| Abstract | [Abstract]: Crown rot (causal organism: Fusarium pseudograminearum) is a significant
disease affecting wheat in Australia. Although first reported over 60 years ago, the disease has become more prevalent in recent years due to the adoption of
minimum tillage and stubble retention practices. Breeding for resistance to crown rot is difficult - phenotypic selection, which is usually done at harvest, is
time-consuming, expensive, and subject to between year variability due to sensitivity to environmental conditions. For these reasons, the coupling of
molecular techniques with conventional plant breeding (marker-assisted selection) has the potential to more rapidly and reliably identify genomic regions
that contribute to resistance. The objective of this study was to identify, validate,
and pyramid quantitative trait loci (QTL) for resistance to crown rot present in a
W21MMT70 x Mendos doubled haploid wheat population. Replicated seedling trials were conducted in 2001, 2003, and 2005. In each
seedling trial, W21MMT70 displayed partial resistance to crown rot whereas
Mendos seedlings were susceptible. A bulked segregant analysis (BSA), using
390 simple sequence repeat (SSR) markers chosen for their coverage of the
wheat genome, was initially conducted based upon the 2001 seedling trial data in
an attempt to rapidly identify genomic regions associated to resistance. The BSA
did not reveal any markers associated with resistance to crown rot. As a result, a
full mapping study was conducted. One hundred and twenty eight (128) SSR
markers were mapped across the population to produce a framework map.
Previously screened AFLP markers were added to the map. Composite interval
mapping revealed eight QTL associated with resistance. Of these, three (located
on chromosomes 2B, 2D, and 5D) were consistently detected in each of the three
seedling trials. Two QTL (on chromosomes 1A and 3B) were detected in two of
the three trials. The 2D, 3B, and 5D QTL were inherited from W21MMT70,
whereas the 1A and 2B QTL were inherited from Mendos. Two software programs were used to identify epistatic interactions between
QTL. While the results of the two programs differed markedly, both programs
detected a highly significant interaction between the W21MMT70 inherited 5D
QTL and a locus on chromosome 2D inherited from Mendos. The overall effect
of the epistatic interactions was not as great as the additive effects of nonepistatic
QTL. Nonetheless, the presence of epistasis may indicate that,
particularly in the case of 5D, the effect of this QTL may be dependent on the
background into which it is introgressed. Validation of three W21MMT70-inherited QTL (on chromosomes 2D, 3B, and
5D) was conducted on three F2 populations with W21MMT70 as one of the
parents. While the 5D QTL was validated in two of the three crosses, neither the
2D nor the 3B QTL were detected in any of the F2 validation populations. It is
likely that the size of the F2 populations (the largest composed of 94 individuals),
in conjunction with the variability that is inherent when screening for resistance
to crown rot, precluded validation of these regions. Validation of the 2B
Mendos-inherited QTL was conducted on a Sunco x Batavia doubled haploid
population because Sunco possesses the same Triticum timopheevi 2B
introgression that is present in Mendos. This validated QTL (designated
Q.CR..usq-2B2) explained 11 % of the phenotypic variance in the Sunco x
Batavia population. To assess the effectiveness of pyramiding QTL for resistance to crown rot, a 2-49
x W21MMT70 population was examined. A number of lines of this population
performed significantly better than each of the parents in the replicated seedling
trial that was conducted. Four QTL, located on chromosomes 1A, 1D, 2D, and
3B, were detected. The 1A and 1D QTL were inherited from 2-49 whereas the
2D and 3B QTL were inherited from W21MMT70. The 1A QTL from 2-49 has
not been previously validated, and this QTL has been designated QCr.usq-1A1.
The 3B QTL (designated QCr.usq-3B1) had the highest effect (LRS 42.1;
explaining 21.0 % of the phenotypic variance) in the 2-49 x W21MMT70
population. The 2D QTL (QCr.usq-2D1) was shown to have a minor effect. The
5D QTL that was inherited from W21MMT70 in the W21MMT70 x Mendos
population was not detected in the 2-49 x W21MMT70 population. A number of
possible explanations for the inability to detect this QTL in the 2-49 x
W21MMT70 population are discussed. |
|---|