Current address: Tasmania Institute of Agricultural Research, University of Tasmania – Cradle Coast Campus, Burnie, Tasmania 7320, Australia
Genotypic diversity in Fusarium pseudograminearum populations in Australian wheat fields
Article first published online: 28 DEC 2009
© 2009 CSIRO
Volume 59, Issue 2, pages 338–347, April 2010
How to Cite
Scott, J. B. and Chakraborty, S. (2010), Genotypic diversity in Fusarium pseudograminearum populations in Australian wheat fields. Plant Pathology, 59: 338–347. doi: 10.1111/j.1365-3059.2009.02219.x
- Issue published online: 28 FEB 2010
- Article first published online: 28 DEC 2009
- Published online 28 December 2009
- crown rot of wheat;
- fusarium head blight;
- population genetics;
- sexual recombination;
- spatial autocorrelation
Despite being closely related to Fusarium graminearum, which has been extensively characterized in many countries around the world, the population biology of Fusarium pseudograminearum, the main causal agent of crown rot of wheat in Australia and many other wheat-growing regions, has been comparatively poorly studied to date. A simple sequence repeat analysis of 163 F. pseudograminearum isolates from three field sites in NSW, Australia identified 128 distinct multilocus genotypes. Observed genetic diversity within fields was high, whilst genetic variation between fields was low. Across all fields genetic linkage disequilibrium was detected, but of the three individual fields, only one also displayed linkage disequilibrium. These results indicate that the isolates obtained were part of the same, highly diverse, population. However, this population may not be freely interbreeding. Whilst isolation incidence of F. pseudograminearum was found to be spatially aggregated within fields, spatial aggregation of genotypes within fields was weak. The study suggests that processes influencing population dynamics may operate at a scale larger than the narrow geographical scale covered in the fields sampled.
Fusarium pseudograminearum (teleomorph Gibberella coronicola) is a destructive pathogen of small grain cereals, especially wheat. It is principally known for causing crown rot of wheat, predominantly in Australia, although recent reports suggest it occurs throughout most of the world’s cereal-producing regions (Nicol et al., 2004; Monds et al., 2005; Smiley et al., 2005a,b). In Australia, crown rot is a chronic problem, particularly in the north-eastern and southern growing regions (Williams et al., 2002). The increased importance of durum varieties, which are highly susceptible to crown rot, has exacerbated the problem in South Australia (Wallwork et al., 2004). Crown rot has gained greater importance in the USA through recent reports suggesting that commercial winter wheat fields in the Pacific North-West could have their grain yield reduced by up to 35% by crown rot (Smiley et al., 2005b). In the same region, field inoculation with F. pseudograminearum demonstrated potential yield losses up to 61% greater than those caused by the native pathogen flora (Smiley et al., 2005b). Additionally, F. pseudograminearum was demonstrated to be a significant cause of the disease fusarium head blight (FHB) in Australia, under favourable conditions (Burgess et al., 1987; Southwell et al., 2003). FHB is a disease that has re-emerged as a significant, if sporadic, threat to global wheat production (Goswami & Kistler, 2004). FHB epidemics are correlated with high economic losses resulting from reduced grain quality and quantity.
Fusarium pseudograminearum has only achieved species recognition in the last decade. Previously described as F. graminearum Group 1 (Burgess et al., 1975), DNA sequence (Aoki & O’Donnell, 1999a), amplified fragment length polymorphism (AFLP; Benyon et al., 2000) and isozyme (Laday et al., 2000) data were used to distinguish F. pseudograminearum from F. graminearum (formerly F. graminearum Group 2). Another key characteristic differentiating F. pseudograminearum from F. graminearum is its heterothallic mating system (Aoki & O’Donnell, 1999b). Individuals possess only one of two mating-type idiomorphs (MAT1-1 or MAT1-2) within their genome (Bentley et al., 2008a). Sexual recombination is uncommon in culture and rarely observed under field conditions (Summerell et al., 2001). Bentley et al. (2008b) observed low levels of female fertility within individuals in laboratory studies, but were able to artificially increase female fertility of individuals to allow mating studies to be undertaken. However, in the absence of such studies, the extent to which sexual recombination affects the evolutionary dynamics of F. pseudograminearum is unclear.
Fusarium pathogens of wheat have been subject to significant research in the last decade, but the majority of study, including population dynamics, has been conducted on F. graminearum (Miedaner et al., 2008). Only a few studies have dealt with F. pseudograminearum populations, despite its distinct mating strategy and epidemiological processes. To date, studies have typically utilized culture collections, or samples collected from widespread geographic regions (Akinsanmi et al., 2006a; Mishra et al., 2006), not always from the same growing season. These studies have generally found a high level of genotypic diversity within a field site. One recent study, using isolates from only three 1-m rows at each site, found physical clustering and aggregation of F. pseudograminearum isolates from two of the three field sites sampled (Bentley et al., 2009). Inferences on evolutionary processes are difficult to make from datasets containing individuals from extensive geographical areas, that may not share a common gene pool (McDonald & McDermott, 1993). Miedaner et al. (2008) noted that comparisons between and within field populations are at the most appropriate scale for making such inferences, preferably using a random, hierarchical sampling strategy. This study aimed to test the hypothesis that individual field populations of F. pseudograminearum constitute components of a larger, panmictic population. To accomplish this, three wheat fields in close proximity to one another in northern New South Wales (NSW) were hierarchically sampled and examined for genetic diversity within and between fields, using simple sequence repeat (SSR) markers, to elucidate the evolutionary processes shaping the F. pseudograminearum population.
