Grain weight (GW) is a key trait of wheat affecting grain yield and quality. Advancements in the understanding of physiological and molecular control of grain weight and size have been significant over the past decade in wheat. Among others, the triple mutation of the TaGW2 gene, a known negative regulator of grain weight, resulted in a successful increase of grain weight both in rice and wheat.

However, the decrease of grain number (GN) was reported in wheat, potentially offsetting yield gains. The objective of this study was to evaluate the response of a triple mutant line of the TaGW2 gene and its wild type on grain yield, grain weight, grain number and the likely causes of the trade-off between both main yield components. A field experiment was conducted with two spring wheat genotypes: a triple mutant line of the gene TaGW2 and its wild type (WT) in Chillán, Chile.

These lines were arranged in a complete randomised block design with four replicates in field plots under optimal management. Data were analysed using ANOVA. As expected, GW was increased (P< 0.05) in the triple mutant by 20% over the WT. This positive impact was recorded in both main stems and tillers. On the other hand, grain yield was similar (P> 0.05) between the lines as GN was negatively affected in the triple mutant line showing similar decreases of GN between main stems (16.5%) and tillers (15.9%).

The trade-off between GW and GN was found due to the reduction of grains per spike and spikes per plant. These results supports that the effect of the triple mutant line of TaGW2 occurs during a long time of the wheat cycle, at least between tillering and anthesis.

The tillering capability of wheat was found affected in the line surprising the action of the TaGW2 gene. Additionally, the effect of the triple mutation was found across the spikelets of the spikes on both GW and GN. Differences in dry weight of spikes was found a week after anthesis, which is proposed as likely cause of the lower grain number per spike in the triple mutant line.