LAUSR

Grain number (GN) is determined by spike growth rate (SGR), fruiting efficiency (FE) and the duration of spike growth period (Ds). However, these three traits are not independent of each other and therefore quantifying their relative contribution to GN is important for improving yield potential. This study aimed to model GN as a function of SGR, FE and Ds in bread wheat and investigate the relative importance of these three traits in determining GN. A large number of commercial varieties, elite wheat breeding lines, and a Multiparent Advanced Generation InterCross population were evaluated for GN, SGR, Ds, and FE in six Australian environments in 2014 and 2015. The model explained 43–98% of the variation in GN in individual environments and 88% of the variation in GN across the six environments. The percentage of GN variance explained by the model was significantly higher than those explained by correlating GN to individual traits. The relative importance of the contribution of the three traits to GN was different and ranked as SGR > FE > Ds. The negative correlation between FE and spike dry weight at anthesis (SDWa) appears to be spurious, suggesting that there was no causality between high FE and low SDWa and vice versa. Thus, FE can be used as an independent trait to increase GN in addition to SGR and Ds. We also investigated if the wheat crop had enough source to fill the increased sink size in the southern Australian high rainfall environment. The source-sink balance analysis indicates that the available source, consisting of concurrent photosynthetic assimilate and stored water-soluble carbohydrate at anthesis, was sufficient to fill the established sink but no surplus assimilates were available to fill a larger sink in southwestern Australia. However, the opportunity to increase yield potential by further increasing grain number might exist in south-eastern Australia environments.