Development of an international standard set of barley differential genotypes for Pyrenophora teres f. teres

International comparison of virulence profiles of Pyrenophora teres f . teres (Ptt), the cause of barley net blotch, is seriously restricted by inconsistencies in differential testers used among researchers. This paper reports an attempt to develop an appropriate set of differentials to standardize characterization of Ptt populations globally. Fourteen barley genotypes (Canadian Lake Shore (CLS), Harbin, c-8755, c-20019, Manchurian, Tifang, CI 9825, CI 5791, CI 9819, Beecher, CI 9214, Skiff, Prior and Corvette) were selected from among genotypes previously used as Ptt differentials. Three cultivars (Pirkka, Haruna Nijo and Harrington) were included to identify a universally susceptible control. Genotypes were inoculated with approximately 1000 Ptt isolates from Russia, Europe, Australia and Canada. The mean reaction frequency of genotypes ranged from highly resistant (CI 9819, CI 5791, c-8755 and CI 9825) to highly susceptible (Harrington, Haruna Nijo and Pirkka). The best differential abilities were demonstrated by Harbin, CLS, c-20019, Manchurian and Prior. Application of cluster analyses identified genotypes with similar reaction patterns, which supported a reduction of genotypes in the set. When combined with an algorithm comparing the ability of individual genotypes to discriminate among Ptt isolates, a further reduction of genotypes was justified. A new, concise set of barley genotypes for differentiating virulences in Ptt was formulated. It is proposed that these genotypes be adopted as the standard, international differential set to characterize and identify the virulence properties of Ptt populations across environments. The new Ptt differential set consists of the genotypes c-8755, c-20019, CI 5791, CI 9825, CLS, Harbin, Prior, Skiff and Harrington.     

Pathogenic variation in populations of Drechslera teres f. teres and D. teres f. maculata and differences in host cultivar responses

The current study examined the variability in the pathogenicity of populations of Drechslera teres f. teres and D. teres f. maculata (the net and spot forms of D. teres) from Ireland and northern Europe. A population of progeny isolates from a mating of net and spot forms was also examined. Significant variation in virulence was found both between and among net form and spot form isolates (p<0.001). In the Irish population, significant differences were found between the net and spot forms, with the spot form isolates more virulent (p<0.05). Progeny isolates were significantly more virulent than net form or spot form populations (p<0.001). Significant differences were found in cultivar reactions, with cv. Botnia most susceptible to both forms of the pathogen (p<0.001). Cultivar Boreal 94145, although quantitatively resistant, was found to be very susceptible to both forms of the pathogen and to progeny isolates. Cultivars CI 5791, CI 2330 and CI 9819 were all less susceptible to infection by both forms, but were more susceptible to spot form isolates. Significant correlations were found between whole plants and detached leaf experiments for the net form isolates only (p<0.001). This study illustrates the importance of including both net form and spot form isolates in resistance studies and the need for a clearer understanding for the genetic basis of resistance to the net and spot forms. It also highlights the limitations of using a detached leaf assay for screening of net blotch of barley.     

Genetic variation of Pyrenophora teres f. teres isolates in Western Australia and emergence of a Cyp51A fungicide resistance mutation

Genome-wide, unlinked, simple sequence repeat markers were used to examine genetic variation and relationships within Pyrenophora teres f. teres, a common pathogen of barley, in Western Australia. Despite the region's geographic isolation, the isolates showed relatively high allelic variation compared to similar studies, averaging 7.11 alleles per locus. Principal component, Bayesian clustering and distance differentiation parameters provided evidence for both regional genotypic subdivision together with juxtaposing of isolates possessing different genetic backgrounds. Genotyping of fungicide resistant Cyp51A isolates indicated a single mutation event occurred followed by recombination and long-distance regional dispersal over hundreds of kilometres. Selection of recently emergent favourable alleles such as the Cyp51A mutation and a cultivar virulence may provide an explanation, at least in part, for juxtaposed genotypes. Factors affecting genotypic composition and the movement of new genotypes are discussed in the context of grower practices and pathogen epidemiology, together with the implications for resistance breeding.     

