SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Polymerase chain reaction (PCR) assays for the detection of various Fusarium species and Microdochium nivale subspecies were compared with conventional visual disease assessment using a field plot of wheat in which the central subplot was inoculated with F. culmorum. Visual disease assessment was performed on a range of samples taken from each of 15 subplots at growth stage 80. At harvest, each sample was divided into its component parts, i.e. grain, glume and rachis, and species-specific PCR analysis was used to detect the presence of F. culmorum, F. poae, F. avenaceum, F. graminearum, M. nivale var. majus and M. nivale var. nivale. Within the inoculated subplot there was good correlation between visual disease assessment and PCR analysis, both techniques indicating a high incidence of F. culmorum in this region. According to the visual disease assessment results, there was also a relatively high incidence of F. culmorum in most other regions of the field plot. However, according to PCR analysis the incidence of F. culmorum in many of the other subplots was relatively low and F. poae, M. nivale var. majus and var. nivale, and F. avenaceum were detected within the grain, glume and rachis tissues of many of the ear samples from these subplots. F. poae predominated in the glume component of ears and M. nivale var. majus and var. nivale in the rachis component. M. nivale PCR results revealed that 64% of infected samples involved var. majus, and 36% var. nivale. PCR analysis has highlighted some difficulties that may arise when using visual assessment for studying disease complexes.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Fusarium ear blight (FEB) of wheat has been linked to at least 17 causal organisms, most records of the disease being associated with five species: Fusarium culmorum, F. avenaceum (Gibberella avenacea), F. graminearum (G. zeae, formerly known as G. saubinetii), F. poae and Microdochium nivale (Monographella nivalis, formerly known as F. nivale) (Parry et al., 1995a). F. culmorum, F. poae and M. nivale are regarded as important causal agents of FEB in the cooler maritime regions of north-west Europe, whereas F. graminearum assumes a greater importance in hotter regions of the world such as parts of the USA, Canada, Australia and central Europe. F. poae was the most frequently isolated species in a UK survey of affected ears carried out in 1989 and 1990 (Polley et al., 1991).

This complex of causal organisms greatly complicates the study of FEB of wheat and conventional techniques, such as visual disease assessment and culture methods, require a degree of taxonomic expertise to distinguish the pathogens at the species level. Visual disease assessment is based on the recognition of the classical symptoms of FEB, such as premature bleaching of spikelets (Wiese, 1987). However, there has been controversy regarding the symptoms of FEB produced by some of the causal agents (Rapilly et al., 1973; Cassini, 1981; Inglis & Cook, 1981; Parry et al., 1995a). This had led to increasing interest in obtaining more sensitive methods for the identification of Fusarium species in infected plant material, including techniques involving molecular diagnosis. Several such diagnostic assays have been developed for the identification and detection of the major fungi involved in FEB of wheat (Höxter et al., 1991; Koopman et al., 1994; Ouellet & Siefert, 1993; Parry & Nicholson, 1996; Nicholson et al., 1996). Most of these are based on the polymerase chain reaction (PCR) and include diagnostic assays specific for F. graminearum (Ouellet & Siefert, 1993; Schilling et al., 1996), F. poae (Parry & Nicholson, 1996) and F. culmorum and F. avenaceum (Schilling et al., 1996). Two varieties are recognized within M. nivale, var. majus and var. nivale (Lees et al., 1995) and species-specific assays have been developed for each of these (Nicholson & Parry, 1996; Nicholson et al., 1996). Using these assays, it is possible to identify and distinguish between the individual pathogens within the disease complex.

In the present work, samples obtained from a field trial inoculated with F. culmorum were analysed by PCR and the results related to visual disease assessment.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Field trial

A field experiment was conducted during the 1993/94 season at Harper Adams Agricultural College, Shropshire. The winter wheat cultivar Avalon was used because of its susceptibility to Fusarium ear blight pathogens (authors' unpublished observations) and the field plot was 30 × 2 m, subdivided into 15 subplots, each 2 × 2 m. Wheat was sown early in October and was kept free from weeds and foliar diseases by application of the appropriate agrochemicals, according to manufacturers' recommendations. At mid-anthesis (GS 65) the central subplot (subplot 8) was inoculated with 167 mL of a conidial suspension of F. culmorum (strain Fu 42) at a rate of 2.1 × 107 conidia per m2 using a pressurized hand sprayer. Following inoculation, the field trial was mist-irrigated until harvest. Visual disease assessment (based on percentage infected spikelets per ear) was performed at GS 80 on a range of samples randomly chosen from each subplot, numbers of samples varying according to subplot (Table 1). FEB symptoms scored were premature bleaching of ears, formation of small pink grains and, in severe cases, mycelial growth. At GS 90, those wheat ears examined at GS 80 were harvested and separated into grain, glume and rachis tissue; DNA was extracted and PCR analysis was used to confirm visual diagnosis.

