzerland (RST = 0.10).

Allelic richness

Differences in allelic richness among P. nodorum populations were calculated by resampling 100 data sets. The average allelic richness across all populations was 4.81, but significant differences in genetic diversity existed among the populations (Table 3). The European

Population differentiation

Estimates of pairwise population differentiation were calculated with F-statistics using a stepwise mutation model (RST) (Table 4). The analysis indicated moderate

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

6 E . H . S T U K E N B R O C K , S . B A N K E and B . A . M c D O N A L D

Table 4 Population pairwise RST Population Australia China Denmark Mexico New York Oregon South Africa Switzerland Texas Australia **** China 0.12 **** Denmark 0.03 0.01 **** Mexico 0.01 0.06 -0.05 **** New York -0.02 0.02 0.02 0.03 **** Oregon 0.04 0.16 0.14 0.10 0.00 **** South Africa 0.10 0.01 0.18 0.19 0.22 0.26 **** Switzerland 0.01 0.13 -0.01 -0.02 -0.04 0.04 0.10 **** Texas 0.04 0.15 0.10 0.09 -0.04 0.01 0.27 0.05 ****

Table 5 Likelihood estimates of , divergence time and migration between pairs of Phaeosphaeria nodorum populations as calculated by the im program (Nielsen & Wakeley 2001; Hey 2005). Donor populations are shown on the left side and recipient populations are indicated along the top. Ninety per cent confidence intervals are shown in parentheses for all estimates Recipient population Australia = 2.4 (2.4 ­ 40.2) Australia Migration (m) Divergence time (t) China Migration (m) Divergence time (t) Europe Migration (m) Divergence time (t) Mexico Migration (m) Divergence time (t) North America Migration (m) Divergence time (t) South Africa Migration (m) Divergence time (t) China = 9.5 (9.5 ­16.8) Europe = 20.1 (9.1­71.3) Mexico = 1.1 (1.1­18) North America = 7.2 (5.2­29.9) South Africa = 2.1 (2.1­ 36.5)

****

0.3 (0.1­ 2.5) 0.4 (0.1­1.2) ****

0.7 (0.3­3.3) 0.7 (0.4­1.4) 7.2 (5.0­14.0) 0.8 (0.4­1.3) ****

0.1 (0.1­2.1) 2.2 (1.3­3.1) 5.5 (2.9­8.1) 1.2 (0.6­1.8) 21.5 (16.7­26.3) 1.1 (0.6­1.8) ****

0.7 (0.3­3.3) 2.2 (1.3­3) 1.0 (0.1­2.1) 0.05 (0.03­0.3) 3.1 (1.6­5.7) 0.6 (0.3­0.9) 0.7 (0.5­1) 0.1 (0.1­0.7) ****

0.1 (0­0.7) 1.8 (0.5­2.7) 1.7 (0.5­4.1) 1.4 (0.8­2.1) 0.5 (0.1­1.8) 0.2 (0.1­0.6) 5.5 (2.5­8.7) 0.4 (0.2­0.6) 0.0 (0­0.5) 0.2 (0.1­0.6) ****

7.3 (3.1­11.5) 0.4 (0.1­1.2) 4.1 (1.9 ­ 6.9) 0.7 (0.4 ­1.4) 2.7 (1.3 ­ 5.3) 2.2 (1.3 ­ 3.1) 1.9 (0.9 ­ 4.3) 2.2 (1.3 ­ 3) 2.3 (1.2­3.5) 1.8 (0.5­2.7)

1.3 (0.5 ­ 4.4) 0.8 (0.4­1.3) 1.9 (0.9­4.5) 1.2 (0.6­1.8) 1.4 (0.2­2.5) 0.05 (0.03 ­ 0.3) 0.1 (0.1­1.9) 1.4 (0.8­2.1)

1 (0.9­1) 1.1 (0.6­1.8) 6.3 (2.8­17.1) 0.6 (0.3­0.9) 2.8 (0.6­9.21) 0.2 (0.1­0.6)

6.9 (1.5­14.7) 0.1 (0.1­0.7) 1 (1­1) 0.4 (0.2­0.6)

0.1 (0.03­0.6) 0.4 (0.2­0.6)

population had the highest level of allelic richness (4.44) and differed significantly from Mexico (3.03), Australia (3.15), South Africa (3.44) and Oregon (3.65). China and Texas (4.15 and 4.16, respectively) had intermediate levels of allelic richness, also differing significantly from Mexico, Australia and South Africa.

