Wheat yield is not causally related to the duration of the growing season

A common conceptual model for crop growth analysis assumes biomass is proportional to the accumulation of radiation intercepted by the canopy during the growing season, and yield is the product of biomass and harvest index. This model assumes that biomass and harvest index are independent, and that biomass has the same value for yield across developmental stages; here we focus on the latter assumption. A corollary of this model is that yield and duration of the growing season are causally related. In this article, we analyse the evidence for and against this causal relationship. We use both a physiological perspective and agronomic evidence to challenge this causality. An alternative, phenology-focused model is outlined that is based on the following principles: crops accommodate environmental variation through grain number, grain number is defined in a species-specific critical developmental period, and grain number is a function of three traits in the critical period: duration, growth rate, and partitioning to reproduction. Against this framework, we analyse dual purpose wheat, where a substantial part of the biomass is removed, effectively shortening the growing season, with little or no effect on yield; timing of nitrogen fertilisation, where delayed application reduces the growth rate early in the season, with little or no effect on yield; and the correlation between season length and yield of winter wheat and spring wheat - wheat yield in Southern Chile, with a sowing-to-maturity season of ∼ 5 months, could be as high as that of winter wheat in the UK, with a sowing-to-maturity season of ∼ 10 months. Physiological principles and agronomic evidence support the conclusion that the duration of, and growth rate and partitioning during the critical period, rather than the duration of the growing season, are the drivers of wheat yield.