Table 1.  Field trial sampling plan aBased on visual disease assessment results.bSubplot inoculated with F. culmorum (Table 2) at GS 65 (2.1 × 107 conidia per m2).Thumbnail image of

Origin and maintenance of fungal isolates

DNA from isolates of Fusarium species and M. nivale subspecies obtained from the John Innes Centre facultative pathogen culture collection (Table 2) were used as positive controls for PCR analysis. The isolates were maintained on potato dextrose agar (PDA) (Difco, UK) containing Penicillin G (50 μg mL−1) and streptomycin sulphate (100 μg mL−1). For DNA preparations, mycelium from 7-day-old colonies grown on PDA were used to inoculate aseptically 50 mL of potato dextrose broth (PDB), using a sterile scalpel. PDB cultures were incubated on an orbital shaker at 20–22°C, 110 r.p.m., for 7 days. Fusarium culmorum was grown at 15°C on 1% agar (w/v) containing milled wheat straw for the production of conidia. After 7 days conidia were washed from plates with sterile distilled water and adjusted to 5 × 105 conidia per mL for inoculation of the field trial.

Table 2.  Code and origin of fungal species Thumbnail image of

DNA extraction

Wheat ears were harvested, separated into glume, grain and rachis tissue, freeze-dried and the dry weight was recorded. Rachis material was ground to a fine powder, which was incubated at 65°C for 2 h in 7 mL CTAB buffer (sorbitol 2.3 g, n-lauryl sarcosine 1.0 g, hexadecyl trimethyl-ammonium bromide 0.8 g, sodium chloride 4.7 g, polyvinylpolypyrolidone 1.0 g, water to 100 mL) together with 15 μL proteinase K (10 mg mL−1) and 10 μL RNAase (10 mg mL−1). Following incubation, an equal volume of chloroform/isoamyl alcohol (24 : 1) was added to the tubes, mixed, and centrifuged at 2600 g for 15 min. The aqueous phase was removed to a fresh tube and two volumes of ethanol (100%) were added, followed by centrifugation as above to precipitate the DNA. The pellet was washed in a 70% solution of cold ethanol and dissolved in TE buffer (10 mM Tris-HCl, 1 mM EDTA).

The method used for DNA extraction from glume and grain samples was similar except that 20 mL of CTAB buffer together with 50 μL of proteinase K and 30 μL of RNAase were used. Because of the presence of substances that inhibited the PCR reaction, glume samples were also subjected to a phenol–chloroform (1 : 1) extraction step prior to chloroform–isoamyl alcohol extraction, using the same procedure. DNA from all tissues was diluted in TE for use in PCR amplification reactions (1 mg dry weight equivalent per 2.5 μL). DNA was extracted from fungal isolates by a method similar to that described by Nicholson & Parry (1996). Mycelium was harvested onto Whatman no.1 filter paper and DNA was extracted as described for rachis material. Fungal DNA was diluted to 10 ng μL−1 in TE for use in PCR reactions.

PCR amplification and agarose gel electrophoresis

Amplification conditions were similar to those described by Nicholson & Parry (1996). Reactions were performed in volumes of 50 μL and contained DNA from 0.8 mg dry weight of plant material or 10 ng of fungal DNA. The reaction buffer consisted of 100 μM each of dATP, dCTP, dGTP and dTTP, 100 nM each of forward and reverse primer for PCR reactions, and 0.8 units of Taq polymerase (Boehringer Mannheim Ltd, Germany) in 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2, 50 mM KCl, 100 μg mL−1 gelatine and 0.05% (w/v) each of Tween 20 and Nonidet P-40. Primers used were those for F. culmorum/F. graminearum (Fc F/R: CAAAAGCTTCCCGAGTGTGTC/GGCGAAGGTTCAAGGATGAC) (Lees, 1995), F. graminearum (Fg11F/R: CTCCGG ATAT-GTTGCGTCAA/GGTAGGTATCCGACATGGCAA) (Nicholson et al., unpublished observations), F. poae (Fp82F/R: CAAGC AAACAGGCTCTTCACC/TGT TCCACCTCAGTGACAGGTT) (Parry & Nicholson, 1996), F. avenaceum (AF/R: CA AGCATTGTCGCCACTCTC/GTTTGG-CTCTACCGGGACTG) (Lees, 1995), M. nivale var. majus (Mnm2F/R: TGCAACGTGCCAGAAGCT/AATCGG-CGCTGTCTACTAAA AGC) (Nicholson & Parry, 1996) and M. nivale var. nivale (Y13NF/R: ACCAGCCGA TTTGTGGTTATG/GGTCACGAGGCAGAGTTCG) (Nicholson et al., 1996). Reaction mixtures were overlaid with mineral oil prior to PCR.

