The importance of grain or kernel number in wheat: A reply to Sinclair and Jamieson

This paper is a response to Sinclair and Jamieson [Sinclair, T.R., Jamieson, P.D., 2006. Grain number, wheat yield, and bottling beer: an analysis. Field Crops Res. 98, 60–67] who propose that bulk carbon and nitrogen accumulation are fundamental to grain yield determination in wheat (Triticum aestivum L.), while challenging the common approach to yield through the separate processes of grain number determination, followed by grain filling, as governed by source–sink balance then. The response focuses on yield determination under potential conditions for which genetic and agronomic progress is clearly associated with increased grain number, herein abbreviated to KNO (kernels m−2). It argues that grain yield in modern cultivars is still limited by post-anthesis sink (KNO) and that understanding KNO determination is therefore useful for predicting physiological routes to higher yield. KNO determination appears to be strongly related to dry matter accumulation in spikes at anthesis (g m−2), governed by events in the last 20–30 days before anthesis, while some modern cultivars show higher grain number per unit spike weight. Post-anthesis photosynthesis and crop dry weight accumulation have increased as KNO has increased with breeding. There is no evidence for effects of N on KNO apart from those operating via dry matter accumulation, or for grain N demand limiting post-anthesis photosynthesis. Beyond this simple model, several other linkages that might exist between the pre- and post-anthesis periods are explored. Such linkages could help maintain the balance between the post-anthesis sink and the source required to fill the sink, and constitute common underlying processes which to some extent reconcile the model of Sinclair and Jamieson (2006) with current mainstream thinking about grain yield in wheat. Reports of an increasing amount of pre-anthesis carbohydrate reserves in the crop with breeding progress is a good example, but overall it is concluded that at least under potential conditions, the commonly accepted approach to grain yield determination is not invalidated by Sinclair and Jamieson (2006).