How close have historical domestication and modern breeding come to ideotypes for yield?

Optimizing crop architecture and allocation of resources to grain yield have been central goals in cereal breeding. Whether historical selection has achieved an ideotype maximizing yield under modern cultivation remains unclear. In this study, we explored how morphological traits have changed in wheat (Triticum aestivum) and whether they approach predicted optima for yield and resource allocation, using functional-structural plant (FSP) modelling. We used phenotypic data from a diverse panel of 142 wild, traditional and modern wheat accessions to parameterise and calibrate a functional-structural plant model. The model predicted which combinations of architectural and allocation trait values conferred maximal grain production under contrasting planting densities representing historical extensive and modern intensive agriculture. Our results show that historical changes to crop traits during domestication and landrace selection have converged on the ideotype by increasing the values of spike, internode and leaf carbon sink capacities. However, the specific internode length and leaf insertion angle became smaller following domestication, while the historical ideotype suggested these traits should increase. Under high planting density, the optimal values of these two traits would also be larger in modern ideotypes. Modern wheats have approached the ideotype by increasing potential spike biomass and decreasing the potential biomass of internode and leaf, but crop architecture could be improved. Historical selection has optimised biomass allocation for landrace cultivation under low densities. Allocation in modern varieties is also optimal for high-density modern agriculture, but further selection on architecture may increase yield potential.