Debieu, Marilyne (2011) Comprehensive phenotypic and genomic analyses of life-cycle variation in Arabidopsis thaliana. PhD thesis, Universität zu Köln.
Natural variation of life history strategies is modulated by both natural selection and physiological constraints. In this study, variation and co-variation of life history traits at the phenotypic and genomic level were studied in details in the model plant species Arabidopsis thaliana in order to elucidate the genetic and environmental factors controlling life-cycle evolution. Natural variation for growth rate, flowering time, primary seed dormancy and secondary seed dormancy were measured in a common environment across a set of 161 A. thaliana genotypes isolated from various locations throughout the species’ range. The results show that natural variation of life-cycle strategies was continuous. An antagonistic pattern of co-variation between flowering time and growth rate was observed. These two traits were negatively or positively correlated respectively depending on the latitude of origin of the genotypes. Functionality of FRIGIDA, a gene previously identified as a major contributor to flowering time variation, was associated with distinct evolutionary trajectories characterized by different patterns of co-variation between life history traits. Indeed, two negative correlations were observed specifically among genotypes carrying functional FRIGIDA allele: between flowering time and primary dormancy as well as between primary dormancy and secondary dormancy. These results indicate that selection on and trade-offs among traits controlling life-cycle strategies change along the distribution range of A. thaliana and are influenced by FRIGIDA functionality. Moreover, natural variation of four life history traits, primary dormancy, secondary dormancy, seedling growth rate and stem leaf number at flowering timing, followed a latitudinal gradient indicating that these traits are involved in local adaptation. Climatic parameters related to temperature and precipitation influenced different life history traits across the distribution range of the species or depending on the functionality of FRIGIDA. The genetic basis of life history variation in Arabidopsis thaliana was studied at fine and large scales, first focusing specifically on the nucleotidic diversity of a dormancy QTL (DOG1), second at the genomic level using the newly developed genome-wide association mapping method (GWA) developed by M. Nordborg and his team (Atwell et al. 2010). For exon1 of DOG1, non synonymous substitution rate was higher than synonymous substitution rate which is rare. It could be the result of independent selective events in populations isolated by glacial cycles. Only rare alleles based on non synonymous substitutions in exon1 were associated with primary dormancy, alleles in high frequency were significantly associated with flowering timing. This probable false positive association was first detected using the unified mixed model method initially developed by E.S. Buckler (2006), in collaboration with B. Stich and then confirmed with GWA method using a larger sample. The detection of this false positive revealed linkage disequilibrium between DOG1 and flowering time genes which is most likely the result of a simultaneous selection on these genes. FLC was significantly associated with primary dormancy and flowering time indicating pleiotropic effect. These results provided some light concerning genetic control of co-variation between life history traits. DOG1 was not significantly associated with primary dormancy probably because expression level of this gene influences as well phenotypic variation. Finally, GWA method allowed detection of new candidate genes controlling life history traits. SPT gene which controls final leaf size was found as a vegetative growth rate QTL.
Actions (login required)