Amplification was performed in a Perkin-Elmer Cetus 480 DNA thermal cycler. The programme used to amplify fungal DNA from infected plant samples varied depending on the specific primers. With F. culmorum/F. graminearum or F. avenaceum-specific primers the cycler was programmed for 40 cycles of 30 s at 95°C, 30 s at 60°C and 40 s at 72°C. Programmes for M. nivale var. nivale, M. nivale var. majus, F. graminearum and F. poae primers were similar except that annealing temperatures of 61, 61, 62 and 62°C, respectively, were used. Aliquots (15 μL) of amplification products were electrophoresed through agarose gels (1.5% w/v), prepared using TAE buffer (40 mM Tris base, 1 mM EDTA, 20 mM acetic acid) and containing 0.05 mg ethidium bromide per 100 mL TAE buffer.

Statistical analysis

Statistical analysis consisted of determination of the correlation coefficients between visual disease assessment and PCR-based assays using the Pearson Product Moment Correlation of arcsine-transformed subplot frequency data. This analysis was performed using Minitab release 10.1© (1994, Minitab Inc., USA). The association between pathogens was tested using Fisher's exact test (Everitt, 1986) for which a correction according to Rom (1990) was used.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Visual disease assessment

This experiment was part of a larger field trial, set up to relate disease severity to yield loss, in which the sample size harvested per subplot increased with proximity to the inoculation zone. As a result, the number of harvested samples available per subplot for the present study varied accordingly. Visual disease assessment was performed at GS 80 on a range of single ear samples from each of the 15 subplots within the field plot (Table 1 and Figs 1a and b) . According to these results, the mean visual disease score attributed to the wheat ears taken from within the F. culmorum-inoculated subplot (subplot 8, samples 56–68) was 96% of spikelets infected by F. culmorum, the majority (77%) of ears from this subplot being given scores of 100% spikelets infected. In the other subplots (subplots 1–7, samples 1–55 and subplots 9–15, samples 69–123), although disease score varied from 0 to 100%, the majority of ears showed disease symptoms. In the subplot adjacent to the inoculated subplot (subplot 9, samples 69–81), 12 of the 13 ears scored had disease ratings of 50% or more (mean = 70%) (Fig. 1). In the other adjacent subplot (subplot 7, samples 43–55), disease severity was much reduced and no wheat ear had more than 35% spikelets infected (mean = 20%). The disease severities for subplots 1–6 were low to moderate, mean disease scores being 2, 5, 21, 10, 12 and 12%, respectively. Only one sample from this region (sample 8, subplot 3) had a relatively high disease severity (60%). The disease severities for the corresponding subplots on the other side of the inoculated subplot (subplots 10–15, samples 82–123) were generally higher, mean disease scores being 18, 31, 17, 19, 50 and 28%, respectively. Compared with subplots 1–6, subplots 10–15 showed greater variation in disease severity from sample to sample, disease scores fluctuating between 0 and 100%. Several samples from these subplots had disease severities greater than 50% (i.e. samples 90, 96, 99, 103, 116, 119, 120 and 122 from subplots 10, 11, 11, 11, 13, 14, 14 and 15, respectively). Therefore, according to the visual disease assessment results, F. culmorum was detected in each subplot within the field trial. With the exception of the inoculated subplot (subplot 8), an adjacent subplot (subplot 9) and subplot 14, the mean disease severities of subplots were low to moderate, with no significant evidence of the development of a disease severity gradient from the inoculated subplot (Fig. 1b).

image

Figure 1. Visual disease assessment results of ear blight of wheat in a field plot in which the central subplot (samples 56–68) was inoculated with conidia of Fusarium culmorum (2.1 × 107 conidia per m2) at GS 65. Results expressed as (a), disease score for each sample and (b) mean disease score per subplot of the field trial (bars indicate standard error of the mean).

