Because of the vegetation cover and anthropogenic disturbances, land-use management strongly influences soil heterotrophic decomposers. Yet, little is known about whether contrasting microbial communities originating from different ecosystems are functionally similar, and only a few studies have disentangled the direct and indirect effects of resource quality on both microbial communities and carbon mineralization rates. To assess the relative importance of aboveground litter quality and belowground microbial communities on litter decomposition, we conducted a reciprocal transplant experiment under controlled conditions using four litters (Triticum aestivum, Fagus sylvatica, Festuca arundinacea and Robinia pseudoacacia) and four soils (culture, plantation, grassland and forest) originating from a land-use gradient. We followed the kinetics of carbon mineralization over 21 dates spanning a 202-day period to assess the variability of responses generated by the plant–soil interactions. Furthermore, at four time points (at 0, 27, 97 and 202 days), the mass loss rates for the main sugars within the cell wall, the microbial biomass (fumigation-extraction), the microbial community structure via phospholipid fatty acid (PLFA), and the activities of four carbon-related hydrolytic enzymes were investigated to assess the functional significance of microbial communities. Our results demonstrated that the importance of soil types and heterotrophic decomposers on carbon mineralization rates was minor (1.2% of the variance explained) compared with the predominant role of litter quality. The structure of the microbial communities responded strongly to both long-term land-use changes and short-term litter additions; specifically, (i) higher proportions of fungi were observed in natural ecosystems compared with agro-systems, and (ii) an opportunistic subset of the bacterial community was stimulated after litter additions. Even if the land-use management and litter quality can shape the microbial community structure in a foreseeable way, we found an important degree of plasticity in the responses of contrasting decomposer communities. In particular, the enzymatic efficiency (defined as the amount of enzyme produced by unit of carbon mineralized) differed among litters but not among soil types, suggesting that the threshold between carbon allocation to growth and acquisition depended more on the ‘resource-use strategies’ of the soil microorganisms than on the community structure. The recalcitrant litters stimulated ‘efficient’ communities characterized by low enzymatic activities, microbial biomass and respiration rates at the opposite of labile litters that stimulated ‘wasteful’ communities characterized by higher activities and metabolic quotient (defined as the amount of carbon respired by unit of biomass). In addition to the direct effects of litter quality, the path analysis reinforced our conclusion that the functional traits of microorganisms via their enzymatic activities are more relevant than their identity for predicting carbon mineralization. Thus, although multiple and coordinated responses of soil microbes can improve our understanding of carbon fluxes, shifts in the plant community composition caused by land-use conversion will have a stronger impact on predictions of carbon mineralization than short-term changes in the microbial community composition.