Increasing grain yield in bread wheat (Triticum aestivum) by selection for high spike fertility index

Spike fertility index (SF) has been proposed as a promising selection criterion for increasing grain yield (GY) in bread wheat. Here, changes in GY and related traits after simulated selection (10% intensity) for high SF or high GY per se were assessed in two RIL populations: (I) Avalon/Glupro and (II) Baguette 10/Klein Chaja, and in (III) advanced lines from a breeding programme. Grain yield, SF, grain number per unit area (GN), grain weight (GW), test weight (TW) and grain protein content (GPC) were determined. Regardless of the environmental conditions, simulated selection for high SF always resulted in GN increases (between 1.6% and 27.4%). Average GY increase observed after selection for high SF (11.5%; N = 10; SEM = 5.8) did not differ (p = .92) from the average GY increase observed after selection for GY per se (11.8%; N = 10; SEM = 4.9). Grain weight, GPC and TW tended to decrease with selection for high SF; however, these trade-offs might be avoided by concurrent selection. Our findings validate the use of SF as a selection criterion for increasing grain yield in bread wheat.