Abstract
The quality of yield prediction is linked to that of leaf area. We first analysed the consequences of flowering time and environmental conditions on the area of individual leaves in 127 genotypes presenting contrasting flowering times in fields of Europe, Mexico and Kenya. Flowering time was the strongest determinant of leaf area. Combined with a detailed field experiment, this experiment showed a large effect of flowering time on the final leaf number and on the distribution of leaf growth rate and of growth duration along leaf ranks, in terms of both length and width. Equations with a limited number of genetic parameters predicted the beginning, end and maximum growth rate (length and width) for each leaf rank. The genotype-specific environmental effects were analysed with datasets in phenotyping platforms that assessed the effects (i) of the amount of intercepted light on leaf width, and (ii) of temperature, evaporative demand and soil water potential on leaf elongation rate. The resulting model was successfully tested for 31 hybrids in 15 European and Mexican fields. It potentially allows predicting the vertical distribution of leaf area of a large number of genotypes in contrasting field conditions, based on phenomics and on sensor networks.