Conceptual Frameworks for Crop Yield Improvement.

Crop yield depends on the environment (E), technology including the genotype (G) and management (M), and their interactions. In the short to medium term (5-10 years), the environment over-rides technology. In the long-term (decades), technology increases yield, whereas the environment contributes to often significant deviations around time trends. Despite of its importance, the environment is often characterised superficially, e.g. nominally as location and season. In this paper, we discuss established and new methods to quantify probabilistically the water and thermal environments for crop production. We present case studies for field pea and chickpea in Australia, and their implications for crop improvement.

The GxE interaction is biologically interesting and agronomically important, as it is often a large component of the phenotypic variance of crop yield. Here we show how an approach based on phenotypic plasticity of crop traits helps to untangle complex GxE interactions. Combining this plasticity perspective with Fst genome scan, we show genetic profiles associated with phenotypic plasticity of yield, nitrogen fixation, carbon isotope discrimination for chickpea in diverse water and thermal regimes returning a range of yield from 1.1 t/ha under moderate stress to 5.1 t/ha under favourable conditions.

Natural selection favours competitive plants whereas selection for seed yield in agriculture favours less competitive types, which conform to the phenotype of Donald’s “communal plant”. Comparison of yield of chickpea under normal crop competition, i.e. central rows in stands sown at 55 plants m^-2 (Desi) or 30 plants m^-2 (Kabuli), and yield measured under relaxed competition in border rows showed high-yielding lines are less responsive to competition, in agreement with Donald’s theory. Fst genome scan highlighted the lack overlap in the genetic architecture underlying yield of crop stands and yield under relaxed competition.