Download figure to PowerPoint

The visual disease results were used to classify each single ear sample as infected or not infected by F. culmorum (scores of 1 and 0, respectively) enabling comparison with the PCR results on the basis of the percentage of infected samples per subplot. The combined results for the samples within each subplot are shown in Table 1, as are the number of samples taken from each subplot. Based on these results, all of the samples in the inoculated subplot appeared to be infected by F. culmorum and, even in the noninoculated subplots, the frequency of infected samples was moderate to high (Table 1 and Fig. 2a). For example, 75% of samples examined from subplot 14 exhibited head blight symptoms, presumed to be caused by F. culmorum

image

Figure 2. Comparison of the frequency of the occurrence of visual disease symptoms (□—□) with the frequency of PCR detection of Fusarium culmorum (•——•) in grain tissue of ear samples from each subplot (a), and comparison of the frequency of PCR detection of Microdochium nivale var. majus (○——○) and var. nivale (◊—◊) and F. poae (▴—▴) in the corresponding grain (b), glume (c) and rachis (d) components of ear samples from each subplot.

Download figure to PowerPoint

F. culmorum and F. graminearum PCR analysis

Fusarium graminearum-specific PCR analysis (primers Fg11F/R) did not detect this pathogen in any of the samples from the field plot, whether from grain, glume or rachis tissue. Therefore, it was concluded that any PCR amplification that occurred when using the F. culmorum/F. graminearum-specific primers (FcF/R) was due to the presence of F. culmorum DNA. As with visual disease assessment results, the PCR results were used to classify the ears within each subplot as diseased or nondiseased, and the results obtained for grain samples are shown in Fig. 2(a). Similar results were obtained for glume and rachis samples (results not shown). Within the inoculated subplot (subplot 8) there was a high incidence of F. culmorum (100% samples infected). PCR detected F. culmorum in the grain, glume and rachis of every sample from this subplot. According to the PCR analysis, the incidence of F. culmorum outside subplots 8 and 9 was relatively low, although most of the ears outside the inoculated subplot that had disease scores of 50% or more were found to contain F. culmorum by PCR analysis (samples 8, 55, 69–73, 75–80, 90, 96, 99, 103, 116, 119 and 122 from subplots 3, 7, 9, 9, 10, 11, 11, 11, 13, 14 and 15 with disease scores of 60, 100, 100, 63, 90, 79, 100, 90, 50, 100, 60, 68, 95, 68, 100, 55, 75, 100 and 80%, respectively). The results of F. culmorum/F. graminearum-specific PCR analysis of some of these ‘high score’ samples are shown in Fig. 3(a). Of these 20 samples from outside the inoculated subplot and given high disease scores, 11 were among the 13 of the samples taken from subplot 9 adjacent to the inoculated subplot. PCR analysis detected F. culmorum in the grain, glume and rachis of all of these 11 samples, with the exception of sample 75 for which the pathogen was detected in the glume and rachis, but not in the grain tissue. The other two samples taken from subplot 9 had lower disease scores (samples 74 and 81, disease scores 5 and 26%, respectively) and PCR analysis did not detect F. culmorum in sample 74, but detected the pathogen in the glume and rachis of sample 81. In the other subplot adjacent to the inoculated subplot (subplot 7), PCR analysis detected F. culmorum in only 8% of samples, but this 8% corresponded to the only sample from this subplot with a high disease score (100%). PCR analysis did not detect the pathogen in the other 92% of samples from subplot 7, which had disease scores between 0 and 33%.

image

Figure 3. Detection of the specific PCR signal for Fusarium culmorum/F. graminearum and F. avenaceum (a), Microdochium nivale var. nivale and var. majus (b) and F. poae and F. graminearum (c) in the corresponding grain, glume and rachis tissue of wheat ear samples. Lanes: 1–8, samples 8, 90, 96, 99, 103, 116, 119 and 122; Fc, Fa, M, N, Fp and Fg, F. culmorum, F. avenaceum, M. nivale var. majus, M. nivale var. nivale, F. poae and F. graminearum genomic DNA; c, control without fungal DNA. Arrows: 1–6, F. avenaceum, F. culmorum/F. graminearum, M. nivale var. majus, M. nivale var. nivale, F. graminearum and F. poae-specific DNA bands (920, 700, 750, 310, 300 and 250 bp, respectively).