Effective population size, divergence time and migration

Estimates of effective population size based on indicated that the largest population of P. nodorum was in Europe ( = 20.1) (Table 5). China and North America had intermediate population sizes ( = 9.5 and 7.2, respectively), while small values were found in Australia (2.4),

South Africa (2.1) and Mexico (1.1) consistent with measures of allelic richness. Estimates of ancestral population sizes were not calculated as these measures were likely to be biased due to missing populations between the regional populations used here. Migration estimates indicated that gene flow occurred in several directions between P. nodorum populations on different continents (Table 5). The major global migration patterns indicated by this analysis are shown in Fig. 1. In general, the European, Chinese and North American populations were donors of immigrants while the Australian, South African and Mexican populations acted as sinks for migration. Europe provided a large number of immigrants mainly to Australia and Mexico, while the outgoing

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

GLOBAL MIGRATION OF PHAEOSPHAERIA NODORUM 7

Fig. 1 Major global migration patterns of Phaeosphaeria nodorum. values are shown for each population and migration rates are indicated on arrows.

migration from China was to Mexico, Australia and Europe. The North American population donated a considerable number of immigrants to both Europe and Mexico. im estimates indicated that some migration also occurred from South Africa to Europe (Table 5). Divergence times between populations were in most cases close to zero, supporting the hypothesis of a population genetic structure with low degree of subdivision. Small t (divergence time) estimates were found between the three large donor populations and between donor and receiving populations. However t estimates between the receiving populations in Australia, South Africa and Mexico suggested that these populations had been separated for a longer time or that they were founded by different migration events. Divergences between the Australian population and North America and Mexico were found to represent the oldest population splits.

Linkage disequilibrium

The index of association (IA) was determined for all field populations of P. nodorum. The hypothesis of linkage disequilibrium was rejected for all populations except the South African population (Table 3). Brown's analysis of multilocus associations found significant disequilibrium in the South African as well as the Mexican populations while the Texan population was close to significance (Table 3). Disequilibrium values were nonsignificant for all other populations, consistent with a recombining population structure.

Discussion

The population genetic structure of Phaeosphaeria nodorum from five different continents showed moderate population differentiation consistent with high levels of gene flow among continents. These findings suggest that longdistance dispersal of P. nodorum has contributed to the genetic structure of the pathogen on a global scale.

Though all P. nodorum populations had high levels of genetic diversity, the allelic richness was significantly different among populations (Table 3). Europe, Texas and China were close to the average allelic richness calculated across all populations, indicating larger effective population sizes in these populations. Mexico, Australia and South Africa, on the other hand, had lower levels of allelic richness consistent with smaller effective population sizes. The effective population size is an important contributor to the genetic structure of a pathogen population. Genetic drift may lead to the loss of alleles over time in small populations while populations with large effective sizes have a greater potential to accumulate genetic variability (Newman & Pilson 1997). The effects of genetic drift can be offset by migration, which may explain why high genetic variability is encountered even in founder populations of P. nodorum. The Mexican population differed from the other P. nodorum populations by originating from an isolated experimental field site inoculated with a limited number of isolates. The remoteness and unnatural origin of this population is a plausible explanation for the low degree of genetic variability and small effective population size. The low levels of genetic variability and small population sizes found in Australia and South Africa are most likely due to relatively recent founder events. Founder events are known to affect the genetic structure of a pathogen by reducing genetic variability and increasing linkage disequilibrium (Taylor et al. 1999; Hewitt 2000). For other pathogen species, similar founder effects have been demonstrated during the spread of pathogens across continental and regional scales, creating populations with different levels of genetic diversity (Goodwin 1997; Engelbrecht et al. 2004; Rivas et al. 2004). The North American P. nodorum population was also expected to be younger because wheat and its pathogens were introduced there less than 500 years ago. However, these populations were characterized by a high degree of genetic variability and large effective population sizes. This suggests that

