Résumé
We have analysed daily patterns of leaf elongation rate (LER) in large data sets with 318 genotypes placed in naturally fluctuating temperature and evaporative demand, and examined the effect of targeted alleles on these patterns. The method consisted, firstly, in expressing elongation rate per unit thermal time, so it became temperature independent; secondly, in a joint analysis of diurnal fluctuations of elongation rate and of micrometeorological conditions in several experiments, and finally, in a comparison of daily patterns between groups of genotypes possessing targeted alleles. (1) Conditions for using thermal time at a time step of 15 min were first tested successfully in 318 recombinant inbred lines (RILs) of three mapping populations. (2) An analysis of 1989 time courses revealed a robust daily pattern of LER per unit thermal time (LERth) over several experiments. LERth was constant during the night and was reproducible (for a given RIL) over up to 10 consecutive nights in different experiments. It declined rapidly during the early morning, closely following the daily pattern of transpiration rate. (3) Groups of RILs carrying alleles conferring a high response to temperature had markedly higher night-time plateau of LER than those with low responses. Groups of RILs with high response to evaporative demand had rapid decreases in elongation rate at the transition between night and day, while this decrease was slower in groups of RILs with low response. These results open the way for using kinetics of responses to environmental stimuli as a phenotyping tool in genetic analyses.