Download figure to PowerPoint

Although FEB symptoms were observed in 30 of the 42 samples from subplots 1–6, PCR analysis detected F. culmorum in only a single sample (sample 8, subplot 3) where it was present in grain (Fig. 3a), glume and rachis tissue. Similarly, 32 of the 42 samples from subplots 10–15 had FEB symptoms while PCR analysis detected F. culmorum in only 13 of the samples from this region. F. culmorum PCR analysis detected F. culmorum in samples 82, 84, 85, 89–91, 93, 96, 99, 107, 119, 122 and 123 from subplots 10–15. Disease scores for samples 82, 89, 93 and 123 were relatively low (5, 11, 5 and 5%, respectively) and the pathogen was detected in only one component part of each sample, i.e. glume (82 and 93) or rachis (89 and 123). Samples 84, 85 and 91 had moderate disease scores (30, 26 and 25%, respectively) and PCR analysis detected the pathogen in the grain and glume of sample 91, in the grain and rachis of sample 84 and in the rachis of sample 85. F. culmorum was detected in samples 90, 96, 99, 119 and 122, which had severe FEB symptoms (disease scores of 95, 68, 100, 100 and 80%, respectively). PCR analysis detected the pathogen in the grain, glume and rachis components of samples 99 and 119, in the grain and rachis components of sample 90, the grain and glume components of sample 122 and only in the rachis component of sample 96. Sample 107 had exhibited no FEB symptoms, but PCR analysis detected F. culmorum in the grain of this sample. When the F. culmorum PCR results were correlated with visual disease assessment (Table 3), the highest correlation coefficient was obtained for the PCR results from glume samples (r = 0.822), as opposed to grain samples (r = 0.759), rachis samples (r = 0.782) or the combined grain, glume and rachis results (r = 0.675).

Table 3.  Correlation between visual disease assessment results and various PCR-based diagnostic assay results obtained for wheat ears and their component parts (grain, glume and rachis tissue) aBased on arcsine-transformed subplot frequency data.Thumbnail image of

M. nivale var. nivale and var. majus PCR analysis

The PCR results obtained for M. nivale var. nivale and var. majus were analysed in the same way as those for F. culmorum (Fig. 2b,c and d). When compared with the surrounding subplots, the frequency of M. nivale var. nivale and var. majus detection was relatively low in samples from the inoculated subplot (subplot 8), which had a high incidence of F. culmorum infection. This was particularly evident in rachis samples (Fig. 2d). M. nivale var. majus was detected in none of the glume components, and in only 8% of grain and 15% of rachis components of samples from this subplot. Similarly, M. nivale var. nivale was detected in none of the glume or rachis components and in only 8% of grain components of the samples. Subplot 9 also had a high incidence of F. culmorum infection but M. nivale var. nivale was not detected in any of the samples (Fig. 2b, c and d); var. majus was not detected in the glume components although it occurred in the grain, rachis or both components of 25% of the samples. PCR analysis revealed that there was a particularly high incidence of these pathogens in the 13 samples from subplot 7, 69% of samples being infected by M. nivale var. nivale and 92% of samples infected with var. majus. The pathogen was detected predominantly in the rachis components.

PCR analysis detected M. nivale var. nivale in 33, 25, 0, 0, 56, and 62% of the samples from subplots 1–6, respectively, while var. majus occurred in 66, 25, 50, 71, 33 and 77%, respectively. Again, both pathogens were found predominantly in the rachis components, often of the same sample, and as shown earlier, PCR analysis did not detect F. culmorum in any of these samples, many of which had low to moderate disease scores. M. nivale var. nivale was detected in 8, 11, 29, 0, 25 and 0% of samples from subplots 10–15, respectively, and var. majus in 23, 66, 86, 50, 50 and 66%. M. nivale var. nivale was found predominantly in the rachis component of these infected samples, often in conjunction with var. majus. M. nivale var. majus was often detected in the corresponding glume and rachis components of samples from subplots 10–15. Of the total of 21 M. nivale var. majus-infected samples from subplots 10–15, 63% were also infected by F. culmorum, and both pathogens were detected in the rachis component of 13 of the 14 co-infected samples. None of the five M. nivale var. nivale-infected samples from subplots 10–15 was infected by F. culmorum. Most M. nivale var. nivale and/or var. majus-infected plants from subplots 10–15 had been given low to moderate disease scores, although in a few cases no FEB symptoms had been seen (samples 97, 106 and 107 from subplots 11, 12 and 12, respectively). M. nivale was also detected in some samples from subplots 10–15 for which disease scores were relatively high, with var. majus in samples 96, 99, 103, 116, 120 (subplots 11, 11, 11, 13 and 14, respectively) and var. nivale in sample 120 (subplot 14). M. nivale var. majus and var. nivale PCR analysis for some of these samples is shown in Fig. 3(b). M. nivale var. majus was detected in the rachis of samples 96 and 99, which were also infected by F. culmorum (Fig. 3a), and which had disease scores of 68 and 100%, respectively. M. nivale var. majus was detected in the grain, glume and rachis component of sample 116 and 120, and in the rachis component of sample 103, while var. nivale was detected in the glume component of sample 120. F. culmorum was not detected in these samples, which had disease scores of 75, 84 and 55%, respectively. PCR analysis showed that var. majus was the predominant M. nivale subspecies within the field plot, with 64% of M. nivale-infected ears showing var. majus and 36% var. nivale. Since both M. nivale var. majus and var. nivale occurred on symptomatic and asymptomatic wheat ears, neither pathogen served to increase the correlation coefficient between visual disease assessment and PCR-based assays (Table 3). In tests of the association between pathogens (Everitt, 1986), the only significant association found was between F. culmorum and M. nivale (P < 0.01), there being an excess of samples where only one of the pathogens was present, i.e. F. culmorum or M. nivale var. majus/var. nivale.

