Agriculture Reference
In-Depth Information
some of the reasons being considered earlier and
in Richards (2006). However, of equal importance
is the need for breeding teams that integrate
skills from physiology to genetics and breeding.
Such complementation currently exists for pathol-
ogists and breeders but should also be extended to
physiologists if meaningful selection for physio-
logical traits is to be implemented, as in the case of
the commercial release of the high transpiration
effi ciency wheat cultivar Drysdale (Fig. 11.8).
Finally, the value of accurate phenotyping is not
restricted to breeding programs and is considered
a critical requisite to good QTL analysis.
molecular markers provides greater opportunity
for trait dissection into component QTLs, and
insight into the underlying genetic basis for varia-
tion and covariation in a trait. For example,
numbers of genes and their interactions, and gene
action, can be determined even for low-heritability
traits. Information around breeding complexity
including linkage, linkage disequilibrium, and
pleiotropy may also be gleaned through mapping
studies, particularly if extended to multiple popu-
lations (Rebetzke et al., 2007a). Key processes
involved in plant growth and development can
then be inferred from what is already known of
the physiology.
For example, development of sequence-based
perfect markers for the
Rht-B1b
and
Rht-D1b
dwarfi ng genes has confi rmed their effects on
tissue insensitivity to endogenous gibberellins
(Ellis et al., 2002), and subsequent reductions in
both coleoptile and early leaf area development
(Rebetzke et al., 2001a). Currently, PCR-based
markers are being developed for gibberellin-
sensitive dwarfi ng genes
Rht4
,
Rht5
,
Rht12
,
and
Rht13
for breeding of long coleoptile, high
early-vigor wheat with short stature (Ellis et al.,
2004).
Perhaps the greatest opportunity for the use of
linked markers in breeding for improved perfor-
mance under drought will apply to selection for
traits with low heritability, recessive gene expres-
sion, or high cost of measurement, such as root
growth (Tuberosa et al., 2002) or WSC concen-
tration (Rebetzke et al., 2008b). In the absence of
recombination, markers linked to a target QTL
have a heritability of 100%, thereby allowing
rapid generation advance. From the indirect
selection formula (equation 11.1), the benefi t of
using a 100%-heritable molecular marker over
direct selection for grain yield (i.e., Δ
G
Y.X
/Δ
G
Y
) is
r
A
/h
yield
. This can be extended to multiple QTLs
so that the sum of absolute genetic correlations
with yield must exceed
h
yield
. Assuming the marker
is diagnostic (robust marker and gene associa-
tion), and evaluation and generation costs are
similar to that of yield, marker-aided selection in
variable environments should produce greater
genetic gain for yield. However, this is likely to
Quantitative trait loci
Little is known or understood of the genetic basis
of wheat performance under drought. Substantial
progress has been made on the development and
availability of molecular markers and their subse-
quent integration into genetic maps across a
number of genotyped populations (Gupta et al.,
2008). The use of molecular techniques is com-
plementary to the use of physiological approaches
in breeding wheat with improved performance
under drought. Further, a range of different
80
60
40
20
0
-20
0
1,000 2,000 3,000 4,000 5,000 6,000 7,000
Environment mean yield (kg ha
-1
)
Fig. 11.8
Yield advantage of low carbon isotope discrimina-
tion-selected cultivar Drysdale grown side-by-side with its
recurrent parent 'Hartog' at 31 sites in 2004 (
), 29 sites in
2005 (
•
), and 12 CSIRO sites (
) from 1995 to 1998. Mean
grain yield advantage was 16%, 9%, and 17% for Australian
Grains Technology Breeding Company (AGT) 2004 and
2005, and CSIRO trials, respectively. (AGT data provided
courtesy of Steven Jeffries.)
