The effects of genotype and environment on tillering patterns in spring wheat
Grain yield in spring wheat, Triticum aestivum L., is in part determined by the number of spikes per unit area. With the exception of high plant density crops, spike numbers are in turn the end-product of the tillering process common to all grass species. A series of experiments was conducted to evaluate the tillering patterns of diverse spring wheat genotypes under the short growing season conditions of central Saskatchewan. In addition, experiments were carried out to determine how consistently tillering patterns were expressed in different years, locations, seeding rates, seeding dates, seeding depths and temperatures. Significant genotypic variation was detected for the following tillering pattern components: time to tiller emergence, maximum tiller number, tiller mortality and spike number. Maximum tiller number consistently exhibited the greatest variability among genotypes with a five and 2.5-fold range at 5 and 100% of the commercial seeding rate (230 seeds/m2), respectively. Genotypes which produced large tiller populations did so by producing more secondary tillers as a result of higher individual tiller emergence and survival frequencies. Cultivars in which tillers emerged when the main stem was less developed had higher tiller grain yields but were late maturing. Lower tillering cultivars were characterized by a lower synchronization of tiller emergence and a greater phytomass differential between the main stem and its second leaf tiller at the mid- to late-tillering phase. On a crop basis there was no relationship between a genotype's yielding ability and either its tillering capacity or tiller mortality. Cultivars with diverse tillering habits (free-tillering, low-tillering and oligoculm) responded differently to changing environments as illustrated by cultivar x seeding rate and cultivar x year interactions for tillering characteristics and yield components. The genotype x environment (G x E) interactions for tillering did not involve changes in cultivar rank while those for a number of yield components did. The G x E interactions for yield components were ascribed to a pronounced reduction in the means of the low-tillering and oligoculm cultivars at higher seeding rates under extended pre-anthesis drought conditions. Several years of testing led to the conclusion that oligoculm cultivars are not suited to the erratic precipitation distributions common to central Saskatchewan. Under field and/or controlled environment conditions increases in seeding depth (1 to 3 cm) had a slight effect on tillering, reducing the rate of tillering, shortening the tillering phase, and reducing maximum tiller number, tiller mortality and spike number. Under controlled environment conditions temperature was shown to have a pronounced effect on the rate of tillering but little impact on maximum tiller number. Similarly, under field conditions delayed seeding had little effect on maximum tiller number. Heritability estimates for maximum tiller numbers based on the variance component method were high and G x E effects negligible, indicating that genotypic tillering characteristics are expressed consistently across environments. A lack of heterosis combined with few deviations from normality in F2 population distributions suggests that tillering capacity is controlled by genes with largely additive effects. It is suggested that selection for maximum tiller number could be carried out during early generations in solid seeded plots. However, the fact that grain yields were not associated with a particular tillering capacity leads to the conclusion that the manipulation of this trait would not be a productive breeding venture.