Early Selection for high spike fertility index in early generations increases grain number per unit area (and yield) of advanced breeding lines in bread wheat

Abstract

Spike fertility index (a.k.a. fruiting efficiency) at maturity (FEm -grain number per spike/g spike chaff at maturity-) and spike harvest index (SHI -grain weight per spike/g spike chaff weight at maturity-) are considered promising selection criteria for improving grain yield (GY) in bread wheat. However, their usefulness in early generations remains to be established. Here, a population derived from a cross between ‘Baguette 10′ and ‘Klein Chajá’ was evaluated from early generations (F2-F3) to recombinant inbred lines (RIL). Associations of FEm and SHI measurements between early generations and RIL were established. Also, responses in grain number/m2 (GN) and GY in the RIL after simulated selection for high FEm or high SHI in the F2-F3 were assessed. Positive correlations were found in FEm measured in the F2-F3 vs. in the RIL (r = 0.32, P < 0.05) and SHI measured in the F3 vs. in the RIL (r = 0.26, P < 0.05). After selection for high FEm in the F2-F3, responses in GN were always positive: GN increased by 5.1 to 22.1% depending on the selection strategy. Responses in GY were positive in three out of four selection strategies, averaging -1.3 to 2.1%. Responses to selection in GN and GY in the RIL using SHI as a selection criterion in the F2-F3 were negative or neutral. Our results support the use of FEm as a selection criterion in early generations for increasing GN and GY in advanced generations of bread wheat breeding programs.