stages [138]. This could be due to a number of factors, including hormonal imbalance, causing loss of
stomatal control on water loss after heading. Also, the plant material used for analysis could deter-
mine the level of heritability [138]. It was shown in a number of crops that the of leaf material is a
better indicator of differences in TE than that of grains [10,37,64,109,111,124]. One of the main rea-
sons could be genotypic differences in the ability to translocate preanthesis-stored carbohydrate re-
serves for grain filling [141].
The effectiveness of indirect selection for TE using will partly depend on the magnitude of the her-
itabilities for TE and and the genotypic correlation between these characters [142]. Broad-sense heri-
tability, which is the proportion of total phenotypic variance that is attributable to genotypic differences,
is a measure of the repeatability of the expression of those genotypic differences [138]. In many crops,
heritabilities for are above 80% [10,38,109,111,124,138,139].
D. Advantages of Using for TE Evaluations
Breeding for improved TE has been limited by the lack of screening tools for identifying desirable geno-
types under field conditions [112]. The^13 C discrimination technique makes it possible to survey a large
number of plants with a simple, albeit expensive, analysis of the leaf tissue [11]. As provides an inte-
grated estimate of TE, it has been suggested that measurement of may better differentiate among geno-
types than most instantaneous physiological assays [124]. Genotypic ranking based on is much more
consistent than that based on gas exchange measurements [112] and thus should be easier to select for in
breeding programs. Also, as remains reasonably constant throughout crop ontogeny, selection could be
made during crop development [74].
Further,is faster and easier to measure than total growth relative to total water use [29]. It is read-
ily determined on field-grown plants because it does not require the plant to be sheltered from rain or that
any other special experimental treatment be maintained. Measurements can be made on small plant sam-
ples collected at maturity with minimal problems of storage and handling. The material can be either leaf,
stem, or grain. Leaves and stems are easier to grind, and use of vegetative material has the potential ad-
vantage that selection can be made early in the crop growth cycle and thus could assist in improving se-
lection efficiency and reducing the time and maintenance costs [29,138].
TRANSPIRATION EFFICIENCY AND GENETIC IMPROVEMENT 845
Figure 3 Change in carbon isotope discrimination in leaves and error variation versus time for well-watered
groundnut cultivars of Tifton-8, Q 18803, and Chico grown in a greenhouse. (From Ref. 74.)