Why work and discuss the basic principles of plant modelling 50 years after the first plant models?

Plant modelling has not yet become a standard method in Plant Biology for integrating complex mechanisms, nor in Genetics for designing new plants. This contrasts with the situation in industry in which, for example, modelling has in good part replaced costly experiments in wind tunnels for the design of cars or planes. Pioneering papers on plant modelling were published as early as the 1950s, in particular by CT de Wit (De Wit, 1959). There has been constant progress since then with, for instance, the first models of water transfer in the soil–plant– atmosphere continuum in the 1960s and 1970s (Gardner, 1960; Sinclair et al., 1976; Molz, 1981) and the development of crop models in the 1980s and 1990s (Sinclair, 1986; Hammer et al., 1987; Williams et al., 1989). Plant and crop models are now regularly used to calculate transpiration at the canopy level, to predict the consequences of climate changes on crop production, or to design new cropping systems. Conversely, their use in Plant Biology has progressed more slowly, and the design of plant ideotypes with models is still in its infancy. The links between Mathematics and Plant Sciences have mainly involved the structure of genomes and gene networks rather than the simulation of phenotypes in a changing and fluctuating environment. Why is this progress so slow compared with that in other areas? Our colleagues in the ‘hard sciences’ are often puzzled that we are still debating the theoretical basis of plant models 50 years after the first models were published. This issue of the Journal of Experimental Botany aims to collect new approaches that will hopefully unlock some of the key questions associated with modelling. Three questions have been addressed.