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

8 E . H . S T U K E N B R O C K , S . B A N K E and B . A . M c D O N A L D multiple introductions from Europe have occurred, probably in addition to a considerable pathogen population expansion as wheat cultivation spread across North America and was accompanied by frequent movement of infected seeds and plant material between regions. Previous analyses of P. nodorum population genetic structure found no population subdivision when partitioning gene diversity into different hierarchical levels including 1 m2 plots within single fields, fields at a regional scale and fields on different continents (Europe and North America; McDonald et al. 1994; Keller et al. 1997b). Based on these findings, we considered our regional populations as representative of five different continents. In agreement with the earlier studies, we found moderate subdivision between the geographically separated populations of P. nodorum (Table 4). The analysis indicated that the overall pattern of pathogen migration was from source populations in Europe, China and North America to sink populations in Australia, South Africa and Mexico. This pattern supports the hypothesis that populations where wheat cultivation was more recently introduced (Fig. 1), except North America, represent younger and more recently founded populations. The observed pattern could be due to either a recent divergence of the populations or significant gene flow among populations. These two processes were considered separately in the program im using an isolation-with-migration model (Nielsen & Wakeley 2001). This analysis indicated that the global populations of P. nodorum diverged recently from their common centre of origin, but that a significant amount of migration simultaneously reduced population differentiation (Table 5). The South African population was genetically differentiated from the other populations. Migration estimates indicated that the South African population received immigrants mainly from Mexico, which is inconsistent with the overall pattern of large source and small sink populations. The Mexican population had the highest clonality and the lowest gene diversity, and this finding could represent a single migration event of a few successful Mexican clones to South Africa through the transport of infected germplasm. Another possibility is that Europe was the source population for both Mexico and South Africa, and the connection between Mexico and South Africa is indirect, reflecting a common ancestral source population. Regular gene flow among populations can make an important contribution to gene diversity in small pathogen populations (e.g. the founder populations in Australia, South Africa and Mexico), introducing new alleles that can then be recombined to form new genotypes (Maynard Smith 1968; Burdon & Silk 1997). But analyses of multilocus associations and degree of clonality suggest that sexual reproduction makes a significant contribution to the genetic structure of P. nodorum, indicating that alleles are regularly recombined except in the two founder populations in South Africa and Mexico. The indication of frequent sexual recombination within naturally infected field populations of P. nodorum is consistent with previous studies based on multilocus associations of RFLP loci and frequencies of mating type alleles (McDonald et al. 1994; Keller et al. 1997a, b; Sommerhalder et al. 2006). Bennett et al. (2005) also found high levels of genotypic diversity and linkage equilibrium in seed populations of P. nodorum. These results were interpreted as evidence that either asexual seed-borne fungus or windborne ascospores could serve as primary inocula of the pathogen. Bennett et al. (2005) argued that rare recombination events are sufficient to maintain a considerable amount of genotypic diversity through a constant mixing of infected seed within and between agricultural regions. In plant pathosystems where sexual and asexual populations could be directly compared, it was shown that frequent sexual reproduction could increase genetic diversity while mainly clonal propagation with infrequent recombination decreased genetic diversity (Roelfs & Groth 1980; Welz & Kranz 1987). Long-distance dispersal of asexual pathogens by infected plant material could lower population differentiation; however, we consider it unlikely that the high degree of genotypic diversity found in the global population of P. nodorum could be maintained without frequent sexual reproduction. The dispersal of plant pathogens over long distances can have important consequences for disease management in agricultural ecosystems (Brown & Hovmøller 2002). Our findings suggest that P. nodorum has the potential to disperse across continental and global scales, while frequent recombination allows the pathogen to recombine different alleles and create new genotypes. This knowledge is important in assessing the risk of introduction of new virulence or fungicide resistance alleles into existing pathogen populations (McDonald & Linde 2002). The global genetic structure of P. nodorum is likely to be related to the expansion of agriculture and European colonization of the New World. Wheat has been cultivated in Europe and China since 4000­5000 bc (Ho 1969; Salamini et al. 2002). European colonists introduced wheat into the Americas, Australia, and South Africa during the past 500 years. The global genetic structure of P. nodorum suggests that the pathogen was dispersed with its host throughout the world during colonization. But more recent migration of the pathogen among continents may still affect the observed population genetic structure. In a recent phylogeographical study of the wheat pathogen Mycosphaerella graminicola, Banke & McDonald (2005) found a strikingly similar global genetic structure indicating that Old World populations in Europe and the Middle East represented ancient centres of origin, while recent migration had occurred as a consequence of the international wheat trade. Though the centre of origin of P. nodorum could not be

