Selection for high grain number per unit stem length through four generations from mutants in a durum wheat population to increase yields of individual plants and crops

• Published in: Field crops research 2012-04, Vol.129, p.59-70
• Main Authors: Pedro, Anna, Savin, Roxana, Parry, Martin A.J., Slafer, Gustavo A.
• Format: Article
• Language: English
• Subjects: Grain yield; Physiological traits; Selection criterion; Triticum aestivum; Triticum durum
• ISSN: 0378-4290; 1872-6852
• DOI: 10.1016/j.fcr.2012.01.016

► We characterised a segregating population of mutant lines, looking for expressions of improved yield. ► We determined a physiological trait representing extreme source-sink balances and selected divergently for it. ► In each generation we determined physiological traits explaining the performance of the selected lines. ► At the end of the selection process the best selections clearly outyielded plants of the controls. ► Increasing the efficiency of conversion of growth during stem elongation into grains may be relevant to further raise wheat yield. In order to identify genotypes of durum wheat expressing improved yield and determining whether selection for individual plant efficiency in producing grains would improve plant and crop yield, we firstly compared the performance of segregating mutants and control cultivars and then selected divergent lines for number of grains per unit stem length, representing extreme source-sink balances further characterising the offspring lines. Research included three field experiments with individual segregating plants within the semi-arid region of the Ebro Valley during the 2006/07, 07/08 and 08/09 growing seasons; a fourth experiment with segregating plants conducted under controlled conditions at Rothamsted in 2008/09 and finally a conventional field plot experiment with the most promising lines derived from the selection. In the initial characterization of the selected mutants, there was a clear response to the selection made in M2, and differences in plant yield were due to differences in grains per plant and biomass. There was a clear response to selection for sink-strength in improving the number of grains exhibited by the offspring lines through the generations (M2–M5) and the overall differences between the two divergent groups were consistently clear throughout the whole process. Response to selection was similarly expressed in field and in controlled conditions, implying high broad sense heritability of grains per plant (when there were no consistent differences in height and then it reflects grains per unit stem length). The best selections outyielded plants of the controls, without reducing the protein concentration of the grains. It seems clear that increasing the efficiency of conversion of growth during stem elongation into grains may be a relevant trait to further raise yield and that the trait used in the present study as a proxy of such efficiency (the number of grains per unit stem length) seemed ap