Population structure of South African and Australian Pyrenophora teres isolates

There are two recognized forms of the disease net blotch of barley: the net form caused by Pyrenophora teres f. teres (PTT) and the spot form caused by P. teres f. maculata (PTM). In this study, amplified fragment length polymorphism analysis was used to investigate the genetic diversity and population structure of 60 PTT and 64 PTM isolates collected across Australia (66 isolates) and in the south‐western Cape of South Africa (58 isolates). For comparison, P. tritici‐repentis, Exserohilum rostratum and Bipolaris sorokiniana samples were also included in the analyses. Both distance‐ and model‐based cluster analyses separated the PTT and PTM isolates into two strongly divergent genetic groups. Significant variation was observed both among the South African and Australian populations of PTT and PTM and among sampling locations for the PTT samples. Results suggest that sexual reproduction between the two forms is unlikely and that reproduction within the PTT and PTM groups occurs mainly asexually.     

Genetics of host-pathogen interactions in the <Emphasis Type="Italic">Pyrenophora teres</Emphasis> f. <Emphasis Type="Italic">teres</Emphasis> (net form) – barley (<Emphasis Type="Italic">Hordeum vulgare)</Emphasis> pathosystem

The genetics of host-pathogen interactions in the Hordeum vulgare – P. teres f. teres pathosystem was studied in twelve resistant barley accessions, i.e. CI 9825, CI 9819, Diamond, CI 4922, CI 5401, Harbin, c-8755, c-21849, c-8721 c-23874, c-19979, c-15811. F₂ analyses of crosses with susceptible genotypes employing various isolates (from Europe, USA, Canada, and Australia) revealed that resistance is mostly isolate-specific and controlled by one or two genes. Segregation in ascospore progeny from two crosses between isolates of different origin revealed that avirulence in P. teres is also determined by one or two genes. An epistatic effect of suppressor genes on avirulence genes is proposed for the genetics of virulence to Diamond, Harbin, CI 5401 and c-8721 in the fungal crosses D (181-6 × A80) and F (H-22 × 92-178/9). Segregation in F₂ of crosses of three new sources of resistance (c-23874, c-19979, c-15811) to the susceptible cv. Pirkka was studied in laboratory and greenhouse tests by using seven P. teres isolates, i.e. 181-6, d8-3, d8-4, d9-1, d9-4, F4 and F74. In addition, virulence to these barley accessions of ascospore progeny from crosses of the same isolates was studied. Based on these studies it was concluded that depending on the isolate used, resistance of c-23874 is determined at least by two genes and in c-19979 and c-15811 by three genes. The results of this parallel analyses of genetics of resistance and genetics of virulence allows the postulation of a gene–for–gene interaction in the P. teres – H. vulgare pathosystem.     

An integrated multivariate approach to net blotch of barley: Virulence quantification,pathotyping and a breeding strategy for disease resistance

Knowledge of pathotype diversity and virulence in local populations of Pyrenophora teres is a prerequisite to screening for durable resistance to net blotch. The current study aimed to quantify the virulence level of Moroccan isolates, identify and designate existing pathotypes, and select resistant genotypes. We developed a method for virulence quantification of P. teres isolates based on a conversion of infection responses into frequencies for use in correspondence analysis. Coordinates of the first axis of this analysis had a virulence spectrum and ranked isolates from virulent to avirulent. Mixed model analysis was also devised for virulence quantification. Coordinates of the first dimension of correspondence analysis were linearly correlated to BLUPs (Best Linear Unbiased Predictors) of the mixed model. A genotype by genotype by environment model (GGE) coupled with cluster analysis differentiated P. teres isolates into ten and nine pathotypes for net- and spot-forms respectively. Populations of these two forms were dissimilar in terms of classes of virulence. For P. teres f. maculata, avirulent, moderately virulent and highly virulent isolates represented one-third of the population, whereas 90% of P. teres f. teres population was composed of avirulent to moderately avirulent isolates. Barley differential sets were subsequently reduced to two new sets that simplified pathotyping through a key code based on resistant or susceptible reactions. Dendrograms of cluster analysis based on GGE analysis depicted the stability of a genotype’s reactions across all isolates, and using only resistant cultivars as sources of resistance to control net blotch disease would, based on this analysis, fail to control all pathotypes. Therefore, we propose an alternative breeding strategy to control net blotch effectively.     