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

GLOBAL MIGRATION OF PHAEOSPHAERIA NODORUM 9 determined in the present study, the large effective population size in Europe and the indicated global migration patterns suggest a similar demographic history. Future studies of P. nodorum will need to include populations from the Fertile Crescent to determine if the pathogen and its host co-evolved in the host's centre of origin.

El Mousadik A, Petit RJ (1996) High levels of genetic differentiation for allelic richness among populations of the argan tree [Argania spinosa (L) Skeels] endemic to Morocco. Theoretical and Applied Genetics, 92, 832­839. Engelbrecht CJB, Harrington TC, Stemmel J, Capretti P (2004) Genetic variation in eastern North American and putatively introduced populations of Ceratocystis fimbriata f. platani. Molecular Ecology, 13, 2995­3005. Fry WE, Goodwin S, Matuszak JM, Spielman LJ, Milgroom MG, Drenth A (1992) Population genetics and intercontinental migration of Phytophthora infestans. Annual Review of Phytopathology, 30, 107­129. Goodwin SB (1997) The population genetics of Phytophthora. Phytopathology, 87, 462­473. Goudet J (2001) FSTAT, a program to estimate and test gene diversities and fixation indices. Version 2.9.3. Available from www.unil.ch/ izea/softwares/fstat.html. Griffiths DC, Hann CAO (1976) Dispersal of Septoria nodorum spores and spread of glume blotch of wheat in the field. Transactions of the British Mycological Society, 67, 413­418. Hewitt G (2000) The genetic legacy of the Quaternary ice ages. Nature, 405, 907­913. Hey J (2005) On the number of New World founders: a population genetic portrait of the peopling of the Americas. Public Library of Science Biology, 3, e193. Hey J, Won Y-J, Sivasundar A, Nielsen R, Markert JA (2004) Using nuclear haplotypes with microsatellites to study gene flow between recently separated Cichlid species. Molecular Ecology, 13, 909­919. Ho PT (1969) The Loess and the origin of Chinese agriculture. American Historical Review, 75, 1­36. Keller SM, McDermott JM, Pettway RE, Wolfe MS, McDonald BA (1997a) Gene flow and sexual reproduction in the wheat glume blotch pathogen Phaeosphaeria nodorum (anamorph Stagonospora nodorum). Phytopathology, 87, 353­358. Keller SM, Wolfe JM, McDermott JM, McDonald (1997b) High genetic similarity among populations of Phaeosphaeria nodorum across wheat cultivars and regions in Switzerland. Phytopathology, 87, 1134­1139. Kimura M, Ohta (1978) Stepwise mutational model and distribution of allelic frequencies in a finite population. Proceedings of the National Academy of Sciences, USA, 75, 2868­2872. Maynard Smith J (1968) Evolution in sexual and asexual populations. American Naturalist, 102, 469­473. Maynard Smith J, Smith NH, O'Rourke M, Spratt BG (1993) How clonal are bacteria? Proceedings of the National Academy of Sciences, USA, 90, 4384­4388. McDonald BA, Linde C (2002) Pathogen population genetics, evolutionary potential and durable resistance. Annual Review of Phytopathology, 40, 349­379. McDonald BA, Miles LR, Pettway RE (1994) Genetic variability in nuclear DNA in field populations of Stagonospora nodorum. Phytopathology, 84, 250­255. Michalakis Y, Excoffier L (1996) A generic estimation of population subdivision using distances between alleles with special reference for microsatellite loci. Genetics, 142, 1061­1064. Milgroom MG (1996) Recombination and the multilocus structure of fungal populations. Annual Review of Phytopathology, 34, 457 ­ 477. Nei M (1973) Analysis of gene diversity in subdivided populations. Proceedings of the National Academy of Sciences, USA, 70, 3321­3323.

