Inclusion of a cover crop is an interesting way to diversify cropping systems and support ecosystem functions and services impaired by intensified monocultures. However, pure covers are hard to obtain without substantial management, and mixed covers—wherein some spontaneous species coexist with sown cover crops—seem to be a good alternative. Greater insight into factors influencing the composition and density of the spontaneous plant community is needed to achieve relatively good control of the performances of this mixed cover. Here we addressed two questions: 1) How does the functional structure of communities with a highly competitive sown cover crop and/or high initial species diversity differ from the functional structure of communities with the opposite characteristics? 2) Can aboveground species traits explain interactions between cover crops and spontaneous species present in the same community? In a fallowed field in Guadeloupe (West French Indies), we studied communities composed of one sown cover crop species and spontaneous species growing simultaneously in the plot. The study was conducted along two unrelated experimentally-devised gradients resulting from cropping practices: a competitivity gradient of a sown cover crop species and an initial species diversity gradient resulting from an initial soil perturbation induced by the false seedbed technique. These factors (soil tillage modality before cover crop sowing and cover crop species identity) were combined in a split-plot design. Seven and twelve months after cover crop sowing, community species densities and biomasses were assessed. Leaf trait values of the 12 most present species were measured at flowering to describe the functional structure of the community using the community weighted trait means and variances, and a functional diversity index. Our results showed high functional divergence in communities with highly competitive cover crops, which suggests differentiation of functional niches betw