Materials and methods
In November 2004, 100 wheat plants at maturity (Zadok’s growth stage 9·1), displaying crown rot symptoms, were collected from each of three locations in northern NSW, within the northeast wheat-growing region of Australia. Sites were situated at Spring Ridge (Spring Ridge 1: S31·33091 E150·17531; and Spring Ridge 2: S31·34181 E150·18028) and Pine Ridge (S31·4652 E150·47651). The two Spring Ridge sites were separated by approximately 1 km, and were approximately 30 km from the Pine Ridge site. Each field was sampled using a hierarchical grid pattern consisting of four parallel transects spaced at 10-m intervals. Five sampling units, at 10-m intervals, were marked along each transect. Each sampling unit was defined as a circle of 1-m radius, from within which five plants were randomly collected from the available diseased plants. The total dimensions of the sampling area at each field were 30 m × 40 m.
Following collection, the eldest tiller was selected for further processing. Fungal isolations were made by removing three, 2-mm-diameter sections from the crown tissue. These were surface-sterilized in bleach (1% available chlorine) for 5 min, and rinsed twice in sterile water for 5 min each time. Crown pieces were blotted dry on sterile filter paper. Crown pieces were then transferred onto quarter-strength potato dextrose agar (PDA) plate cultures containing 100 μg streptomycin sulphate and 10 μg tetracycline hydrochloride mL−1. Cultures were incubated at ambient light and temperature for 7–10 days. Plates then were flooded with sterile distilled water and the resultant spore suspension streaked onto water agar plates using a flame-sterilized metal loop. Water-agar cultures were maintained under ambient conditions for 12–24 h. A single germinating macroconidium was transferred to a fresh quarter-strength PDA plate for each isolate. Monoconidial cultures were maintained at 25°C under ambient light for 14 days, before storage as 5-mm-square agar plugs under 15% glycerol at −80°C.
Isolates were recovered from storage and DNA was extracted using a modified cetyltrimetholammonium bromide (CTAB) protocol. Fungal mycelia (50 mg) were scraped from 10-day-old PDA cultures and ground manually in a sterile 1·5-mL microcentrifuge tube with a sterile micropestle. Five hundred microlitres of TES lysis buffer (100 mm Tris, pH 8·0; 10 mm ethylenediaminetetraacetate (EDTA), pH 8·0; 2% (weight/volume; w/v) sodium dodecyl sulphate), pre-warmed to 60°C, and 50 μg proteinase K were added to the ground material. The resultant suspension was incubated at 60°C for 60 min. Subsequently, 140 μL 5 m NaCl and 64 μL 10% w/v CTAB were added and the suspension incubated at 65°C for 10 min. An equal volume of chloroform:isoamyl alcohol (24:1), was added to each tube and centrifuged at 14 000 g for 10 min and the aqueous phase transferred to a fresh tube. This step was repeated to improve DNA purity. DNA was precipitated by adding 0·6 volume of cold (4°C) isopropanol and 0·1 volume of 3 m sodium acetate (pH 5·2) and incubating samples at −20°C overnight. DNA was pelleted by centrifugation at 14 000 g for 30 min and the supernatant discarded. DNA pellets were washed twice, by adding 100 μL cold (4°C) 70% ethanol and centrifuging at 4000 g for 5 min, each time discarding the supernatant. DNA pellets were resuspended in 100 μL TE (10 mm Tris, pH 8·0; 1 mm EDTA, pH 8·0). RNA was digested by adding RNAse A at 10 mg mL−1 and incubating at 37°C for 45 min. Extractions were stored until use at −20°C.
Species identifications were based on a combination of cultural, morphological (Burgess et al., 1994; Aoki & O’Donnell, 1999a) and molecular characteristics (Aoki & O’Donnell, 1999a). Monoconidial isolates were recovered from storage and cultured on both PDA and carnation leaf agar (CLA; Burgess et al., 1994) and incubated at 25°C for 10–14 days under alternative periods of 12 h combined black light (F20T9BL-B 20W FL20S.SBL-B NIS) and standard fluorescent light (35098 F18E/33 General Electric), and 12 h darkness. Subsequently, colony morphology and pigmentation of PDA cultures, and macroconidium morphology from CLA cultures, were recorded. The identity of putative F. pseudograminearum isolates was confirmed by PCR amplification with the F. pseudograminearum-specific primer pair Fp1-1/Fp1-2 (Aoki & O’Donnell, 1999a). PCR reaction concentrations and temperature profiles are listed in Scott & Chakraborty (2006). Isolates identified as F. pseudograminearum were stored in the CSIRO Plant Industry culture collection under the accession numbers: Spring Ridge 1 (CS4147-9, CS4151, CS4154-5, CS4158-9, CS4161-8, CS4170-8, CS4180, CS4183-4, CS4189-92, CS4198, CS4201-7, CS4209-18, CS4220-31, CS4392, CS4397-440), Spring Ridge 2 (CS4232-5, CS4237, CS4242, CS4244-46, CS4253-4, CS4260, CS4263-70, CS4272, CS4275-9, CS4281-3, CS4285-6, CS4288, CS4295) and Pine Ridge (CS4298-300, CS4302-5, CS4307-13, CS4317, CS4319-21, CS4323-6, CS4330-1, CS4333-5, CS4337-8, CS4341, CS4346, CS4349, CS4353, CS4356-9, CS4361, CS4361, CS4365-6, CS4368-73, CS4375-87, CS4395-6).
