Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

abstracts

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The efficiency of resource capture is an important component of crop performance. In the case of soil resources (water and nutrients), this efficiency relies on the distribution of roots in the soil and, in particular, on its matching the distribution of these resources. Root system architecture (RSA; a concept that encapsulates the form and the structure of the root system) can be used as a bridge between root distribution and the underlying biological processes. Because the nature of RSA variation is quantitative and its analysis is complicated by its strong responses to environmental conditions, the genetic analysis of root traits in cultivated grass species has mainly been investigated through the identification of quantitative trait loci (QTLs). For practical reasons, most of these QTL studies focused on global traits like the total mass or volume of the root system, the root:shoot ratio or the distribution of root mass within specific soil layers, and on few architectural traits like the maximum length, number and diameter of root axes. The major objective of this thesis was to explore the pros and cons of using fine and dynamic representations of root systems as phenotypic input in QTL studies. It relies on a QTL analysis of such representations of the adult RSA in rice. Firstly, this detailed architectural dissection helped to clarify the biological processes underlying the QTL effects of global root traits through their dissection into root elongation or root production QTLs. Secondly, QTLs were identified for static estimators of dynamic root traits (elongation and emergence rates) in spite of their high response to environmental influence. Thirdly, QTLs for the elongation rate were redundant of QTLs for the maximum root length, i.e. a trait that corresponds to the integration of the elongation rate over time. Fourthly, a modelling attempt to extract dynamic information from static root measurements highlighted the inadequacy of linear functions to describe the evolution of the emergence of adventitious root axes in rice. Fifthly, a classification of root traits into categories relating specifically to elongation vs. production processes was proposed and allowed to substitute partly for the lack of standards in root trait measurements among previous root QTL studies.