Clonal architecture is involved in performance of clonal fragments, as it determines spatial distribution of ramets. It is expected to rely on the species-specific expression of several architectural traits (structural blue-print). However, in contrasting environments, realized clonal architectures may differ, due to phenotypic plasticity. In this paper, we compared clonal architectures between two rhizomatous ecologically close Cyperaceae (Carex divisa and Eleocharis palustris) in non-defoliated and defoliated conditions. Two questions were addressed. (1) How much do the structural blue-print and resulting colonization and occupation of space differ between both species? (2) Does the structural blue-print constrain plastic responses of clonal architecture to defoliation? Traits related to performance, spatial pattern, architecture and biomass allocation of clonal fragments were monitored through an original non-destructive mapping method. In nondefoliated conditions, both species showed similar biomass but contrasting architectures and patterns of biomass allocation to rhizomes that resulted in different spatial patterns. The rhizome network of C. divisa, which consisted in only two primary rhizomes but several branches, was involved in resource storage rather than in spatial colonization. Conversely, E. palustris produced on average six primary rhizomes that grew in the whole horizontal plane, maximizing both occupation and colonization of space. These differences in structural constraints coupled with allometric relationships, resulted in differential responses to defoliation. In C. divisa, the costs associated to defoliation caused a decrease in branching, limiting the area occupied and number of ramets produced by clonal fragments, but increasing ramet density. Conversely, the weakly branched rhizome network of E. palustris was not affected by defoliation. Both spatial strategies (consolidation vs. colonization) are likely to provide ecological advantages allowing their coexistence in grazed meadows.