Number of kernels in wheat crops and the influence of solar radiation and temperature

Abstract

The number of kernels per m2 (K) in well managed and watered wheat crops was studied using results of experiments in Mexico and Australia in which short spring wheat cultivars were subjected to independent variation in radiation, largely via artificial shading, and in temperature. Also crops subjected to differences in weather (year), sowing date and location within Mexico, revealed responses to the natural and simultaneous variation which occurs in radiation and temperature. Responses in K were interpreted in terms of spike dry weight at anthesis (g/m2) and number of kernels per unit of spike weight. K was linearly and most closely related to incident solar radiation in the 30 days or so preceding anthesis, herein termed the spike growth period; for the cultivar Yecora 70 with full ground cover the slope was 19 kernels/MJ. This response seemed largely due to a linear response of crop growth rate to intercepted solar radiation. The proportion of dry weight increase partitioned to the spike increased somewhat with reduced radiation. Also increasing temperature in the range 14–22 °C during this period reduced K (slope approximately 4% per °C at 15 °C). The cause appeared to be lower spike dry weight due to accelerated development. The number of kernels per unit spike weight at anthesis was little affected by radiation or temperature, and averaged 78 ± 2/g for the cultivar Yecora 70. With natural variation in radiation and temperature, K was closely and linearly correlated with the ratio of mean daily incident or intercepted radiation to mean temperature above 4•5 °C in the 30 days preceding anthesis. As this ratio, termed the photothermal quotient, increased from 0•5 to 2•0 MJ/m2/day/degree, K increased from 70 to 196 x 102/m2. These responses of K to weather, sowing date and location were closely associated with variation in spike dry weight. It was concluded that the ratio of solar radiation to temperature could be very useful for estimating K in wheat crop models. Also the analysis of K determination in terms of spike dry weight appeared promising, and suggests that wheat physiologists should place greater emphasis on the growth period immediately before anthesis. © 1985, Cambridge University Press. All rights reserved.