Acknowledgements

We thank Søren Rosendahl for helpful comments and discussion on the manuscript and Andreas Näf and Patrick Brunner for assistance with the data analyses. This work was funded by the Swiss Federal Institute of Technology (ETH), Zurich and the Swiss National Science Foundation (Grant 3100A0-104145).

References

Arseniuk E, Góral T, Scharen AL (1998) Seasonal patterns of spore dispersal of Phaeosphaeria spp. and Stagonospora spp. Plant Disease, 82, 187­194. Babujee L, Gnanamanickam SS (2000) Molecular tools for characterization of rice blast pathogen (Magnaporthe grisea) population and molecular marker-assisted breeding for disease resistance. Current Science, 78, 248 ­ 257. Banke S, McDonald BA (2005) Migration patterns among global populations of the pathogenic fungus Mycosphaerella graminicola. Molecular Ecology, 14, 1881­1896. Banke S, Peschon A, McDonald BA (2004) Phylogenetic analysis of globally distributed Mycosphaerella graminicola populations based on three DNA sequence loci. Fungal Genetics and Biology, 41, 226­238. Beerli P, Felsenstein J (2001) Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proceedings of the National Academy of Sciences, USA, 98, 4563 ­ 4568. Bennett RS, Milgroom MG, Bergstrom GC (2005) Population structure of seedborne Phaeosphaeria nodorum on New York wheat. Phytopathology, 95, 300 ­ 305. Brennan RM, Fitt BDL, Taylor GS, Colhoun J (1985) Dispersal of Septoria nodorum pycnidiospores by simulated raindrops in still air. Phytopathology, 112, 281­ 290. Brown AHD, Feldman MW, Nevo E (1980) Multilocus structure of natural populations of Hordeum spontaneum. Genetics, 96, 523­536. Brown JKM, Hovmøller MS (2002) Aerial dispersal of pathogens on the global and continental scales and its impact on plant disease. Science, 297, 537 ­ 541. Burdon JJ, Marshall DR, Luig NH, Gow DJS (1982) Isozyme studies on the origin and evolution of Puccinia graminis f.sp. tritici in Australia. Australian Journal of Biology, 35, 231­ 238. Burdon JJ, Silk J (1997) Sources and patterns of diversity in plantpathogenic fungi. Phytopathology, 87, 664 ­ 669. Burt PJA, Rutter J, Gonzales H (1997) Short-distance wind dispersal of the fungal pathogen causing Sigatoka diseases in banana and plantain. Plant Pathology, 46, 451­ 458. Chen RS, McDonald BA (1996) Sexual reproduction plays a major role in the genetic structure of populations of the fungus Mycosphaerella graminicola. Genetics, 142, 1119 ­1127. Chen RS, Boeger JM, McDonald BA (1994) Genetic stability in a population of a plant pathogenic fungus over time. Molecular Ecology, 3, 209­218.

