The functional trait framework provides a powerful corpus of integrated concepts and theories to assess how environmental factors influence ecosystem functioning through community assembly. While common in plant ecology, this approach is under-used in microbial ecology. After an introduction of this framework in the context of microbial ecology and enzymology, we propose an approach 1) to elucidate new links between soil microbial community composition and microbial traits; and 2) to disentangle mechanisms underlying “total” potential enzyme activity in soil (sum of 7 hydrolase potential activities). We address these objectives using a terrestrial grassland ecosystem model experiment with intact soil monoliths from three European countries (Switzerland, France and Portugal) and two management types (Conventional-intensive and Ecological-intensive), subjected to 4 rain regimes (Dry, Wet, Intermittent and Normal) under controlled conditions in a common climate chamber. We found tight associations between proxies of microbial ecoenzymatic community-weighted mean traits (enzymatic stoichiometry and biomass-specific activity) and community composition, bringing new information on resource acquisition strategy associated with fungi, Gram positive and Gram negative bacteria. We demonstrate that microbial biomass explained most of the total enzyme activity before altered rain regimes, whereas adjustments in biomass-specific activity (enzyme activity per unit of microbial biomass) explained most variation under altered rain regime scenarios. Furthermore, structural equation models revealed that the variation of community composition was the main driver of the variation in biomass-specific enzyme activity prior to rain perturbation, whereas physiological acclimation or evolutionary adaptation became an important driver only under altered rain regimes. This study presents a promising trait-based approach to investigate soil microbial community response to environmental changes and potential consequences for ecosystem functioning. We argue that the functional trait framework should be further implemented in microbial ecology to guide experimental and analytical design.