Fusarium pseudograminearum isolates were genotyped using a set of 12 SSR markers identified from the F. graminearum genome (Broad Institute of MIT and Harvard http://www.broad.mit.edu) and confirmed as being present in F. pseudograminearum (Scott & Chakraborty, 2008; Table 1).
|Locus||Allelic range (bp)a||All populations||Spring Ridge 1||Spring Ridge 2||Pine Ridge|
SSR loci were amplified individually in PCR reactions containing 1 × PCR reaction buffer (67 mm Tris-HCl, pH 8·8; 16·6 mm (NH4)2SO4; 0·45% w/v Triton X-100; 0·2 mg gelatin mL−1), 2·5 mm MgCl2, 240 nm primers, 200 μm dNTPs, 1·0 U Taq polymerase (Biotech Int.) and 25 ng target DNA. Reaction volumes were made up to 15 μL with sterile distilled water. Forward-sense primers for each primer pair were labelled with WellRED dyes at the 5′ terminus (Table 1; Sigma-Aldrich). Amplifications were conducted with an initial denaturation at 94°C for 3 min, followed by 30 cycles of denaturation at 94°C for 45 s, annealing for 45 s and extension at 72 °C for 2 min, then concluded with a final extension phase at 72°C for 7 min. Annealing temperatures were dependent upon primer set (Table 1). Amplification was confirmed by gel electrophoresis in 1% agarose gels in 0·5 × TBE (45 mm Tris-borate; 1 mm EDTA), containing 0·5 mg ethidium bromide L−1 and UV light illumination. Sizes of amplicons were determined by capillary separation using the CEQ 8000 Genetic Analysis System in the presence of the GenomeLab DNA 400 Base Pair Size Standard (Beckman Coulter, Inc.). Alleles were automatically determined using the CEQ 8000 Genetic Analysis System and manually edited where necessary.
Prior to amplification and separation, DNA extracts from individual isolates were randomly allocated to positions within a set of 96-well plates. In addition, water controls and duplicate DNA extracts were randomly positioned within each plate set as checks.
The number of distinct multilocus genotypes (MLG) within each field and across the entire dataset was determined using genclone v1.0 software (Arnaud-Haond & Belkhir, 2007). Following MLG identification, three datasets were constructed: one in which all isolates were maintained (original), and two that were corrected for the presence of putative clones, whereby only a single isolate representing each MLG present in each field (field-clone-corrected), or transect (transect-clone-corrected), was retained. Except where indicated, all further analyses were conducted on the field-clone-corrected dataset.
Haploid diversity of individual loci within fields and across all fields, mean field haploid diversity and Nei’s unbiased estimated of genetic identity between fields were estimated using genalex v. 6.2 software (Peakall & Smouse, 2006).
Multilocus linkage disequilibrium, both within fields and across the entire dataset, was analysed by calculating the index of association (IA) (Maynard Smith et al., 1993), Agapow & Burt’s (2001) statistic () and Gordon’s (1997) test statistic for haploid organisms (VO/VE). represents a standardized version of IA, values of which, unlike IA, do not increase with increasing numbers of loci present in a dataset, and instead range between 0, indicating no linkage disequilibrium, and 1, complete linkage disequilibrium (Agapow & Burt, 2001). Both IA and were determined using multilocus v. 1.2 software (Agapow & Burt, 2001) and the significance of values greater than 0 was estimated by 1000 randomly recombined datasets, with the location of missing data points fixed within each dataset. VO/VE was calculated in genalex, and 999 randomly re-ordered datasets were created to test for values significantly greater than 0. In addition, linkage disequilibrium between pairs of individual loci, within fields and across the dataset, was calculated using genepop v. 3·4 (Raymond & Rousset, 1995). The significance of the interaction between loci pairs was tested using a Markov chain set of 100 000 with 1000 steps of dememorization, and a sequential Bonferroni correction for pairwise comparisons was applied (Rice, 1989).
To assess the extent of population differentiation, analysis of molecular variance (amova) was performed using genalex. Total genetic variance was partitioned into covariant components, accounting for allelic differences between individuals within fields and between individuals across different fields. The haploid analogue of Wright’s fixative index,ΦPT(field), was estimated as a measure of allelic differentiation between field populations. Relative gene flow between field populations was estimated by the statistic Nm. Additionally, amova was repeated using the transect-clone-corrected dataset, whereby total genetic variation was partitioned into covariance components covering variation within transects, between transects within fields and between fields.