F. poae and F. avenaceum PCR analysis

PCR analysis was also employed to determine if any of the samples were infected by F. poae (Fig. 2). In subplots 8 and 9, which had a high incidence of F. culmorum infection, the incidence of F. poae was highest in the glume component of samples, occurring in 23 and 15% of the glumes, respectively. F. poae infected many glume samples throughout the field plot; for example, it was detected in 54% of the glume samples from subplot 7, but not in any of the corresponding grain or rachis samples. The infected samples had disease scores of 0–33% and PCR analysis did not reveal F. culmorum in any of them. M. nivale var. nivale and var. majus were detected in many of the F. poae-infected samples from subplot 7, although these pathogens were predominantly found in the rachis rather than the glume components.

PCR analysis detected F. poae in 66, 100, 50, 58, 56 and 38% of samples from subplots 1–6, respectively. Incidence in grain and glume components was relatively high (Fig. 2), and the pathogen was detected in either or both of these components of a sample. It was generally not detected in the rachis components. For many of the F. poae-infected glume and grain samples, the corresponding rachis components were infected by M. nivale var. nivale and/or var. majus. In subplots 10–15, F. poae was detected in 23, 66, 71, 17, 0 and 33% of samples, respectively (Fig. 2), predominantly in the glume component and rarely in the rachis component. The majority of infected samples had low to moderate disease scores (5–35%), although a few showed no FEB symptoms (samples 97, 106 and 107). F. poae was also detected in samples that had relatively high disease scores (samples 96, 99, 103 and 116) and the PCR analysis of these is illustrated in Fig. 3(c). F. poae was found in conjunction with F. culmorum (Fig. 3a) and M. nivale (Fig. 3b) in samples 96 and 99 (disease scores 68 and 100%, respectively). F. poae, F. culmorum and M. nivale were detected in the rachis component of sample 96. Both F. poae and F. culmorum were detected in the grain, glume and rachis tissues of sample 99, while M. nivale was found in the rachis. F. poae occurred in conjunction with M. nivale in samples 103 and 116 (disease scores 55 and 75%, respectively), but no F. culmorum was found. In sample 103, F. poae was detected in the grain and glume components and M. nivale in the rachis. Both F. poae and M. nivale occurred in the grain, glume and rachis of sample 116. As for M. nivale var. majus and var. nivale, when the PCR results for F. poae were combined with those for F. culmorum and/or M. nivale and correlated with visual disease assessment, the various correlation coefficients were lower than that obtained for visual disease assessment and F. culmorum PCR (Table 3).