Control of Stagonospora nodorum and Septoria tritici in Wheat by Pre-treatment with Drechslera teres, a Non-host Pathogen

A field study is described which explored the possibility of controlling Stagonospora nodorum and Septoria tritici on wheat using a barley pathogen, Drechslera teres. Pre-treatment of wheat cv. Hussar flag leaves with D. teres resulted in a significant reduction in disease caused by S. nodorum and S. tritici, resulting in a significant increase in grain yield. When cv. Brigadier leaves were treated with D. teres prior to inoculation with S. nodorum there was an initial increase in disease expression whilst D. teres had no effect on symptoms produced by S. tritici on cv. Brigadier. There was significantly less disease on leaves of cvs. Hussar and Brigadier pre-treated with D. teres prior to inoculation with an equal mixture of S. nodorum and S. tritici compared to plants pre-treated with water. It is concluded that D. teres and other non-host pathogens show potential as biological control agents for S. nodorum and S. tritici.     

The genetic diversity of Icelandic populations of two barley leaf pathogens, Rhynchosporium commune and Pyrenophora teres

Twelve species of fungi have been found growing on barley leaves in Icelandic fields. The study presented here on the population structure of two of these species, the pathogens Rhynchosporium commune and Pyrenophora teres f. teres, reveals high levels of genetic diversity, low levels of migration, and a significant differentiation from other European populations, despite the short history of continuous barley cultivation in Iceland. The gene diversity for R. commune in Iceland was 0.55 compared to a range of 0.43?C0.73 for six European populations. The gene diversity for P. teres was higher in Iceland than in populations from Russia and Finland. The two mating types were found to overlap in distribution for both fungi making sexual reproduction a possibility, supported by the few clones and the gametic equilibrium within the Icelandic populations. When the high levels of diversity, low levels of migration, and the genetic differentiation observed between Icelandic and Scandinavian populations are put into context with the short history of barley cultivation in Iceland, it raises questions regarding the origin of the Icelandic fungal populations. It also underlines the importance of proper analysis of pathogens prior to starting resistance breeding projects. The findings are an addition to the ongoing analysis of the global diversity of barley fungal pathogens in general and an important input into future barley breeding projects in Iceland in particular.     

Tracking Phytophthora infestans with SSR markers within and between seasons – a field study in Sweden

The dynamics of a late blight epidemic and sexual reproduction in Phytophthora infestans were studied in an experimental field in mid‐Sweden. The field was inoculated with six isolates of P. infestans taken from another potato field where sexual reproduction of the pathogen was suspected. Three weeks after inoculation single‐lesion leaflets were sampled and the resulting isolates characterized using microsatellites (SSRs) and mating type as markers. Among the 151 isolates analysed, the inoculum genotypes constituted more than 80% of the genotypes found, with three other genotypes making up the remainder. The following year, P. infestans obtained from soil samples taken from this field were analysed, and six novel genotypes were identified. Genotypes from the previous summer’s population were not detected. Analysis of the genotypes recovered was consistent with them being recombinants, with the previous summer’s population acting as parents. These findings are consistent with the hypothesis that oospores produced during a summer epidemic in Sweden can overwinter and cause infection the next year.     

Genetic variation of single nucleotide polymorphisms identified at the mating type locus correlates with form-specific disease phenotype in the barley net blotch fungus Pyrenophora teres

Mating-type (MAT) locus-specific single nucleotide polymorphisms (SNPs) have been shown to be sufficient for conventional PCR-based differentiation of Pyrenophora teres f. teres (Ptt) and P. teres f. maculata (Ptm), the cause of the net and spot form, respectively, of barley net blotch. Here, we report the cloning and characterization of the MAT locus from 10 California isolates that cause atypical blotch symptoms on barley. Analysis of the full-length nucleotide sequences of one MAT1-1 (1,993?bp) and nine MAT1-2 (2,149 or 2,161?bp) idiomorphs revealed high (98?C99?%) similarity to those of Ptt isolates. However, distinct SNP patterns were identified in the newly cloned MAT idiomorphs. Two new MAT1-2-specific SNPs were found to be conserved in one Australia and eight California isolates that all cause similar atypical blotch symptoms. Phylogenetic analysis indicated that all 10 California isolates form a separate branch (or clade) within the Ptt group, except for one that appears to be ancestral to both Ptt and Ptm. PCR primers designed based on the identified SNP patterns were used successfully to differentiate each atypical isolate from highly virulent forms. This study extends our previous work and, taken together, the results demonstrate that the genetic variation at the MAT locus correlates with variation in the form-specific disease phenotype, and that MAT-specific SNPs can serve as reliable and convenient markers for subspecies level differentiation in P. teres, an economically important plant-pathogenic ascomycete (http://en.wikipedia.org/wiki/Single_nucleotide_polymorphism).     