Spatial clustering of F. pseudograminearum isolates was assessed by fitting the binomial and beta-binomial distributions to the isolation incidence data from each field (Hughes & Madden, 1993; Madden & Hughes, 1995). The binomial distribution has a single parameter (π), defined as the probability of each individual plant being diseased, and is thus indicative of a regular disease pattern (Madden & Hughes, 1995). The beta-binomial distribution has two parameters, p, the expected probability of disease, and θ, the variation in disease incidence between sampling units (Hughes & Madden, 1993). Therefore, the beta-binomial distribution is indicative of aggregated disease patterns. Goodness-of-fit of the two distributions was assessed by calculating the C(α) (Hughes & Madden, 1993) and log-likelihood ratio (LRS) statistics (Turechek & Madden, 1999), under the null hypotheses that the beta-binomial distribution did not significantly improve the goodness-of-fit to the data relative to the binomial distribution. Additionally, the index of dispersion (D), the ratio of observed variation to expected variation under the binomial distribution, was calculated (Madden & Hughes, 1995). A value of D significantly greater than 1·0 was evidence of overdispersion of the data. Distribution fitting and statistical testing was undertaken in r (R Development Core Team, 2009).
Spatial autocorrelation between F. pseudograminearum isolates and F. pseudograminearum genotypes were both assessed, by calculating Smouse & Peakall’s r, which is closely related to Moran’s I (Smouse & Peakall, 1999). For autocorrelation between isolates, an isolate distance matrix was constructed by comparing each pair of sampling points, giving a distance score of zero (0) when F. pseudograminearum was either present or absent at both sampling points and one (1) when F. pseudograminearum was present at one point only. For autocorrelation between genotypes, genetic distance matrices were constructed using the methods of Smouse & Peakall (1999), assuming a stepwise mutation model for SSR markers. For each analysis, complementary geographic distance matrices were constructed for each field and r calculated over a series of 10 distance classes set at 5-m intervals. Significance of autocorrelation was tested using both 9999 random permutations and 9999 bootstrap replicates. The null hypothesis of no autocorrelation was rejected in favour of autocorrelation for each individual lag only when r was greater than the upper 95% confidence interval calculated by random permutation and when the 95% error bars calculated by bootstrapping did not intersect the x-axis at r = 0. Individual field population analyses were combined and the overall significance of spatial autocorrelation across all populations was assessed using the methodology of Peakall et al. (2003). All calculations for spatial autocorrelation were conducted using genalex.
Spatial aggregation of MLG within individual fields at the sampling scale employed in this study was tested using the method of Milgroom et al. (1991), originally designed for analysis of vegetative compatibility groups. Briefly, two matrices, one of distance between all isolates collected and a second indicating the position of isolates of the same MLG, were constructed and multiplied together to produce the sum of the mean distances between identical MLG (Γ). Distance between isolates was based on the metric distance between sampling units from which isolates were obtained. The significance of Γ, relative to the sum of mean distance between all isolates, was calculated by 10 000 random permutations. The average distance between isolates of the same MLG was calculated by dividing Γ by MLGm,, where MLGm, is the total number of MLG represented by multiple isolates in a given field. All calculations were conducted in r.
All spatial analyses were conducted using the original dataset, except spatial autocorrelation of genotypes, which was conducted on both the original and clone-corrected datasets.
A total of 253 fungal isolates were obtained from 300 crowns sampled. Of these, 193 were confirmed as Fusarium spp. and 163 were identified as F. pseudograminearum using morphological and molecular characterization. Isolates other than F. pseudograminearum were not identified to species level. Spring Ridge 1, Spring Ridge 2 and Pine Ridge yielded 68, 33 and 62 F. pseudograminearum isolates, respectively (Table 1).
The number of alleles observed for individual loci varied between one (Fg2_5b) and 20 (Fg3_9a; Table 1). All alleles of the loci Fg2_5b (one allele) and Fg2_2f (two alleles) were observed in all field populations. Remaining loci were represented by a subset of their alleles within either one, or all, individual field populations. Haploid diversity of loci was greatest for Fg3_9a both across all fields and within each individual field. The monomorphic locus Fg2_5b was subsequently excluded from further analyses.
SSR genotyping identified 128 distinct multilocus genotypes (MLG) across all fields. In the Spring Ridge 1 population, 58 MLG were observed, with seven MLG represented by multiple isolates (Table 2). Of these seven MLG, five were represented by two isolates and one each by three and four isolates. Only two sampling units at Spring Ridge 1 provided more than one isolate of the same MLG. All 33 F. pseudograminearum isolates from Spring Ridge 2 constituted distinct MLG. From Pine Ridge, 58 MLG were obtained, 54 of which were represented only once. The remaining four MLG were all represented by two isolates. All sampling units at Pine Ridge contained no more than one representative of each MLG. Following clonal correction, 149 and 157 isolates remained within the field-clone-corrected and transect-clone-corrected datasets, respectively.
