Leaf rust resistance is desirable in most areas of the world. In Ontario, leaf rust may be severe, especially in cultivars which are resistant to powdery mildew. Many of the currently used sources of resistance to leaf rust are single gene and race specific, with the inherent risk posed by the possibility of race changes in the rust population. A source of an alternative, and potentially more durable, form of resistance is desirable. High levels of partial resistance, based on an accumulation of minor genes, may be the solution. The use of a partially rust resistant parent (TR306) in creating the two-rowed barley doubled haploid population, as part of the North American Barley Genome Mapping project, provided an opportunity to attempt to analyse a doubled haploid population of for partial resistance to leaf rust.
The Harrington/TR306 doubled haploid mapping population (150 HT lines and the two parents) were planted as hill-plots in four and two randomized replications in 1992 and 1993, respectively, at the University of Guelph's Elora research station. The leaf rust susceptible variety Leger was used as a running check, every tenth hill-plot, for disease severity and incidence. Leger was also used as a spreader around the plots with a rust susceptible winter barley bulk population as a spreader between the rows of hill-plots.
In 1992, the severity of leaf rust was scored on flag leaves due to prevalence of powdery mildew infecting the lower leaves in the hills. In 1993, the amount of leaf rust was scored on the penultimate leaf as mildew was not as severe and rust came in earlier. The severity of rust was scored on a scale 0-9. An average of 10-15 leaves were used to score each hill-plot. Moderately susceptible to susceptible infection types were observed.
The combined leaf rust severity from both years is presented in Table 1. Parental line TR306 had a score of 4.5 (moderately resistant) and the parental line Harrington had a score of 7.0 (susceptible). Susceptible check Leger had a score of 7.0.
The lowest score was 3.9 for the line HTOIO and the highest score was 8.1 for the line HT139. Transgressive segregation for reaction to leaf rust was observed (Table 1) with eight lines having a lower score than more resistant parent TR306 and seventeen lines having a higher score than a more susceptible parent Harrington. However, most of the lines (122) had a score ranging in between parental scores (4.5-7.0).
The distribution of disease severity observed may imply the presence of partial resistance to leaf rust in the parental lines. If that is so, TR306 would possess most of partial resistance alleles, while Harrington would have few or none of the desirable alleles for this trait. This hypothesis would explain a slight skewness of population from normality towards the more susceptible Harrington. The distribution of doubled haploids between two parental lines supports this conclusion.
Quantitative or partial resistance is generally environmentally affected and measured with numerical parameters. While trying to measure a level of quantitative resistance one should bear in mind that genetic variability among the parental lines determines the structure of the resulting population. In other words, normality of progeny distribution and presence of transgressive segregates depend directly on number of differing alleles and their expression. In terms of reaction to leaf rust, this cross may be designated as susceptible (Harrington) × moderately resistant (TR306). Random sampling of parental gametes, using the doubled haploid method, appears to be an ideal tool for sorting out the basis of resistance in these barley lines.
Field data per se should be backed-up with laboratory data. This appears to be important specifically in terms of revealing presence or absence of any race specific gene(s) and separating their effects from those of partial resistance genes. Very low severity scores may be related to accumulated partial resistance if pathogen infection pressure in the field is substantial. However, the presence of major, race specific resistance genes in the host and low levels of matching virulence in the pathogen could give similar results, and so their presence in the population must be unequivocally determined. From previous observations of these particular parents, we concluded that no effective, race-specific genes were present in this population; nothing in the currently reported results suggests otherwise.
These data will serve for analysis of QTLs for leaf rust resistance on the basis of RFLP markers developed for cross Harrington/TR306 by the North American Barley Genome Mapping Project which partially supported this research.