Coarse and fine regulation of wheat yield components in response to genotype and environment

Abstract

Wheat yield is frequently analysed in terms of yield components (spikes per m2, grains per spike, grain size), and correlations between components are commonly reported, but partially understood. Compensations between components are one of the main barriers to improve yield using this approach. Understanding better the nature of these relationships is therefore important. We compiled a large database (n=846) of wheat yield (ranging from 1.3 to 14.8 Mgha−1) under field conditions to further investigate the relationships between yield components from the viewpoint of phenotypic plasticity. We found two main hierarchies between plasticities. Hierarchy 1, grains per m2≫grain size, has been widely described before and is supported empirically and theoretically; here this hierarchy lends confidence in our analytical method. Hierarchy 2, spikes per m2>grains per spike, is demonstrated for the first time. These hierarchies are interpreted in terms of coarse and fine regulation of grain yield (Hierarchy 1), and grains per m2 (Hierarchy 2). This means small changes in yield or grains per m2 can be accounted for by any of their components, but large changes in yield can only be accommodated by grains per m2, and large changes in grains per m2 are primarily associated with heads per m2. Further, we showed that the rankings of responses of yield components depend on the nature of the driver, i.e. genotype or environment. Both hierarchies are much stronger when the driver is environmental factors (e.g. water supply). When the driver is genotype, and this means we are not strictly dealing with plasticity but trait responsiveness to genotype, the rankings are weaker. Strong, negative relationships between components were found for small changes in yield or in grains per m2 for both environmental and genetic factors, implying strong trade-offs. For large changes, trade-offs were generally less relevant but much stronger for components of yield than for components of grains per m2. For large changes in grains per m2 associated with environmental factors, the negative relationship between grains per spike and of spikes per m2 was weak but when they were associated with genetic factors, the negative relationship was strong. In conclusion, our study reinforces the hierarchy of plasticities between components of grain yield, reveals a hierarchy between the components of grains per m2, and demonstrates that the strength of the trade-off between components of either yield or grains per m2 depends upon the magnitude of yield changes and, for large changes, the source of variation. A practical corollary is that breeding and agronomy could potentially achieve modest to intermediate yield increase involving any components, which would exhibit some trade-offs. Agronomic interventions are more likely to lead to large yield improvement, mainly through increase in grains per m2 with negligible trade-offs in grain size.