Emergence and elongation dynamics of maize (Zea mays L.) roots in aeroponics : diversity and genetic studies

Root systems are important components of crop performance. Roots are primarily responsible for plant anchorage and for the acquisition of soil resources. However, the use of roots traits as selection criteria in breeding programs is fraught with practical difficulties: the low throughput of root assessment methods, our limited understanding of the genetic bases of root growth and development, the extreme phenotypic plasticity of root systems and the lack of field-proven ideotypes. This thesis uses a novel aeroponics phenotyping infrastructure to unravel the genetic diversity and the genetic determinism of dynamic root traits in young maize plants (5th leaf stage). It also exploits model-based formalisms of root system dynamics to perform in-depth analysis of genetic correlations between traits. The genetic diversity experiments, conducted with a wide diversity panel of 86 inbred lines, revealed an abundant genotypic variation for root elongation, emergence and tropisms, with broad-sense heritabilities ranging between 35 and 95%. Surprisingly, the genotypic variation did not relate with the geographic origin, which suggests that climate scenarios are not a major driver of root system adaptation. Our study of the genetic determinism was achieved with a QTL analysis for the same traits, using three mapping populations derived from crosses of CMLP1 x CMLP2, FV2 x Io and FV2 x FV252 (respectively P, D and E). A total of 38 QTLs was identified using these three mapping populations. Individual QTLs explained more than 10% of the total phenotypic variation. They contributed also to a large extent of transgressive segregation. In both studies, a remarkable stability of the elongation rate was observed at the individual root level. In other words, individual roots adopted a given elongation rate at the time of their emergence, and tended to maintain this elongation rate throughout the experiment. We have conducted a preliminary analysis to correlate this elongation rate with indicators of source and sink activity. At this stage, however, we have not detected any significant relationships. In our thesis, we have also designed and evaluated a low-tech high-throughput root phenotyping method, in order to show that basic procedures of image analysis of roots grown in artificial conditions can be as powerful as comprehensive manual root tracing to reveal differences between genotypes. The new insights and questions brought by this thesis on the dynamics of root growth and development will drive future research and support crop improvement and adaptation for low-input agriculture and low-tech R&D.