Land abandonment across Europe is a widespread phenomenon that promotes secondary succession, reshaping ecosystem structure and functioning. Soil microorganisms play a central role in these processes through their contribution to soil organic matter (SOM) decomposition and nutrient cycling. However, understanding the relationship between microbial diversity and carbon sequestration requires information that extends beyond taxonomic composition alone. We investigated changes in soil carbon dynamics following agricultural land abandonment across a latitudinal gradient in Italy, relating them to microbial taxonomic diversity, activity and inferred functional potential. We specifically asked whether moving beyond taxonomic diversity metrics towards functionally annotated indices could provide new ecological insights into microbial responses to forest recovery and their links with soil carbon dynamics, assessed through CHN analysis, bulk density and basal respiration measurements.Microbial diversity was characterized using eDNA metabarcoding targeting 16S and ITS markers, while extracellular enzymatic activities were used as proxies of microbial functioning. Bacterial functional annotations included biotic interactions, trophic strategies, nitrogen-cycle functions and metabolic traits involved in carbon transformation, including chemoheterotrophy, chitinolysis, ligninolysis, methanogenesis and degradation of aromatic compounds and hydrocarbons. Fungal taxa were annotated according to trophic modes, functional guilds and growth forms, including ectomycorrhizal, arbuscular mycorrhizal, endophytic and saprotrophic fungi. Across the Italian gradient, bacterial function related to carbon and nitrogen cycling increased with afforestation (PERMANOVA, P < 0.001), whereas trophic strategies were driven by microclimatic conditions.Variation in enzymatic activity patterns across northern and southern sites underscores the combined effects of land-use change and climate in shaping microbial functioning. Overall, functional diversity captured ecologically meaningful microbial responses to forest succession and carbon dynamics more effectively than taxonomy alone. Our approach provides a promising framework for integrating eDNA- functional information into SOM models, helping to unravel the microbial "black box" and strengthen the links between microbial community composition, functional potential and soil carbon turnover.