Fruiting efficiency in wheat: physiological aspects and geneticvariation among modern cultivars

Efforts to enhance the yield of wheat require increases in the number of grains per m2 (GN) which depends on spike fertility, i.e. the number of fertile florets or grains per spike. Fruiting efficiency (FE), the number of grains set per unit spike dry weight at anthesis, has been proposed as an alternative trait to improve spike fertility and thus GN. The use of FE in realistic breeding programs requires genetic variation within well-adapted, high-yielding modern cultivars and a lack of trade-off with other yield components. We quantified FE variation within Spanish elite material and determined whether the variation implied trade-offs with spike dry weight or grain weight (GW). Field experiments under contrasting nitrogen conditions were carried out with nine elite genotypes (eight modern cultivars and one advanced line) in which the number of fertile florets and grains were determined for different spikelets and the whole spike. Furthermore, the partitioning of dry matter was investigated at anthesis and maturity. Overall, the yield was strongly related to GN because environmental and genetic factors contributed to this relationship, although as expected for elite material the environmental effect was much greater than the genotype. Environmental effects on GN mainly influenced spike dry weight at anthesis (SDWa) whereas most genotypic differences were due to FE variation. Averaging across environments, the FE was c. 110–170 grains gspike−1. Although some G × E interactions were observed, there was no reversal in ranking between genotypes with extreme FEs across environments, and genotypic differences across environmental conditions were consistent (h2 = 0.46). The relationship between FE and the individual weight of fertile florets tended to be negative, although rather than a constitutive reduction in floret size driving the improvements in FE, this relationship appeared to reflect the increased proportion of smaller grains (distal florets in each spikelet) due to changes in spike fertility. FE was not related to grain weight (either the average of all grains or that of the most proximal grains). Therefore, differences in FE among the elite genotypes did not imply constitutively-linked changes in grain weight. When chaff dry weight was used as a proxy for SDWa to estimate genotypic variation in FE, we observed the consistent underestimation of FE and also changes in the ranking of genotypes.

Funding was provided partially by ADAPTAWHEAT (a project funded by the European Commission under the 7th Framework Programme for Research and Technological Development) and partially by project AGL2012-35300 funded by the Ministry of Economy and Competitiveness of Spain.