|Site||N a||MLG b||H c||IA (P)d||(P)e||Vo /Ve (P)f||Locus pairs in LDg|
|Spring Ridge 1||68||58||0·246||0·1849 (0·085)||0·0230 (0·085)||1·141 (0·234)|
|Spring Ridge 2||33||33||0·308||1·0461 (<0·001)||0·1096 (<0·001)||2·309 (<0·001)||1_1b/1_7b; 1_1b/3_3a; 1_1b/4_6a; 1_7b/3_3a; 1_7b/4_6a; 3_3a/4_6a|
|Pine Ridge||62||58||0·255||−0·1633 (0·952)||−0·0195 (0·952)||1·016 (0·829)|
|All populations||163||149||0·270||0·3596 (<0·001)||0·0387 (<0·001)||1·501 (<0·001)||1_1b/1_7b; 1_1b/3_3a; 1_1b/4_6a; 1_7b/3_3a; 1_7b/4_6a; 3_3a/3_9a; 3_3a/4_6a|
Mean haploid diversity (H) was greatest in Spring Ridge 2 (H = 0·308; Table 2), whilst H values for Spring Ridge 1 and Pine Ridge were closely matched (0·246 and 0·255, respectively). Within fields, haploid diversity of individual loci (h) ranged from zero to 0·846, 0·877, and 0·815, in Spring Ridge 1, Spring Ridge 2 and Pine Ridge, respectively. Nei’s unbiased genetic identity was 0·995 between fields Spring Ridge 1 and Spring Ridge 2, 0·997 between Spring Ridge 1 and Pine Ridge, and 0·989 between Spring Ridge 2 and Pine Ridge.
Significant multilocus linkage disequilibrium across all fields was detected by all three indices (IA, and Vo/Ve) used in this study (Table 2). However, when fields were analysed individually, significant linkage disequilibrium was only detected in the Spring Ridge 2 population. No evidence of linkage disequilibrium was provided by any index in Spring Ridge 1 or Pine Ridge.
Pairwise comparison between loci detected significant linkage disequilibrium in seven of 55 pairs of loci across all fields (P < 0·05; Table 2). Within individual fields, six of 55 pairs of loci were in linkage disequilibrium (P < 0·05) in Spring Ridge 2, but no significant pairs were detected in the remaining two fields.
amova indicated that 95% of the observed molecular variation was attributable to differences between isolates within fields (Table 3). The remaining 5% of variation was attributed to variation between fields.
|d.f.a||Variance||Variance (%)||Statistic||Value||P b|
|Between transects within fields||9||16·926||0·17||ΦPR(trans)||−0·002||0·419|
Pairwise comparisons of ΦPT(field) indicated significant differences between Spring Ridge 1 and Pine Ridge (0·077, P = 0·015), and Spring Ridge 2 and Pine Ridge (0·071, P = 0·040), but not between Spring Ridge 1 and Spring Ridge 2 (0·000, P = 0·342). Calculated effective numbers of migrants between fields were 216·9 between Spring Ridge 1 and Spring Ridge 2, 454·1 between Spring Ridge 1 and Pine Ridge and 115·8 between Spring Ridge 2 and Pine Ridge.
For amova conducted on the transect-clone-corrected dataset, 97% of variation occurred within transects, but the fixative index did not indicate that this variation was significant (Table 3). Another 3% of variation occurred between fields, and variation between fields was significant (ΦRT(trans) = 0·027, P = 0·01). Minimal variation between transects within fields was observed.
Indices of dispersion for each of the three fields were significantly greater than 1·0 (P ≤ 0·03; Table 4), indicating spatial aggregation of F. pseudograminearum isolates. The LRS test indicated that beta-binomial distribution provided a better fit to all data sets than the binomial distribution. The C(α) test concurred with the outcome of the LRS test for the Spring Ridge 1 and Spring Ridge 2 datasets, but was not significant (P = 0·09) for Pine Ridge.
|Population||nc||BBD analysisa||MLG spatial autocorrelationb|
|D (P)||C(α) (P)||LogLik (P)||MLGm||Γobs||Γexp||P|
|Spring Ridge 1||68||2·26 (0·0013)||4·39 (<0·001)||36·9 (<0·001)||7||139·1||158·4||0·15|
|Spring Ridge 2||33||1·83 (0·0145)||2·89 (0·0019)||51·4 (<0·001)||0||NA||NA||NA|
|Pine Ridge||62||1·69 (0·0305)||2·38 (0·0876)||59·5 (<0·001)||4||30||23·0||<0·0001|
Spatial autocorrelation of F. pseudograminearum isolates was observed overall and in each individual field in the smallest distance class (0–5 m; Fig. 1). Positive autocorrelation was also detected in the overall analysis in the distance class 5–10 m. Autocorrelation at this spatial lag was also observed in Spring Ridge 2, but not in Spring Ridge 1 or Pine Ridge.
Significant positive spatial autocorrelation of genotypes was not detected over any distance class tested, when field populations were combined into a single analysis in either the original or clone-corrected datasets (Fig. 1). Significant positive spatial autocorrelation was also lacking in all individual field populations over all distance classes in both datasets.
Spatial distance between identical genotypes was significantly less than expected under 10 000 random permutations in the Pine Ridge population (P < 0·0001; Table 4). However, no evidence of spatial aggregation of MLG was observed in the Spring Ridge 1 population. The Spring Ridge 2 population could not be assessed as all isolates collected belonged to individual MLG.
