Positive contribution of stem growth to grain number per spike in wheat

Abstract

This paper challenges the common belief that spike:stem competition determines grain number in wheat. A field experiment was conducted using three contrasting treatments (high shoot density, shading, and extended photoperiod) and six genotypes of varying earliness and spike:stem ratio. Floret primordia were counted in medium and top spikelets of main spikes throughout stem elongation. Floret death occurred simultaneously in both spikelets, and the corresponding critical time was used for comparison with growth analysis. Whole shoot growth rate was highest at the time of floret death, but both stem and spike growth continued to accelerate for a period of time thereafter. Spike respiration measurements did not indicate special requirements at the critical time. Spike carbohydrate concentration started to decline more or less before floret death, depending on genotype, but exactly at the inflexion time for spike elongation, thus suggesting a developmental process rather than a trophic limitation. Multiple correlation indicated that grain number positively correlated not only with spike growth rate at the time of floret death, but with both growth rates of spike and non-spike (stem + leaves) shoot parts. Grain number was poorly predicted by partitioning to spike of either shoot dry mass or shoot growth rate at critical time. In contrast, floret surviving proportion was highly correlated to partitioning and multiple correlation did not indicate a significant effect on survival of non-spike growth rate. It is concluded that even though spike:stem competition decreased floret survival, it was only loosely correlated to grain number.