Differential effects of partial spikelet removal and defoliation on kernel growth and assimilate partitioning among wheat cultivars

Abstract

Among soft red winter wheat ( Triticum aestivum L.) cultivars, compensatory kernel set does not occur with 50% spikelet removal (SR) at anthesis, but there are differential responses in final kernel mass. It seems that these cultivar differences resulted from sink limitations for nonresponsive cultivars and from source limitations for responsive cultivars. Individual kernel masses and kernel growth rates of sink-limited cultivars are not expected to respond to an increased ratio of source to sink and should be less affected than source-limited cultivars when this ratio is reduced. To compare kernel growth traits of six cultivars that differed in response to 50% SR in a previous experiment, a range of SR treatments and defoliation were used to manipulate ratios of source to sink. Two field experiments were conducted for four years. In the first experiment, all spikes on a plant had 0, 25, 50, 75 and 100% SR treatments at anthesis. In the second experiment, sink size was reduced at anthesis by restricting 50% SR to a single culm, and source size was reduced at anthesis by removing all leaf blades from a single culm. Partial SR increased final kernel mass of only responsive cultivars (15–25%), but differences in final kernel mass between the 50 and 75% SR-treated plants were not significant. At maturity, water-soluble carbohydrate levels in stems of all cultivars with 75% SR were less than or equal to those with 100% SR, but greater than those with ≤ 50% SR. Partial SR increased the kernel growth rate of responsive cultivars, which increased the final mass of individual kernels. Kernel N concentrations also were increased. Defoliation decreased kernel growth rate and individual kernel mass of all cultivars; relative decreases were greater for responsive cultivars than for nonresponsive cultivars. The mass and kernel growth rate of individual kernels of responsive cultivars were normally source-limited, while source and sink strengths of nonresponsive cultivars were more closely balanced than those of responsive cultivars.