Selection for Reduced Carbon Isotope Discrimination Increases Aerial Biomass and Grain Yield of Rainfed Bread Wheat G. J. Rebetzke,* A. G. Condon, R. A. Richards, and G. D. Farquhar ABSTRACT targeting traits closely associated with improved plant adaptation to stress. A number of characteristics have Genetic gain is characteristically slow when selecting directly for been proposed as indirect selection criteria for genetic increased grain yield under water-limited conditions. Genetic in- improvement of drought resistance in a breeding pro- creases in grain yield may be achieved through increases in aerial biomass following selection for greater transpiration efficiency (TE gram (e.g., Ludlow and Muchow, 1990). Some charac- as aerial biomass/water transpired). Strong negative correlations be- ters enable plant survival under drought, while others tween TE and carbon isotope discrimination () in wheat (Triticum afford temporal protection at critical periods in crop aestivum L.) suggest that selection of progeny with low  may increase development (Richards, 1996). Although these charac- TE and aerial biomass under water-limited conditions. This study teristics may protect plants from dehydration stress, they investigated how early generation, divergent selection for  affected may not improve yield under drought. Genotypic in- aerial biomass and grain yield among 30 low- and 30 high-, ‘Hartog’- creases in wheat yields under drought should be associ- like, BC 2 F 4:6 progeny and the recurrent, high- parent Hartog. Lines ated with increases in aerial biomass (Fischer and Wood, were evaluated in nine environments varying for seasonal rainfall 1979). Greater water use efficiency (WUE as aerial bio- (235–437 mm) and hence grain yield (1.3–6.2 Mg/ha). Selection for mass/total water use) should provide more biomass for low  in early generation progeny was associated with significantly (P  0.01) smaller , higher grain yield (+5.8%), aerial biomass crops growing in water-limited conditions (Passioura, (+2.7%), harvest index (+3.3%), and kernel size (+4.8%) in tested 1977). Transpiration efficiency (TE as aerial biomass/ lines. Kernel number was the same for low- and high- selected water transpired) is an important component of crop groups. Grain yield advantage of the low  group increased with WUE especially in regions where stored soil water is reductions in environment mean yield (r =-0.89, P  0.01) and total a major component of crop water use (Condon and seasonal rainfall (r =-0.85, P  0.01) indicating the benefit of low Richards, 1992). Repeatable genetic variation has been , and therefore high TE for genetic improvement of grain yield in reported for TE in wheat (Condon et al., 1990; Ehdaie lower rainfall environments. Narrow-sense heritability on a single- et al., 1991; Malik et al., 1999), yet its direct use for plot basis was much greater for  (h 2 = 0.63  0.10) than for either breeding has been constrained by the lack of a suitable aerial biomass (0.06  0.05) or grain yield (0.14  0.04). Strong genetic screening methodology in large segregating populations. correlations between  and both aerial biomass (r g =-0.61  0.14) and grain yield (-0.58  0.12) suggest  could be used for indirect Carbon isotope discrimination () is correlated nega- selection of these traits in early generations. Selection of low  (high tively with TE in wheat (Farquhar and Richards, 1984; TE) families for the advanced stages of multiple-environment testing Ehdaie et al., 1991) and in other C 3 species (see Hall et should increase the probability of recovering higher-yielding wheat al., 1996). Selection for low  may provide a useful families for water-limited environments. method for indirect selection of TE and perhaps biomass and grain yield in cereal breeding programs for water- limited environments (Hall et al., 1996; Voltas et al., W heat is commonly grown in regions where grain 1999). Plant  has many features that make it desirable yields are limited by low seasonal rainfall (Ste- for implementation in a breeding program targeting in- phens and Lyons, 1998; van Ginkel et al., 1998). Growers creased yield. For example, leaf  integrates TE over in these regions rely on wheat varieties selected for the period for which leaf tissue is formed, leaf tissue improved yield under drought. Empirical selection in can be sampled rapidly on a large number of families, the development of wheat varieties for water-limited and repeatability for  is high for leaf tissue formed in conditions is difficult, slowing genetic gain for yield in the absence of moisture stress (Condon and Richards, these environments (Feyerherm et al., 1984; Blum, 1992). 1988). Differences in seasonal distribution and severity Genotypic selection for low  shows potential for of water limitation can vary substantially from season improvement of TE of wheat (see Hall et al., 1996), yet to season for a given testing location. In turn, large its utility for indirect improvement of grain yield in an genotype  environment interactions contribute to a applied breeding program has not been investigated. lower heritability, thereby reducing confidence in selec- The effectiveness of correlated genetic gain requires tion of superior yielding genotypes (Calhoun et al., 1994; that the secondary trait have both a high narrow-sense van Ginkel et al., 1998). heritability and strong additive genetic correlation with Genetic gain in grain yield may be achieved through the primary trait (Falconer and Mackay, 1996). Previous reports for broad-sense heritability of  in wheat were high when expressed on an entry-mean basis (Ehdaie G.J. Rebetzke, A.G. Condon and R.A. Richards, CSIRO Plant Indus- et al., 1991; Condon and Richards, 1992). Phenotypic try, P.O. Box 1600, Canberra ACT 2601 Australia; G.D. Farquhar, Australian National Univ., P.O. Box 475, Canberra ACT 2601 Austra- correlations between  and grain yield are typically high lia. Received 1 Sept. 2001. *Corresponding author (G.Rebetzke@pi. (Condon and Richards, 1993), and either positive or csiro.au). negative, depending on the plant tissue analyzed and the yield potential of environments sampled. The objec- Published in Crop Sci. 42:739–745 (2002). 739