Basic HTML Version
Molecular Plant Breeding 2010, Vol.1 No.5
http://mpb.sophiapublisher.com
Page 6 of 10
Barc70, wmc757 and wmc283 were detected to be
significantly associated with GPS in one location,
respectively. Gwm44 and bac70 were also detected to
be significantly associated with TKW in two locations.
In conclusion, most marker-trait associations can be
detected in two locations, but no one marker showed
significant association with traits in three locations, this
might be caused by missing data (only three traits were
investigated in all of the locations).
2 Discussion
2.1 Marker allele diversity
High levels of polymorphism were observed for the
markers, with a range of 2~16 alleles and average of
6.1 alleles per marker locus, which indicated that the
diversity of wheat accessions in this study was
relatively high. The above results also indicated that
EST-SSR markers in this study had lower levels of
polymorphism as compared to SSR markers. This
suggested that genomic-SSR marker is more powerful
than EST-SSR marker in evaluation of genetic
polymorphism, Similar results were obtained by Leigh
et al. (2003) in 56 old and new varieties of bread
wheat on the UK Recommended List.
2.2 Linkage disequilibrium
The presence of LD is a prerequisite for association
mapping. Some factors increase LD, such as mutation,
mating system, genetic isolation, selection and so on. In
the other hand, factors such as recurrent mutations, high
recombination and mutation rate decrease LD (Gupta et
al. 2005). Besides, the presence of rare alleles can also
affect LD to lead inaccurate estimates (
Abdurakhmonov
and Abdukarimov, 2008). Hence, in this study, markers
with rare alleles were replaced with missing values for
LD analysis and association mapping. The background
LD, which was defined as the distribution of LD
between unlinked loci, must be considered to evaluate
the extent of LD. Thus, in this study the 95th percentile
of the distribution of unlinked r
2
estimates was used as
baseline LD, instead of r
2
= 0.1 or r
2
= 0.2 used in
previous studies (Palaisa et al. 2003).
Our results showed that LD on chromosome 4A
extended up to ~3 cM with
r
2
=0.054 in 103 wheat
accessions surveyed with 31 SSR markers. While
previous studies indicated that LD was different in
different populations of wheat. Maccaferri et al. (2004)
reported that LD extended up to 10 and 20 cM with
D’= 0.67 in a collection with 134 durum wheat
accessions. Somers et al. (2007) reported that LD
extended approximately 2~3 cM with r
2
>0.2, but few
loci at longer distances showed high levels of LD (r
2
=
0.7 and 1.0) extending 25.5 and 41.2cM. Tommasini
et al. (2007) reported that LD on chromosome 3B
extended up to 0.5 cM in 44 wheat varieties or 30 cM
in 240 RIL populations of winter wheat based on 91
SSR and STS markers. Rhone et al. (2007) found a
strong LD extending up to 20 cM with r
2
>0.6. Crossa
et al. (2007) reported that the r
2
declined within ~40
cM below critical values, which occurred at a larger
distance than in previous studies in wheat. Probably
there are at least three likely causes for the difference
of LD extent. The first one is that higher diversity can
cause shorter LD extent, because of the many
recombination events that have accumulated. In
agreement with that, Tenaillon et al. (2001) reported
that the extent of LD in nine U.S. inbred lines was
higher than in 16 exotic landraces. The second one is
that different chromosome has different LD level. For
example,
Breseghello
and Sorrells (2006) reported
that LD decayed within a distance of < 1 cM (2D) and
< 5 cM (5A) with r
2
= 0.065 in 95 soft winter wheat
cultivars. The last one is that different marker type has
different LD estimates. Remington et al. (2001) found
that SSR loci showed proportionally higher estimates
of LD than SNPs.
It was also found that LD decay in wheat was far
higher than LD decay found in maize. Remington et al.
(2001) reported the rapid decay within 1500 bp to
values of
r
2
< 0.1 when assessed with SSRs. Jung et al.
(2006) reported the extent of LD within 500 kb in
surveying
adh1
locus.
Andersen
et al. (2007) reported
that LD persisted over the length of the phenylalanine
ammonia lyase (PAL) gene locus (3.7 kb) with the
r
2
>
0
.
2 in a survey of 32 European maize inbred lines.
The difference between wheat and maize in LD-decay
could be due to mating system. Maize, as a highly
outcrossing crop species, has much higher
recombination than hexaploid wheat, which is an
almost completely self-pollinating species, so rapid