Differential Selection on Rhynchosporium secalis During Parasitic and Saprophytic Phases in the Barley Scald Disease Cycle

ABSTRACT Competition among eight Rhynchosporium secalis isolates was assessed during parasitic and saprophytic phases of the disease cycle in field experiments conducted at two locations and over two growing seasons. The eight isolates were inoculated onto six barley populations exhibiting varying degrees of resistance. Microsatellite analysis of 2,866 isolates recovered from the field experiments showed significant, and sometimes opposite, changes in the frequencies of R. secalis genotypes during the growing season (parasitic phase) and between growing seasons (saprophytic phase). Isolates that showed the most complex virulence in greenhouse seedling assays had the lowest fitness in the field experiment. Significant differences in isolate fitness were found on different host populations and in different environments. Selection coefficients were large, indicating that evolution can occur rapidly in field populations. Although inoculated isolates had the lowest overall fitness on the moderately resistant landrace cv. Arabi Aswad, some isolates were more virulent and consistently increased in frequency on this landrace, suggesting a risk of directional selection and possible erosion of the resistance following its widespread deployment in monoculture. These results provide the first direct evidence that R. secalis pathogen genotypes differ in their saprophytic ability and parasitic fitness under field conditions.     

Optimization of in vitro growth conditions of Pyrenophora teres for production of the phytotoxin aspergillomarasmine A

In liquid cultures of Pyrenophora teres, three phytotoxins may be found: L, L - N -(2-amino-2-carboxyethyl) aspartic acid (toxin A), anhydroaspergillomarasmine A (toxin B) and aspergillomarasmine A (toxin C). In particular, toxins A and C cause chlorotic and necrotic symptoms in detached barley leaves, toxin C being the most damaging, whereas toxin B is only weakly phytotoxic. When P. teres is grown in liquid modified Fries medium, toxin B is the main toxin accumulated, possibly due to a ring closure of toxin C at the low pH value of the medium. The amount of toxin B produced by 11 isolates of P. teres was compared in modified Fries medium. Generally, the most virulent isolates of P. teres produced higher amounts of toxin B than the less virulent isolates. During growth, the pH of the media decreased from 6.7 to about 3.0–3.5, followed by a slight increase to about 3.5–4.0. All isolates, except one, produced toxin B, whereas only two isolates produced toxin C and toxin A. Maintaining the pH at about 6.5 by sterile titration with 1 M NaOH resulted in a shift in toxin accumulation from toxin B to toxin C. The addition of tris or phosphate buffer to the media resulted in higher pH during the growth period, an increase in the total amount of toxins produced, and a shift in toxin accumulation from toxin B to toxin C. The higher pH value probably prevented the conversion of toxin C into toxin B. No toxins were produced in two routinely used media, potato glucose broth and grass broth. Toxin B and toxin C were purified by ion exchange chromatography and precipitation with HCl.     

Genetic correlations in resistance to morpholine and piperidine fungicides inPyrenophora teres populations

Highly significant genetic variation (P<0.001) in resistance to the morpholine fungicides fenpropimorph, tridemorph and dodemorph and the piperidine fungicide, fenpropidin was found in different populations ofPyrenophore teres in North America and Europe which had not been previously exposed to these fungicides. Resistance phenotypes were continuously distributed for each fungicide in each population. Cross resistance relationships were determined by estimating genetic correlation coefficients in resistance to all pairwise combinations of fungicides. The majority of the correlation coefficients were highly positive for all fungicide combinations in all populations; eight of 36 (22%) coefficients were not significantly different from 1 (P>0.05). This result is consistent with the hypothesis that many of the same genes, or genes in gametic disequilibrium, control resistance to more than one fungicide in most populations ofP. teres and that these fungicides comprise a single cross resistance group. Three of 36 (8%) correlation coefficients were not significantly different from 0 (P>0.05) indicating that, in these populations, independent genes controlled resistance to these fungicides. The results of this study indicate that although most of the same genes control resistance to morpholine and piperidine fungicides inP. teres, differences in frequencies of these genes among populations can result in different cross resistance relationships from one population to another.     