PCR revealed a low incidence of F. avenaceum within the field plot and it was detected in only 18 of the 123 samples (results not shown). These infected plants generally had low to moderate disease scores (5–33%), although FEB symptoms were not detected in two of the F. avenaceum-infected samples (6 and 43). F. avenaceum was also detected (along with M. nivale var. majus and F. poae) in the grain of sample 116, which had a disease score of 75% (Fig. 3). In contrast to the M. nivale var. majus, M. nivale var. nivale and F. poae PCR results, when the F. avenaceum PCR results were combined with those for F. culmorum and correlated with visual disease assessment (r = 0.843), a more linear relationship was observed than between visual disease assessment and F. culmorum PCR (r = 0.822) (Table 3).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The use of species-specific PCR analysis to identify the fungal species present in wheat ears that exhibited FEB symptoms within the field plot has highlighted many potential problems associated with visual assessment of this disease complex. Only within the central F. culmorum-inoculated subplot and an adjacent subplot (9) was a high correlation found between visual disease assessment and the presence of F. culmorum as detected by PCR analysis. Both visual disease assessment and PCR analysis revealed the presence of F. culmorum in all 13 ears sampled in the inoculated subplot and in all of the ears exhibiting symptoms in the adjacent subplot. However, PCR analysis failed to detect F. culmorum in some of the samples taken from the remaining subplots that were given disease scores assumed to be related to the presence of F. culmorum originating from the inoculated subplot. F. culmorum was detected in the most of the ears (both within and outside the inoculated subplot) exhibiting the highest disease scores (above 50%). However, since PCR analysis did not indicate that there was a gradient in the field plot in terms of the frequency of detection of the pathogen, it cannot be assumed that the F. culmorum inoculum for all of these samples originated from the inoculated subplot. It could be inferred that perhaps F. culmorum was more pathogenic than the other Fusarium species and Microdochium subspecies detected within the field plot, under the prevailing environmental conditions. In a recent review of FEB, Parry et al. (1995a) suggested that F. culmorum along with F. graminearum were consistently the most pathogenic of the Fusarium species infecting cereal ears.

PCR analysis indicated that the presence of other Fusarium species and M. nivale subspecies within the field plot may account for the disease observed in many areas of the field trial, particularly where low to moderate disease scores were recorded in subplots 1–6 and 10–15 of the field trial. It is likely that the high level of the pathogenic isolate of F. culmorum applied to the inoculated region limited the establishment of other species within this area of the field plot.

The separation of ear samples into their component parts (grain, glume and rachis) for the purpose of PCR analysis permitted tissue localization of the various Fusarium species and M. nivale subspecies. This highlighted the fact that F. poae predominated in glume material, but was often not detected in the corresponding grain tissue and was rarely found in rachis tissue, suggesting that initial infection by this pathogen was via the glumes. In a survey of Fusarium ear diseases of winter wheat in Great Britain during 1989–90, Polley & Turner (1995) reported that F. poae was the Fusarium species most commonly isolated from glumes affected by ear blight symptoms and that this pathogen was associated with distinct glume spot lesions. Thus it may be that the mode of infection and colonization by F. poae differs from that of other pathogens associated with Fusarium ear blight.

Outside the inoculated and adjacent subplots, F. poae, F. avenaceum and M. nivale were detected in ears that did not contain F. culmorum and that had moderate to high disease scores. Indeed, one interesting observation that emerged from this experiment was the occurrence of F. culmorum, F. avenaceum and M. nivale on the grain and rachis components, but not the corresponding glume components, of several high disease score samples taken from outside the inoculated subplot. Although these only accounted for a small proportion of samples, these observations implied that the bleaching of ears, which led to high visual disease scores for these samples, was not a result of glume infection, but rather was an indirect consequence of pathogen infection, possibly following the impediment of nutrient translocation to the glume.

PCR analysis also confirmed that M. nivale var. majus was more abundant in ears than var. nivale. Using a PCR-restriction fragment length polymorphism (RFLP) technique, Parry et al. (1995b) found that, of 91 Microdochium isolates obtained from grain taken from seven sites throughout the UK, 93% were var. majus and 7% var. nivale. In the present study it appeared that M. nivale var. majus was more frequently detected in rachis than in grain or glume tissue.

PCR analysis detected F. avenaceum in comparatively few samples and F. graminearum was not found within the field plot. Tests using F. avenaceum as inoculum have resulted in severe ear blight in the UK (Parry et al., 1995a), although this species generally represents a small proportion of the isolates obtained from FEB-affected crops in the UK and other cool maritime regions of north-west Europe. There are few reports of F. graminearum on wheat crops in the UK (Moore, 1948) and this species is generally important in hotter regions of the world (Parry et al., 1995a).

The fact that pathogens were detected in asymptomatic samples may be due to several factors including sensitivity of PCR-based assays and variability of disease symptoms (i.e. distinctiveness from F. culmorum disease symptoms). Among the pathogens, the only significant association was that found between F. culmorum and M. nivale var. majus/var. nivale. The significant probability of independence found between F. culmorum and M. nivale var. majus/var. nivale may suggest some sort of antagonistic interaction between these pathogens.

This work has demonstrated how PCR analysis may be used to gain insight into, and overcome, some of the problems associated with the diagnosis and understanding of FEB of wheat.

