Characterization of the functional structure of weed community in agroecosystems

Despite numerous studies, weed community assembly mechanisms still remain unidentified. However, a better understanding of these mechanisms is crucial to design management strategies promoting weed biodiversity while maintaining agricultural production. The aim of this PhD was to understand the mechanisms underlying the functional structure of weeds in agroecosystems.Three research axes were conducted. The first axis enabled to propose a general analytical framework of community assembly in the context of a statistical approach called null model. The second axis, based on empirical analyses, revealed (1) the predominant role of stochastic processes in weed community assembly between crop types and (2) the coexistence of two, shade-tolerance and shade-avoidance, ecological strategies, thus explaining the higher functional diversity observed in crop edge, a field element adjacent to the field. Finally, on the third axis, based on experimentation, we have shown that the competition between the crop and the weed community promotes a unique shade-avoidance strategy resulting on a functional mimicry of the crop by the most successful weed plants. Thus, both stochastic and deterministic processes influence weed community assembly, besides at different spatial and temporal scales.