This is apparently the first report of SSR markers being used to study the population genetics of F. pseudograminearum. Previous molecular studies have been based around dominant marker systems, including amplified fragment length polymorphisms (AFLP; Akinsanmi et al., 2006a; Bentley et al., 2008a, 2009), restriction digest fragments (Mishra et al., 2006) and inter-simple sequence-repeats (ISSR; Mishra et al., 2006). Dominant markers are biallelic, thus less informative than multiallelic markers, such as SSR (Mariette et al., 2002). AFLP studies can also suffer with respect to SSR, as a proportion of the observed variation may be nonheritable or unrelated to the target organism (Sunnucks, 2000). These deficiencies can be overcome by increasing the number of marker loci scored (Mariette et al., 2002). Additionally, in this study, SSR markers were located on each of the four F. pseudograminearum chromosomes, and provided a wide coverage of the genome (Scott & Chakraborty, 2008). Because of their anonymous nature, this cannot be said with certainty for AFLP markers.
Sample collection was conducted from three sites within a single wheat-growing region in northern NSW in the 2005 growing season. At these sites, F. pseudograminearum infection of sampled wheat tillers ranged between 33 and 68%. Based on 11 polymorphic SSR markers, between 85 and 100% of F. pseudograminearum isolates collected per field constituted distinct MLG. This high level of diversity is consistent with previous studies of F. pseudograminearum populations. Using AFLP markers Bentley et al. (2008a) found that between 56 and 100% of isolates collected from individual fields were of distinct MLG. Likewise, a Canadian study using ISSR markers found that 100% of isolates collected were genetically distinct (Mishra et al., 2006).
However, despite the high proportion of genetic diversity observed within fields, genetic variation between fields was low. Pairwise comparisons between fields calculated genetic identities of 99% or greater, whilst the estimated number of migrants between fields ranged between 116 and 454. Likewise, amova on a per-field basis indicated that 95% of the observed genetic variation was attributable to variation occurring within fields. These results suggest that the three fields sampled in this study are part of a single metapopulation. This agrees with the work of Mishra et al. (2006), who found that 88 and 94% of variability present was within fields, based on restriction digestion of the intergenic spacer region and ISSR analysis, respectively. This extent of within-field variation is higher than the 50% observed by Bentley et al. (2008a), although the observed regional differences or the different marker systems employed (AFLP vs. SSR) in that study could account for this. The hierarchical sampling structure employed in the present study allowed for further breakdown of the genetic variance into within-transect and between-transect, components, in individual fields. In this analysis, 97% of variability occurred within individual transects, indicating that field populations of F. pseudograminearum in the study area are highly diverse at spatial scales much smaller than the field. However, minimal diversity was observed between transects within the same field, suggesting an even spread of diversity across the field populations.
High genetic diversity may be the result of several evolutionary mechanisms. Frequent genetic recombination, either sexual or parasexual, can result in a diverse range of allele combinations within a given population. Significant exchange of genotypes between field populations and/or high mutation rates can similarly result in high genetic diversity, especially if selective pressures are absent. In the case of F. pseudograminearum, the ability to alternate between saprophytic and parasitic life stages (Wearing & Burgess, 1977) may balance the effect of potentially differential selection pressures and thus maintain higher levels of polymorphism, even in the absence of recombination. Estimated gene flow between field populations in this study was high, but whether this was true gene flow, resulting in genetic recombination, or genotype flow, exchanging whole genotypes only, is unclear. Perithecia of F. pseudograminearum are only infrequently observed in field systems (Summerell et al., 2001, Bentley et al., 2008b), but the heterothallic nature of F. pseudograminearum means that a single recombination event can result in high levels of genetic diversity in the progeny (Bentley et al., 2005). Linkage disequilibrium analyses conducted across all fields were indicative of significant linkage disequilibrium. This is counter to the results of Mishra et al. (2006), who found no evidence of linkage disequilibrium. However, when fields were analysed individually, only Spring Ridge 2 was indicative of linkage disequilibrium, whilst the hypothesis of randomly mating populations could not be discounted in Spring Ridge 1 and Pine Ridge. Such inconsistencies have been reported previously (Bentley et al., 2008a), whereby a lack of significance indicating linkage disequilibrium was attributed to low numbers of clones analysed. However, the reverse situation occurred here, whereby the field with the lowest number of individuals was the only field to indicate significant linkage disequilibrium. Similar discrepancies were found by Naef & Defago (2006) who found seven out of 12 populations of F. graminearum in Switzerland exhibiting evidence of linkage disequilibrium. All populations were from the same wheat cultivar host, and the relatively close proximity of Spring Ridge 1 and Spring Ridge 2 would preclude climatic influences. It is possible that microclimatic conditions significantly influence the occurrence of sexual recombination, which would explain the discrepancy between fields. Alternatively, the extent of female infertility present within a given population (Bentley et al., 2008b), may be influential. In any case, the obtained results cast doubt that F. pseudograminearum in this region constitutes a panmictic population.
