Components of resistance to diseases in winter oilseed rape cultivars: CORDISOR

Abstract

The project had two aims: (i) to test whether measuring components of resistance to stem canker (caused by Leptosphaeria maculans) and light leaf spot (caused by Pyrenopeziza brassicae), especially symptomless growth, will improve selection of resistant cultivars; and (ii) to produce new methods to rate cultivar resistance to stem canker and light leaf spot in winter oilseed rape cultivar selection and evaluation trials.

The quantity of L. maculans DNA in a crown (stem base) sample from an individual winter oilseed rape plot before harvest was very closely related to the mean severity of canker symptoms recorded on the same plants. However, plant to plant variation (sample error) in both symptom severity and amount of pathogen DNA reduced the closeness of the relationship between amount of DNA in one plant sample and canker severity in a separate plant sample from the same plot at the same time. Sample sizes were ≤10 plants per plot, so increasing sample size to 30 plants per plot would reduce this variation but would increase collection and processing time and costs of DNA extraction and quantitative PCR (qPCR) analysis. The relationship between amount of L. maculans DNA and pre-harvest canker severity in the same plot became less good as L. maculans DNA was quantified in crowns at progressively earlier time-points. Amounts of L. maculans DNA in crowns were small in the winter but increased greatly from flowering onwards when canker symptoms became visible. L. maculans DNA in petioles of leaves with visible phoma leaf spots sampled in the autumn or winter could not be used as a good predictor of final canker severity on cultivars although there were clear reductions in pathogen DNA caused by fungicide use. In some cases, amounts of L. maculans DNA were large for some cvs that subsequently had relatively low canker severity, suggesting that a potential mechanism of resistance may operate in the stem or at the boundary between petiole and stem. Cultivars were ranked for amount of L. maculans DNA in stems in winter. The more susceptible cvs like Bristol and Winner tended to have more L. maculans DNA in stems than the more resistant cvs like Hearty and ES Astrid.

Difficulties in identifying the most appropriate time to sample meant that the relationship between amount of L. maculans DNA in petioles and final canker severity at harvest was not a reliable method to predict cultivar maculans in petioles of senescing leaves than in those of the lowest green leaf.

Symptom development and production of L. maculans DNA were investigated in controlled conditions using an extended set of varieties (six European cvs used in the CORDISOR field experiments and 30-40 Chinese cvs thought to be more susceptible to L. maculans) that were inoculated at single positions on leaves using a small drop of ascospore suspension. Large differences between cultivars in visible symptoms and in amounts of L. maculans DNA were measured at different times and distances from the inoculation point and these are thought to be potentially good indicators of 'field' resistance to L. maculans. Large differences in severity of visible symptoms also occurred following inoculation of wounded cotyledons with conidia (asexually-produced spores) of a set of different isolates (races) of L. maculans. More L. maculans DNA was produced in leaf 2 than leaf 1, when both were inoculated at the same time (i.e. leaf 1 slightly older).

The project showed for the first time that Pyrenopeziza brassicae infection of the main shoot tip (meristem) of oilseed rape plants in winter was a common and widespread occurrence. Amounts of DNA of P. brassicae in leaves or main shoots (meristems) related well to subsequent severity of light leaf spot (LLS) symptoms at some dates (particularly when visible symptoms were severe in late winter/early spring, when visual assessments are normally made for Recommended List evaluation). It is probable that the early samples taken in December and January, particularly in Scotland, had not had sufficient thermal time to allow the pathogen to grow and spread on plants (increasing biomass) through secondary infection, which limited the ability to discriminate between resistant and susceptible cvs. Sampling in late winter/spring appears to provide the best material for qPCR studies to discriminate cultivars for resistance to P. brassicae. There was always much more P. brassicae DNA in samples of leaf blades (lamellae) of recently expanded leaves than in the main shoot tip (meristem). The relationship between the amount of P. brassicae DNA and LLS severity in a range of cvs was not dependent on the total amount of DNA at the particular site or time of sampling. Measurements of LLS severity by visual assessment of plots (1-9 score) often did not relate well to mean disease severity measured by close inspection of ten plants. Incubation of ten newly expanded leaves per plot in plastic bags in a cool room for a few days provided a good visualisation of LLS symptoms to discriminate differences in cultivar resistance to P. brassicae. There were significant differences in amounts of pathogen DNA for both L. maculans and P. brassicae between sites that had high or low severity of visible disease and between untreated and fungicide-treated plots.

In these experiments, yield loss due to canker occurred only if mean canker severity was > 2.5 (over half stem cankered) by mid-June proceeding to reduce yield by 15% per additional unit of canker severity (0-6 scale). The occurrence of stem splitting, caused by very warm spring weather during stem extension, did not increase canker severity. There were few differences in number and time of leaf production between cvs and little difference in leaf or petiole length between cvs but these factors did vary considerably between sites. Canker severity assessed as cross-sectional area of canker at the stem base (at maximum severity) was very closely related to canker severity assessed as percentage of stem girdled by canker. Discrimination between different cvs could be improved by recording the percentage of canker (either as area or girdling) as a continuous variable, rather than on the 0-6 scale currently used in HGCA RL assessments, although data recording would be more time-consuming.

With respect to the first aim of the project, it is clear that measurement of components of resistance can help in the breeding of more resistant cvs. The project showed new information on the development of stem canker and pathogen DNA in the stem in thermal time - pathogen growth was not linear but increased from the onset of flowering at different rates for different cvs.  This was particularly important as the project also demonstrated that yield losses due to stem canker were insignificant if canker severity could be managed to be less than 2.5 on the 0-6 scale. 

Regarding the second aim, to produce new methods to rate cultivar resistance to stem canker and light leaf spot in winter oilseed rape cultivar selection and evaluation trials, the project showed that exisiting, relatively 'low-tech' methods were best but that, particularly with inoculated controlled environment studies, there appears to be great potential in measuring pathogen growth from its DNA in order to screen genotypes for resistance to L. maculans at early growth stages.

The project also showed that infection of meristems by P. brassicae was widespread before Christmas, and as this is though to lead to symptoms of plant stunting, indicates that late-autumn foliar fungicide applications should be used against light leaf spot.  The amount of DNA of P. brassicae on leaves sampled from field plots in late-winter was a reasonable indicator of cultivar resistance to LLS but this has not been tested in controlled conditions.