Effect of Sexual Recombination on Pathotype Frequencies in Barley Powdery Mildew Populations of Artificially Inoculated Field Plots

Changes in pathotype frequencies due to sexual reproduction during summer were assessed in barley plots inoculated at the start of the growing season with powdery mildew isolates of the pathotype GL1 (Va22Vh) which belonged to a clonal lineage frequent in Northern France in 1992–1996 and was absent or rare in the local population at the time of the study. The capacity for recombination among isolates belonging to the pathotype GL1 was confirmed by performing crosses in the greenhouse. The field experiment was repeated over two years, with two field plots inoculated with either a single isolate (one mating type) or four isolates (two mating types). Following artificial inoculation, the frequency of GL1 increased to between 40% and 80% of the total conidia population at the end of the asexual epidemics before summer. When only one isolate (one mating type) of the pathotype GL1 was present, the frequency of the GL1 pathotype decreased to 4–5% of the ascospore population, following sexual recombination between the inoculated isolate and the local natural population. When isolates of both mating types of GL1 were inoculated, however, the GL1 pathotype remained dominant (28%) in the ascospore population. A pathotype dominant at the end of the summer might possibly have over-summered through asexual reproduction, or alternatively it may originate from sexual reproduction. The observed GL1 frequencies and numbers of virulences per isolate (complexity) observed on the field plots were compared to those calculated with a model assuming random or non-random mating.     

Response of strawberry genotypes to inoculation with isolates of Verticillium dahliae differing in host origin

Eight strawberry genotypes known to differ in susceptibility to verticillium wilt were inoculated with eight isolates of Verticillium dahliae originally obtained from six different host crops: strawberry, potato, watermelon, mint, eggplant (aubergine) and cauliflower. Inoculation experiments were conducted in replicated field trials during two successive years. Known susceptible genotypes developed typical symptoms of verticillium wilt in both years. Although isolates manifested a wide range of aggressiveness, differences were significant only on the most susceptible strawberry genotype. Two isolates originating from strawberry were among the most aggressive of the eight tested, whereas the least aggressive isolate was obtained from cauliflower. Six strawberry genotypes that were regarded as resistant to verticillium wilt based on previous tests were also resistant in the present study, regardless of the isolate used. Overall, strawberry genotypes represented the largest source of variation in these experiments, with variance components approximately 10-fold greater than those associated with either isolate or the isolate × genotype interaction. The results suggest it should be possible to develop resistance to verticillium wilt in strawberry that is broadly effective against isolates of diverse host origin.     

Phenotypic and molecular genetic characterization indicate no major race‐specific interactions between Xanthomonas translucens pv. graminis and Lolium multiflorum

Bacterial wilt of forage grasses, caused by the pathogen Xanthomonas translucens pv. graminis (Xtg), is a major disease of forage grasses such as Italian ryegrass (Lolium multiflorum). The plant genotype‐bacterial isolate interaction was analysed to elucidate the existence of race‐specific responses and to assist the identification of plant disease resistance genes. In a greenhouse experiment, 62 selected plant genotypes were artificially inoculated with six different bacterial isolates. Significant differences in resistance were observed among L. multiflorum genotypes (P < 0·001) and in virulence (intensity of disease symptoms) among Xtg isolates (P < 0·001) using the area under the disease progress curve (AUDPC). No significant genotype‐isolate interaction (P > 0·05) could be observed using linear regression modelling. However, additive main effects and multiplicative interaction effects (ammi ) analysis revealed five genotypes which did not cluster close to the origin of the biplot, indicating specific interactions between these genotypes and some bacterial isolates. Simple sequence repeat (SSR) markers were used to identify marker‐resistance associations using the same plant genotypes and bacterial isolates. The SSR marker NFA027 located on linkage group (LG) 5 was significantly associated with bacterial wilt resistance across all six bacterial isolates and explained up to 37·4% of the total variance of AUDPC values. Neither the inoculation experiment nor the SSR analyses revealed major host genotype‐pathogen isolate interactions, thus suggesting that Xtg resistance, observed so far, is effective across a broad range of different bacterial isolates and plant genotypes.     

Virulence profile and genetic structure of a North Dakota population of Pyrenophora teres f. teres, the causal agent of net form net blotch of barley

