Agriculture Reference
In-Depth Information
is unlikely to be broadly applicable to increasing
crop productivity under drought.
et al., 1980). Thus future genetic gain in yield
through empirical breeding is likely to arise
through subtle changes in harvest index, coupled
to an increase in total biomass.
There is some evidence that genotypic varia-
tion for total biomass exists in cultivated (Austin
et al., 1980; Shearman et al., 2005) and novel
genetic sources (e.g., chromosomal segments
1BL.1RS or 7DL.7Ag—Foulkes et al., 2007).
However, their effects are commonly small,
making selection diffi cult. Phenotypic selection
for biomass is challenging as it is diffi cult to
measure accurately and has low heritability,
especially in early, segregating generations of a
breeding program (Rebetzke et al., 2002).
Processes of yield determination can be further
partitioned to refl ect use of available soil water, as
done in the widely used model fi rst enunciated by
Passioura (1977). The model describes grain yield
determination under water-limited conditions in
terms of water use or evapotranspiration (ET),
effi ciency of water use (WUE), and harvest
index:
DEFINING THE BREEDING TARGET
The multifaceted nature of drought and the
requirement for improved effi ciency of water use
dictate a reliance on multiple traits that account
for the nature, timing, and variability of water
limitation in the target environment. Empirical
selection for yield in dry environments has led
wheat breeders in many parts of the world to
develop well-adapted wheat cultivars primarily
through selection for earlier fl owering and reduced
height (Richards et al., 2002; Slafer et al., 2005).
As alleles for appropriate phenology and height
become fi xed, the opportunity then arises for
identifi cation and selection of alleles for new traits
such as greater early vigor, increased transpira-
tion effi ciency, or greater remobilization of stored
stem carbohydrates, which extend adaptation to
water-limited environments.
In its simplest form, grain yield (GY) refl ects
the partitioning of dry matter to grain and can be
expressed mathematically as:
GY = ET × WUE × HI
Crop WUE can be further partitioned into two
components refl ecting the portion of total water
transpired by the crop (T/ET) and the transpira-
tion effi ciency of biomass production (W). The
resulting framework is given as:
GY = TDM × HI,
in which TDM is the total aboveground dry
matter or biomass at harvest maturity, and HI is
harvest index, a measure of the total dry matter
allocated to grain. Genetic advances in grain yield
have largely been achieved through increases in
harvest index (Austin et al., 1980; Perry and
D'Antuono 1989; Shearman et al., 2005; Zhou et
al., 2007). Increases in harvest index have been
attributable to changes in phenology (anthesis
date) and incorporation of height-reducing genes,
including the major
Rht-B1b
and
Rht-D1b
dwarf-
ing genes promoted widely during the Green
Revolution. The yield component most affected
through this change is kernel number (Perry and
D'Antuono 1989; Shearman et al., 2005; Zhou et
al., 2007). There appears to be a limit to the pro-
portion of biomass partitioned to the grain, and
this value is estimated to be 50% to 60% (Austin
GY = ET × T/ET × W × HI
This is not the only possible partition of grain
yield, as others have been developed to refl ect
grain yield components (e.g., kernel number and
size), and others around radiation capture and
conversion to dry matter. The benefi t in the Pas-
sioura framework refl ects the broad processes by
which crops actually achieve yield in water-
limited environments (Condon et al., 2004). A
number of characteristics have been identifi ed
affecting ET, T/ET, W, and HI, and thereby
yield improvement under drought. Many of these
traits are physiologically independent, allowing
genetic effects to be accumulated through