Spatial distribution of isolates was found to be aggregated in all fields sampled, which is as expected for a soilborne pathogen (Gosme et al., 2007) and is consistent with recent results obtained by Bentley et al. (2009). Spatial autocorrelation analysis indicated that the spatial scale of this aggregation was up to 10 m. However, only limited evidence of spatial clustering of F. pseudograminearum genotypes was observed. Whilst aggregation of MLG was indicated in the Pine Ridge population, this result must be treated with caution. In this field, only four MLG were represented multiple times, and in no instances was the same MLG isolated from within the same sampling unit. In Spring Ridge 1, where a greater number of MLG were present more than once, and on two occasions in the same sampling unit, the observed distance between identical MLG was not significantly less than expected under a random distribution. The vast majority of genetic variation occurred within transects, with minimal variation between transects according to amova, suggesting the absence of spatial clustering of genotypes. The fixative index for within transects was not significant, but this result was backed by the absence of spatial autocorrelation based on genetic similarity, in any field, even at the smallest lag distance. This contradicts the results of Bentley et al. (2009), who found evidence of clustering of clonal haplotypes. However, in that study, sampling for spatial analysis was undertaken within a single, one-dimensional, 1-m transect per field. Sampling of this nature would maximize the chance of detecting clonal isolates of a soilborne pathogen. If soilborne spread of clonal isolates was the only method by which F. pseudograminearum propagated itself over multiple seasons, it would be expected that spatial aggregation of genotypes would be detected at the larger, two-dimensional sampling scale employed in the present study. As this was not the case, it is therefore concluded that other mechanisms of dispersal, and/or sexual recombination, must occur to minimize the aggregation of clonal genotypes. Potential mechanisms of dispersal over longer distances could include airborne inoculum, or human-mediated movement of infested seed or crop residues. The role of airborne macroconidia and/or ascospores has yet to be adequately studied (Summerell et al., 2001).
Fusarium pseudograminearum has a significant saprophytic phase in its lifecycle (Wearing & Burgess, 1977). It must be noted that F. pseudograminearum was only isolated from diseased wheat plants in this study and as such is biased towards the pathogenic proportion of the F. pseudograminearum population. It may be of interest in the future to compare the genetic diversity of individuals in saprophytic and parasitic life stages, at a given point in time within field populations.
The implication that sexual recombination plays a role in the population dynamics of F. pseudograminearum agrees with the observations of Bentley et al. (2008b), who observed the presence of perithecia at field sites. Coupled with high pathogenic (Akinsanmi et al., 2006b) and genetic variation, this indicates that F. pseudograminearum has a great potential to adapt to different selection pressures, which should be considered in the implementation of disease control strategies. If highly resistant host cultivars reliant upon only one or a few resistance genes are widely used in crop rotations, the ability of the pathogen population to evolve to mitigate the effect of these genes cannot be discounted. Additionally, pathosystems that involve cycles of recombination followed by selection of subsequent asexual progeny, have the potential to erode a host’s quantitative resistance over time (McDonald & Linde, 2002). Fungicide use, although not common for crown rot, likewise represents a similar issue for growers. Increased resistance to benzimidazole fungicides in F. graminearum has been reported in China (Gale et al., 2002). Therefore, the inherent variability of F. pseudograminearum should be taken into account in current resistance breeding programmes to ensure the long-term viability of new varieties, and in the development of viable long-term disease control strategies.
We wish to thank Steven Simpfendorfer and Ross Perrott for their assistance with the collection of isolates used in this study. Funding for this work was provided by the Grains Research and Development Corporation of Australia, the CRC for Tropical Plant Protection and CSIRO Plant Industry.
- 2001. Indices of multilocus linkage disequilibrium. Molecular Ecology Notes 1, 101–2. , ,
- 2006a. Genetic diversity of Australian Fusarium graminearum and F. pseudograminearum. Plant Pathology 55, 494–504. , , , ,
- 2006b. Pathogenic variation of Fusarium isolates associated with head blight of wheat in Australia. Journal of Phytopathology 154, 513–21. , , , ,
- 1999a. Morphological and molecular characterization of Fusarium pseudograminearum sp. nov., formerly recognized as the Group 1 population of F. graminearum. Mycologia 91, 597–609. , ,
- 1999b. Morphological characterization of Gibberella coronicola sp. nov., obtained through mating experiments of Fusarium pseudograminearum. Mycoscience 40, 443–53. , ,
- 2007. GENCLONE: a computer program to analyse genotypic data, test for clonality and describe spatial clonal organization. Molecular Ecology Notes 7, 15–7. , ,
- 2005. Gibberella coronicola: an imperfectly understood teleomorph. Proceedings of the 15th Biennial Australasian Plant Pathology Society Conference, 2005, Geelong, Australia, 382. , , , ,
- 2008a. Genetic structure of Fusarium pseudograminearum populations from the Australian grain belt. Phytopathology 98, 250–5. , , , , ,
- 2008b. Sexual compatibility in Fusarium pseudograminearum (Gibberella coronicola). Mycological Research 112, 1101–6. , , ,
