Stable carbon isotope discrimination (Δ) showed both intraspecific and interspecific variations in leaves collected from field plants representing a wide range of peanut germplasm grown in similar environments. It was predicted, on the basis of theoretical models relating Δ and water-use efficiency (W), that there could be as much as 60% variation in W. The Δ of leaf material was used as a guide to select peanut genotypes for testing, in the glasshouse, for a correlation between Δ and W. Pot plants of nine peanut genotypes were grown in conditions of unlimited water availability or very restricted water supply. Water-use efficiency and Δ of leaves was measured on all plants. A strong correlation (r = -0.81) was found between Δ and W. Water-use efficiency actually varied as much as twofold. Carbon content and Δ were measured in all parts of some plants. The Δ values of all parts were highly correlated with and therefore well-represented by the Δ of the leaves. Calculation of W on a molar basis removes variation associated with carbon content of dry matter. On this basis, W was correlated with the discrimination against 13C of the whole plant. The link between Δ and W is not direct but the strong correlation between Δ and W suggests that other factors, which are discussed, do not interfere greatly. The data provide more evidence that Δ might be used as a selection technique for W in C3 species.

Water-soluble carbohydrates (WSCs; composed of mainly fructans, sucrose [Suc], glucose [Glc], and fructose) deposited in wheat (Triticum aestivum) stems are important carbon sources for grain filling. Variation in stem WSC concentrations among wheat genotypes is one of the genetic factors influencing grain weight and yield under water-limited environments. Here, we describe the molecular dissection of carbohydrate metabolism in stems, at the WSC accumulation phase, of recombinant inbred Seri/Babax lines of wheat differing in stem WSC concentrations. Affymetrix GeneChip analysis of carbohydrate metabolic enzymes revealed that the mRNA levels of two fructan synthetic enzyme families (Suc:Suc 1-fructosyltransferase and Suc:fructan 6-fructosyltransferase) in the stem were positively correlated with stem WSC and fructan concentrations, whereas the mRNA levels of enzyme families involved in Suc hydrolysis (Suc synthase and soluble acid invertase) were inversely correlated with WSC concentrations. Differential regulation of the mRNA levels of these Suc hydrolytic enzymes in Seri/Babax lines resulted in genotypic differences in these enzyme activities. Down-regulation of Suc synthase and soluble acid invertase in high WSC lines was accompanied by significant decreases in the mRNA levels of enzyme families related to sugar catabolic pathways (fructokinase and mitochondrion pyruvate dehydrogenase complex) and enzyme families involved in diverting UDP-Glc to cell wall synthesis (UDP-Glc 6-dehydrogenase, UDP-glucuronate decarboxylase, and cellulose synthase), resulting in a reduction in cell wall polysaccharide contents (mainly hemicellulose) in the stem of high WSC lines. These data suggest that differential carbon partitioning in the wheat stem is one mechanism that contributes to genotypic variation in WSC accumulation.Water-soluble carbohydrates (WSCs) can accumulate in the stem and leaf sheath of cool-season cereals (e.g. wheat [Triticum aestivum], barley [Hordeum vulgare], and oats [Avena sativa]) during the period from stem elongation to the early phase of grain filling and serve as temporary carbohydrate reserves, commonly called the stem carbohydrate reserves (Schnyder, 1993;

Grain yield and grain weight of wheat are often decreased by water-limitation in the north-eastern cropping belt of Australia. Based on knowledge that CIMMYT lines are well-adapted in this region, a recombinant inbred line (RIL) population between two elite CIMMYT bread wheats (Seri M82 and Babax) was evaluated under water-limited environments. Fourteen productivity traits were evaluated in 192 progeny in up to eight trials. For three aggregations of the environments (all, high yield or low yield), multiple quantitative trait loci (QTL) were detected, each explaining <15% of variation. Co-location of multiple trait QTL was greatest on linkage groups 1B-a, 1D-b, 4A-a, 4D-a, 6A-a, 6B-a, 7A-a and an unassigned linkage group. Two putative QTL (LOD > 3) from Seri (6D-b and UA-d) increased grain yield and co-located with a suggestive (2 < LOD < 3) and a putative QTL for increased stem carbohydrate content (WSC), respectively; the latter QTL also co-located with a putative anthesis QTL for earlier flowering. Both QTL were detected only in high yield (>4t ha(-1)) environments. A third increased grain yield QTL (7A-a) from Babax co-located with QTL for increased grain number. Six putative QTL increased grain weight and co-located with QTL for harvest index, grains per spike and spike number. Three putative QTL for increased grains per spike co-located with strong QTL for earlier flowering, increased grain weight and fewer spikes. A group of progeny that exceeded the mean grain yield and grain weight of commercial checks had an increased frequency of QTL for high WSC, large grain size, increased harvest index and greater height, but fewer stems, when compared to low yielding (20% less), low grain weight progeny. These findings were consistent with agronomic analyses of the germplasm and demonstrate that there should be opportunities to independently manipulate grain number and grain size which is typically difficult due to strong negative correlations.