Footnotes
  1.  To whom correspondence should be addressed.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

FMD was supported by AgrEvo UK Ltd. The work of the Cereals Research Department is supported by the Ministry of Agriculture, Fisheries and Food.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Cassini R, 1981. Fusarium diseases of cereals in Western Europe. In: Nelson PE, Toussoun TA and Cook RJ, eds Fusarium: Diseases, Biology and Taxonomy. University Park, PA, USA: The Pennsylvania State University Press, 5863.
  • 2
    Everitt BS, 1986. The Analysis of Contingency Tables. London, UK: Chapman and Hall.
  • 3
    Höxter H, Miedaner TH, Sander E, Geiger HH, 1991. Quantitative assessment of Microdochium nivale in rye with ELISA. Zeitschrift für Pflanzenkrankheiten und Pflanzenschutz 98, 137.
  • 4
    Inglis DA, Cook RJ, 1981. Calonectria nivalis causes scab in the Pacific Northwest. Plant Disease 65, 9234.
  • 5
    Koopmann B, Karlovsky P, Wolf G, 1994. Differentiation between Fusarium culmorum and Fusarium graminearum by RFLP and with species-specific DNA probes. In: Schots A, Dewey FM, Oliver R, eds Modern Assays for Plant Pathogenic Fungi, Identification, Detection and Quantification. Wallingford, UK: CAB International, 3746.
  • 6
    Lees A, 1995. Diagnosis and Control of Foot Rot Pathogens of Wheat. Buckinghamshire, UK: Open University, PhD thesis.
  • 7
    Lees A, Nicholson P, Rezanoor HN, Parry DW, 1995. Analysis of variation within Microdochium nivale from wheat: evidence for a distinct sub-group. Mycological Research 99, 1039.
  • 8
    Moore WC, 1948. Report on Fungal, Bacterial and Other Diseases of Crops in England and Wales for the Years 1943–46. London, UK: Ministry of Agriculture and Fisheries Bulletin no. 139.
  • 9
    Nicholson P, Lees AK, Maurin N, Parry DW, Rezanoor HN, 1996. Development of a PCR assay to identify and quantify Microdochium nivale var. nivale and Microdochium nivale var. majus in wheat. Physiological and Molecular Plant Pathology 48, 25771.
  • 10
    Nicholson P, Parry DW, 1996. Development and use of a PCR assay to detect Rhizoctonia cerealis the cause of sharp eyespot in wheat. Plant Pathology 45, 87283.
  • 11
    Ouellet T, Siefert KA, 1993. Genetic characterisation of Fusarium strains using RAPD and PCR amplification. Phytopathology 83, 10037.
  • 12
    Parry DW, Jenkinson P, McLeod L, 1995a. Fusarium ear blight (scab) in small grain cereals − a review. Plant Pathology 44, 20738.
  • 13
    Parry DW, Nicholson P, 1996. Development of a PCR assay to detect Fusarium poae in wheat. Plant Pathology 45, 38391.
  • 14
    Parry DW, Rezanoor HN, Pettitt TR, Hare MC, Nicholson P, 1995b. Analysis of Microdochium nivale isolates from wheat in the UK during 1993. Annals of Applied Biology 126, 44955.
  • 15
    Polley RW, Turner JA, 1995. Surveys of stem base diseases and Fusarium ear diseases in winter wheat in England, Wales and Scotland, 1989–90. Annals of Applied Biology 126, 459.
  • 16
    Polley RW, Turner JA, Cockerell V, Robb J, Scudamore KA, Sanders MF, Magan N, 1991. Survey of Fusarium species infecting winter wheat in England, Wales and Scotland, 1989–90. In: Home Grown Cereals Authority Report 39. London, UK: Home Grown Cereals Authority Publication.
  • 17
    Rapilly F, Lemaire JM, Cassini R, 1973. Les Fusarioses. In: Les maladies des Céréales, Paris, France: INRA/ITCF, 11938.
  • 18
    Rom DR, 1990. A sequentially rejective test procedure based on a modified Bonferroni inequality. Biometrika 77, 6635.
  • 19
    Schilling AG, Möller EM, Geiger HG, 1996. Polymerase chain reaction-based assays for species-specific detection of Fusarium culmorum, F. graminearum and F. avenaceum. Phytopathology 86, 51522.
  • 20
    Wiese MV, 1987. Compendium of Wheat Diseases. St. Paul, MN, USA: American Phytopathological Society Publication.