- 2009. Spatial aggregation in Fusarium pseudograminearum populations from the Australian grain belt. Plant Pathology 58, 23–32. , , , , ,
- 2000. Molecular genetic investigations and reclassification of Fusarium species in the sections Fusarium and Roseum. Mycological Research 104, 1164–74. , , ,
- 1975. Surveys of fusaria associated with crown rot of wheat in eastern Australia. Australian Journal of Agricultural Research 26, 791–9. , , ,
- 1987. Head blight of wheat caused by Fusarium graminearum group 1 in New South Wales in 1983. Australasian Plant Pathology 16, 72–8. , , , ,
- 1994. Laboratory Manual for Fusarium Research. Sydney, Australia: Fusarium Research Laboratory, University of Sydney and Royal Botanic Gardens. , , , , ,
- 2002. Population analysis of Fusarium graminearum from wheat fields in eastern China. Phytopathology 92, 1315–22. , , , , ,
- 1997. The genetic structure of Escherichia coli populations in feral house mice. Microbiology 143, 2039–46. ,
- 2007. Size, shape and intensity of aggregation of take-all disease during natural epidemics in second wheat crops. Plant Pathology 56, 87–96. , , ,
- 2004. Heading for disaster: Fusarium graminearum on cereal crops. Molecular Plant Pathology 5, 515–25. , ,
- 1993. Using the beta-binomial distribution to describe aggregated patterns of disease incidence. Phytopathology 83, 759–63. , ,
- 2000. Isozyme evidence for two groups of Fusarium graminearum. Mycological Research 104, 788–93. , , , ,
- 2002. Pathogen population genetics, evolutionary potential, and durable resistance. Annual Review of Phytopathology 40, 349–79. , ,
- 1993. Population genetics of plant pathogenic fungi. Bioscience 43, 311–9. , ,
- 1994. BBD – computer software for fitting the beta-binomial distribution to disease incidence data. Plant Disease 78, 536–40. , ,
- 1995. Plant disease incidence – distribution, heterogeneity, and temporal analysis. Annual Review of Phytopathology 33, 529–64. , ,
- 2002. Sampling within the genome for measuring within-population diversity: trade-offs between markers. Molecular Ecology 11, 1145–56. , , , ,
- 1993. How clonal are bacteria? Proceedings of the National Academy of Sciences, USA 90, 4384–8. , , , ,
- 2008. Population genetics of three important head blight pathogens Fusarium graminearum, F. pseudograminearum and F. culmorum. Journal of Phytopathology 156, 129–39. , , ,
- 1991. Spatial pattern analysis of vegetative compatibility groups in the chestnut blight fungus, Cryphonectria parasitica. Canadian Journal of Botany 69, 1407–13. , , ,
- 2006. Genetic diversity and recombination within populations of Fusarium pseudograminearum from western Canada. International Microbiology 9, 65–8. , , , ,
- 2005. Fusarium graminearum, F. cortaderiae and F. pseudograminearum in New Zealand: molecular phylogenetic analysis, mycotoxin chemotypes and co-existence of species. Mycological Research 109, 410–20. , , , , ,
- 2006. Population structure of plant-pathogenic Fusarium species in overwintered stalk residues from Bt-transformed and non-transformed maize crops. European Journal of Plant Pathology 116, 129–43. , ,
- 2004. Strategy for the identification and breeding of resistance to dryland root rot complex for international spring and winter wheat breeding programs. In: FischerT, TurnerN, AngusJ, McIntyreL, RoberstonM, BorrellA, LloydD, eds. Proceedings of the 4th International Crop Science Congress, 2004, Brisbane, Australia. New Directions for a Diverse Planet. 283. , , , , , ,
- 2006. GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6, 288–95. , ,
- 2003. Spatial autocorrelation analysis offers new insights into gene flow in the Australian bush rat, Rattus fuscipes. Evolution 57, 1182–95. , , ,
- R Development Core Team, 2009. R: a Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing.
- 1995. Genepop (version 1.2) – population genetics software for exact tests and ecumenicism. Journal of Heredity 86, 248–9. , ,
- 1989. Analyzing tables of statistical tests. Evolution 43, 223–5. ,
- 2006. Multilocus sequence analysis of Fusarium pseudograminearum reveals a single phylogenetic species. Mycological Research 110, 1413–25. , ,
- 2008. Identification of 11 polymorphic simple sequence repeat loci in the phytopathogenic fungus Fusarium pseudograminearum as a tool for genetic studies. Molecular Ecology Resources 8, 628–30. , ,
- 2005a. Pathogenicity of fungi associated with the wheat crown rot complex in Oregon and Washington. Plant Disease 89, 949–57. , , , ,
- 2005b. Crop damage estimates for crown rot of wheat and barley in the Pacific Northwest. Plant Disease 89, 595–604. , , , , ,
- 1999. Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity 82, 561–73. , ,
- 2003. An outbreak of fusarium head blight of durum wheat on the Liverpool Plains in northern New South Wales in 1999. Australasian Plant Pathology 32, 465–71. , , , ,
- 2001. Natural occurrence of perithecia of Gibberella coronicola on wheat plants with crown rot in Australia. Australasian Plant Pathology 30, 353–6. , , , , ,
- 2000. Efficient genetic markers for population biology. Trends in Ecology & Evolution 15, 199–203. ,
- 1999. Spatial pattern analysis and sequential sampling for the incidence of leaf spot on strawberry in Ohio. Plant Disease 83, 992–1000. , ,
- 2004. Resistance to crown rot in wheat identified through an improved method for screening adult plants. Australasian Plant Pathology 33, 1–7. , , , ,
- 1977. Distribution of Fusarium roseum‘Graminearum’ Group 1 and its mode of survival in eastern Australian wheat belt soils. Transactions of the British Mycological Society 69, 429–42. , ,
- 2002. The application of species-specific assays based on the polymerase chain reaction to analyse fusarium crown rot of durum wheat. Australasian Plant Pathology 31, 119–27. , , , ,