Résumé : Nitrogen is a macronutrient that greatly influences plant growth and development. The availability of that element shapes the morphology and influences biomass allocation between organs. Nitrate is the most prevalent nitrogen form in agricultural soils. In the model species Arabidopsis thaliana, low nitrate concentration stimulates the growth of the lateral roots, while homogeneous elevated concentration represses it. This work explores the natural variation of root morphology in the model species Arabidopsis thaliana. between Columbia-0 (Col-0) and Den-Hors 2 (Dhs-2). This dissertation will focus on the phenotypical and genetic characterization of the root architecture development and the flowering time. This master thesis aims to characterize in the laboratory conditions, a natural Arabidopsis accession known as De Horse-2 (Dhs-2), collected on the Texel Island in the Netherlands. That accession is lacking lateral branching, in contrast to the reference Columbia-0 (Col-0) accession. Also, Dhs-2 is flowering later than Col-0. First, a thorough root phenotypic characterization was conducted with Col-0 and Dhs-2. The macroscopic root morphology, the density of lateral root primordia and the kinetic of lateral root development were inspected. Seedlings were cultivated in vitro on agar plates with two divergent nitrate concentrations. The Dhs-2 accession had a longer primary root with a larger meristematic zone, and produced any or very few visible lateral roots, compared to Col-0. Microscopic observations also showed that Dhs-2 had a smaller density of lateral root per unit length of primary root hair than Col-0. Furthermore, Dhs-2 showed a slowdown in the initiation and emergence of lateral root primordia, more precisely between stages III and V. These observations support that cell layers in the primary root cortex are exerting biomechanical constraints on the primordium growth. Secondly a genetic characterization of lateral root and flowering time traits was also explored in Dhs-2. Bulk segregrant analysis can help identify the genetic loci underlying phenotypic trains by comparing DNA from two pools of individuals at the extremes of the trait issuance in an interbred population. F2 individuals with extreme lateral root and flowering time traits are pooled and genotyped to find associated genomic regions with the allele frequency differences (AFD) pinpointing relevant loci. A segregating F2 population was generated after crossing Col-0 and Dhs-2. It appeared that the Dhs-2 lateral-root-less phenotype was likely controlled by one recessive locus, and the late flowering by one dominant locus. The Dhs-2 root phenotype could be an adaptation to the sand soil conditions, where the primary root grows downwards to the lower strata without elongating the lateral roots, in order to absorb water and other mobile soil resources.