Are crop and detailed physiological models equally ‘mechanistic’ for predicting the genetic variability of whole-plant behaviour? The nexus between mechanisms and adaptive strategies
Tardieu F, Granato ISC, Van Oosterom EJ, Parent B, and Hammer G L
in silico Plants
Tailoring genotypes for the variety of environmental scenarios associated with climate change requires modelling of the genetic variability of adaptation mechanisms to environmental cues. A large number of physiological mechanisms have been described and modelled, e.g. at transcript, metabolic or hormonal levels, but they remain to be assembled into whole-plant and canopy models. A ‘bottom-up’ approach combining physiological mechanisms leads to a near-infinite number of combinations and to an unmanageable number of parameters, so more parsimonious approaches are required. We propose that natural selection has constrained the large diversity of mechanisms into consistent strategies, in such a way that not all combinations of mechanisms are possible. These constraints, and resulting feedbacks, result in integrative ‘meta-mechanisms’, e.g. response curves of traits to environmental conditions, measurable via high-throughput phenotyping, and resulting in robust and stable equations with heritable genotype-dependent parameters. Examples are provided for the responses of developmental traits to temperature, for the response of growth and yield to water deficit and evaporative demand, and for the response of tillering to light and temperature. In these examples, it was inoperative to combine upstream mechanisms into whole-plant mechanisms, whereas the evolutionary constraints on the combinations of physiological mechanisms render possible the use of genotype-specific response curves at plant or canopy levels. These can be used for a new generation of crop models capable of simulating the behaviour of thousands of genotypes. This has significant consequences for plant modelling and its use in genetics and breeding.