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

10 E . H . S T U K E N B R O C K , S . B A N K E and B . A . M c D O N A L D

Newman D, Pilson D (1997) Increased probability of extinction due to decreased genetic effective population size: experimental populations of Clarkia pulchella. Evolution, 51, 354 ­ 362. Nielsen R, Wakeley J (2001) Distinguishing migration from isolation: a Markov chain Monte Carlo approach. Genetics, 158, 885­896. Ou SH (1980) Pathogen variability and host resistance in rice blast disease. Annual Review of Phytopathology, 18 (167 ­1), 87. Rivas GG, Zapater MF, Abadie C, Carlie J (2004) Founder effects and stochastic dispersal at the continental scale of the fungal pathogen Mycosphaerella fijiensis. Molecular Ecology, 13, 471­ 482. Roelfs AP, Groth JV (1980) A comparison of virulence phenotypes in wheat stem rust populations reproducing sexually and asexually. Phytopathology, 70, 855 ­ 862. Salamini F, Ozkan H, Brandolini A, Schafer-Pregl R, Martin W (2002) Genetics and geography of wild cereal domestication in the Near East. Nature Reviews Genetics, 3, 429 ­ 441. Shah DA, Bergstrom GC, Ueng PP (1995) Initiation of Septoria nodorum blotch epidemics in winter wheat by seedborne Stagonspora nodorum. Phytopathology, 85, 452 ­ 457. Simon J (1997) Resampling: The New Statistics. Resampling Stats, Arlington, Virginia. Solomon PS, Parker K, Loughman R, Oliver RP (2004) Both mating types of Phaeosphaeria (anamorph Stagonospora) nodorum are present in Western Australia. European Journal of Plant Pathology, 110, 763­766. Sommerhalder R, McDonald BA, Zhan J (2006) The frequencies and spatial distribution of mating types in Stagonospora nodorum are consistent with recurring sexual reproduction. Phytopathology, 96, 234­239. Stoddart JA, Taylor JF (1988) Genotype diversity: estimation and prediction in samples. Genetics, 118, 705 ­ 711. Stukenbrock EH, Banke S, Zala M, McDonald BA, Oliver RP (2005) Isolation and characterization of EST-derived microsatellite loci from the fungal wheat pathogen Phaeosphaeria nodorum. Molecular Ecology Notes, 5, 931­933. Taylor J, Jacobson D, Fisher M (1999) The evolution of asexual fungi: reproduction, speciation and classification. Annual Review of Phytopathology, 37, 197­246. Templeton AR, Routman E, Phillips CA (1995) Separating population structure from population history: a cladistic analysis of the geographical distribution of mitochondrial DNA haplotypes in the tiger salamander, Ambystoma tigrinum. Genetics, 140, 767­782. Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358­1370. Welz G, Kranz J (1987) Effects of recombination on races of powdery mildew populations. Plant Pathology, 36, 107­113. Yeh FC, Yang RC, Boyle TBJ, Ye ZH, Mao JX (1999) POPGENE, the user-friendly shareware for population genetic analysis. Molecular Biology and Biotechnology Centre, University of Alberta, Canada. Zhan J, Pettway RE, McDonald BA (2003) The global genetic structure of the wheat pathogen Mycosphaerella graminicola is characterized by high nuclear diversity, low mitochondrial diversity, regular recombination, and gene flow. Fungal Genetics and Biology, 38, 286­297.

This work is part of the PhD research of Eva H. Stukeubrock. The aim of the project is to understand the evolution and population genetics of the fungal plant pathogen Phaeosphaeria nodorum. Søren Banke's research focuses on the inference of historical population events using phylogenetic and coalescence tools. This work is a part of ongoing research to better understand the evolutionary and population biology of fungal plant pathogens in Bruce McDonald's lab at the Swiss Federal Institute of Technology (ETH), Zurich.

© 2006 Blackwell Publishing Ltd, Molecular Ecology, 10.1111/j.1365-294X.2006.02986.x

Information

10 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

1202727

You might also be interested in

BETA
Septoria and Stagonospora Diseases of Cereals: A Compilation of Global Research