Agriculture Reference
In-Depth Information
interaction due to poor performance in the more-
droughted (ET2) environments (Chapman 2008).
This example illustrates the need for breeders
to manage (e.g., CIMMYT simulated drought;
Trethowan et al., 2005) or at least interpret trials
in such a way as to sample the different types of
drought via timing and location of trials. However,
frequently the interannual variation is such that
breeding evaluation trials should be characterized
individually if drought stress pattern is to be used
in the characterization of genotype performance
(Chapman 2008).
Yield potential and genetic gain in
water-limited environments
Heritability and genetic gain are typically higher
for wheat grown in favorable environments (Ud-
Din et al., 1992; Cooper et al., 1997). Access to
favorable testing environments may therefore
provide an opportunity to select among lines for
high yield potential. However, there are few
instances where selection in well-watered envi-
ronments has translated to broad adaptation in
very low-yielding, water-limited environments.
This partly refl ects large genotype × environment
interaction and a subsequently low genetic corre-
lation for grain yield across contrasting stress
levels (Ud-Din et al., 1992; Cooper et al., 1997).
Genetic variance and heritability are also some-
times lower in water-limited environments
(Ud-Din et al., 1992).
Using a theoretical framework, Rosielle and
Hamblin (1981) showed tolerance to stress and
mean productivity were negatively correlated
when the genetic variance under stress was
reduced compared to nonstress conditions. In
turn, selection for tolerance to stress commonly
results in reduced yield in nonstress environ-
ments and a decrease in mean productivity.
However, selection for improved mean produc-
tivity generally increases yields in both stress and
nonstress environments. Simmonds (1991) indi-
cated that indirect selection for low-yielding envi-
ronments via productivity in intermediate- to
high-yielding environments was ineffective. This
was particularly true when genotype × environ-
ment interactions refl ect crossover changes in line
ranking common to performance under lower-
yielding, water-limited conditions (Ceccarelli
1994). This assertion has been confi rmed empiri-
cally by Ud-Din et al. (1992) and Cooper et al.
(1997). Both showed that selection for improved
productivity in stress and other low-yielding con-
ditions must be undertaken in appropriate target
environments.
Selection for improved yield under drought
has produced yield gains of 4.4% yr −1 in the
CIMMYT semiarid wheat breeding program
(Trethowan et al., 2002). Selection of semiarid,
adapted lines with improved productivity under
BREEDING FOR IMPROVED
PERFORMANCE UNDER DROUGHT
The identifi cation and release of cultivars with
the capacity to perform well across a range of
environments represents a major challenge to
wheat breeders. Genetic gain for grain yield has
been greatest in managed, high-yielding environ-
ments, particularly with the development of
improved disease resistance and introduction of
dwarfi ng genes to increase harvest index and
reduce lodging (Richards et al., 2002). In con-
trast, genetic progress has been smallest in water-
limited environments (Araus et al., 2002) with
long-term yield gains of less than 0.5% year −1
reported for many water-limited environments
(Byerlee and Morris 1993).
Development of broadly adapted cultivars is
diffi cult when the cropping region encounters
both well-watered and water-limited environ-
ments (Calhoun et al., 1994). To date, essentially
two strategies have guided selection for yield
under drought. The fi rst refl ects selection for
improved performance under favorable condi-
tions and an expectation that line ranking will be
substantially maintained under less favorable
conditions. The second strategy focuses on selec-
tion for improved productivity in targeted, water-
limited environments. In this case, selection may
take the form of empirical selection for yield per
se , or take an analytical selection approach target-
ing physiological and/or developmental traits
underpinning the known biology of adaptation to
drought.
Search WWH ::




Custom Search