A Pyrenophora teres f. teres population in North Dakota was analyzed for virulence variation and genetic diversity using 75 monospore isolates that were collected across a 4-year period (2004 to 2007) from two North Dakota State University agricultural experiment stations at Fargo and Langdon. Pathogenicity tests by inoculation onto 22 barley differential lines at seedling stage revealed 49 pathotypes, indicating a wide range of pathogenic diversity. Two-way analysis of variance of disease ratings revealed a significant difference in the virulence among isolates and in the resistance among barley lines, as well as in the interactions between the two. 'CI5791', 'Algerian', and 'Heartland' were three barley lines showing a high level of seedling resistance to all North Dakota isolates tested; however, many previously reported resistance genes have been overcome. Forty multilocus genotypes were identified from this set of isolates by genotyping at 13 simple-sequence repeat loci. High percentages of clonal cultures were detected in the samplings from 2005 and 2007 in Fargo and 2005 in Langdon. Using a clone-corrected sample set, the mean gene diversity (h) was estimated to be 0.58, approximately the same for both locations. The calculated Wright's F(ST) value is small (0.11) but was significantly >0, indicating a significant differentiation between the Fargo and Langdon populations. In the gametic disequilibrium test, only 3 of 78 possible pairwise comparisons over all isolates showed significant (P < 0.05) nonrandom association, suggesting a random mating mode. Our results suggest that the populations from the two locations are derived from a common source and undergo frequent recombination. This research provides important information for barley breeders regarding development and deployment of cultivars with resistance to net form net blotch in this region.     

Head infection and accumulation of Fusarium toxins in kernels of 12 barley genotypes inoculated with Fusarium graminearum isolates of two chemotypes

Twelve barley genotypes were inoculated with two F. graminearum isolates of different chemotype I₁ #148 (producing nivalenol/deoxynivalenol) and I₂ #108 (deoxynivalenol/acetyldeoxynivalenol). For both I₁ and I₂ isolates, respectively, reductions (%) in number of kernels head^-1 10.6 and 14.3; yield 39.6 and 35.7; weight of 1000 kernels 36.9 and 23.2 were observed in inoculated plants from control values. Chemical analysis revealed the presence (average concentration mg kg^-1) of deoxynivalenol (1.3) and nivalenol (3.2) in kernels of all genotypes inoculated with the I₁ isolate, and zearalenone (0.2) in three samples. After inoculation with the I₂ isolate, deoxynivalenol (37.8) and zearalenone (0.4) were found in kernels of all genotypes, while 3-acetyldeoxynivalenol and 15-acetyldeoxynivalenol, respectively, were determined in five and four samples only. No significant correlation between examined characteristics was found for either the I₁ or I₂ isolate. The results obtained contribute information on the accumulation of toxins in cereal grain inoculated with F. graminearum isolates of different chemotypes.     

Effect of early infection on pathotype frequencies in barley powdery mildew (Blumeria graminis f.sp. hordei) populations in field plots

A field experiment with barley powdery mildew ( Blumeria graminis f.sp. hordei ) was designed in order to study how the time of arrival of inoculum in the field influenced pathotype frequencies in the resulting populations. Three isolates belonging to pathotypes that were absent or rare in the local aerial inoculum were used to inoculate field plots of winter barley cv. Plaisant. Two successive inoculations with different combinations of the three isolates were performed with an approximately two-generation delay, and frequencies of inoculated pathotypes were assessed four and nine generations after the first inoculation. Pathotypes of the first inoculated isolates generally persisted throughout the period of sampling; this is described as an 'early arrival' effect. During the epidemics the inoculated isolates were not replaced by isolates from the natural airborne inoculum. Pathotype frequencies depended mainly on the time of arrival of inoculum in the plot, but frequencies also depended on the isolate that had been inoculated. The most frequent isolate, GL1, belonged to the clonal lineage dominant in powdery mildew populations on winter barley in the north of France. These results confirmed that the composition of a powdery mildew population in a field is largely determined by the composition of the initial inoculum.     

Head blight and biosynthesis of Fusarium toxins in barley kernels field inoculated withFusarium culmorum

Heads of 12 spring barley genotypes (eight cultivars, four lines) were inoculated with conidial suspensions of twoFusarium culmorum isolates (I₁ and I₂). Both isolates caused the following significant reductions (%) when compared with the control: number of kernels head^–1 12 and 16, weight of 1000 kernels 47 and 24, yield 49 and 37, for I₁ and I₂ respectively. Both isolates were able to produce average levels (mg kg^–1) of deoxynivalenol of 67.1 and 13.9, 3-acetyldeoxynivalenol of 9.4 and 1.4 and zearalenone of 0.3 and 0.4 for I₁ and I₂ respectively. In kernels of three genotypes inoculated with the same isolates, 15-acetyldeoxynivalenol at an average level (mg kg^–1) of 1.3 and 1.5 was detected, while in kernels of six genotypes inoculated with I₂, 0.2 of nivalenol was found. A significant correlation for yield and toxin level was found for less pathogenicF. culmorum isolate I₂.