Résumé : The urge for new sustainable weed control systems in Europe is increasing as modern herbicides have proven their limit. Alternatives using allelopathic plants as cover crops to reduce weed growth, are perceived as a compelling approach. Rye is an allelopathic crop which has been shown to produce phytotoxic defensive secondary metabolites called allelochemicals such as benzoxazinoids (BXDs). In that view, the belowground interactions between, rye (Secale cereale L.) and pigweed (Amaranthus retroflexus L.), which is a widespread weed, have been studied in this master thesis. A focus has been made on the effect of two potential belowground signaling molecules (loliolide and jasmonic acid)inducing the production of BXDs in rye. Three different aspects have been developed.Firstly, the influence of substrate characteristics on plant growth, root architecture along with BXDs composition and concentration has been analyzed. Both rye and pigweed have been cultivated, alone or in co-culture, in two different substrates: microbeads of glass and a mixture of clay and attapulgite. It has been demonstrated that both plants had greater growth in the clay and attapulgite mixture, suspectedto be caused by substrate differences in particle size, water retention and/or pore space between particles. However, fewer BXDs were detected in that substrate, presumably due to the sorption capacity of clay compared to glass microbeads. These findings further support the critical importance of the physiochemical properties of soils when investigating plant morphology and plant chemistry such as allelopathy.Secondly, the physical and chemical interactions between rye and redroot pigweed along with the effect of potential signaling molecules (loliolide and jasmonic acid) on rye root architecture and allelochemicals (e.g. BXDs) production have been studied. On the one hand, rye treated with pigweed root exudates showed greater root growth for almost all root architecture parameters and lower BXDs concentrations compared to rye grown in co-culture with pigweed. It could be assumed that the physical competition between rye and pigweed when grown in co-culture might have overcome the potential effect of chemicals from root exudates. Moreover, low concentration of pigweed root exudate might have reduced the entire chemical outcome, leading to a possible hormesis effect. On the other hand, rye treated with low dose corresponding to 0.5 nM of loliolide and jasmonic acid, which has never been pursued before, showed lower root growth for all root architecture parameters as well as higher BXDs concentrations for three compounds (DHBOA-Glc, HBOA-Glc and DIMBOA). The increase of BXDs concentration motivates the hypothesis that loliolide and jasmonic acid could act as potential belowground signaling molecules inducing the production of defensive metabolites, while the root growth reduction highlights their inhibition effect.To close this master thesis, the detection and quantification of loliolide in rye root and shoot by HPLCUV have been carried out. The method of loliolide extraction has been optimized by using fresh plants biomass and the extraction solution made of acetonitrile, water and formic acid. Loliolide could be detected in rye roots. Those results should nonetheless be further confirmed by repeating the experiment with more replicate.