||Soils, plants, and microbial communities respond to global change perturbations through coupled, nonlinear interactions. Dynamic ecological responses complicate projecting how global change disturbances will influence ecosystem processes, such as carbon (C) storage. We developed an ecosystem-scale model (Stoichiometrically Coupled, Acclimating Microbe-Plant-Soil model, SCAMPS) that simulates the dynamic feedbacks between aboveground and belowground communities that affect their shared soil environment. The belowground component of the model includes three classes of soil organic matter (SOM), three microbially synthesized extracellular enzyme classes specific to these SOM pools, and a microbial biomass pool with a variable C-to-N ratio (C:N). The plant biomass, which contributes to the SOM pools, flexibly allocates growth toward wood, root, and leaf biomass, based on nitrogen (N) uptake and shoot-to-root ratio. Unlike traditional ecosystem models, the microbial community can acclimate to changing soil resources by shifting its C:N between a lower C:N, faster turnover (bacteria-like) community, and a higher C:N, slower turnover (fungal-like) community. This stoichiometric flexibility allows for the microbial C and N use efficiency to vary, feeding back into system decomposition and productivity dynamics. These feedbacks regulate changes in extracellular enzyme synthesis, soil pool turnover rates, plant growth, and ecosystem C storage. We used SCAMPS to test the interactive effects of winter, summer, and year-round soil warming, in combination with microbial acclimation ability, on decomposition dynamics and plant growth in a tundra system. Over 50-year simulations, both the seasonality of warming and the ability of the microbial community to acclimate had strong effects on ecosystem C dynamics. Across all scenarios, warming increased plant biomass (and therefore litter inputs to the SOM), while the ability of the microbial community to acclimate increased soil C loss. Winter warming drove the largest ecosystem C losses when the microbial community could acclimate, and the largest ecosystem C gains when it could not acclimate. Similar to empirical studies of tundra warming, modeled summer warming had relatively negligible effects on soil C loss, regardless of acclimation ability. In contrast, winter and year-round warming drove marked soil C loss when decomposers could acclimate, despite also increasing plant biomass. These results suggest that incorporating dynamically interacting microbial and plant communities into ecosystem models might increase the ability to link ongoing global change field observations with macro-scale projections of ecosystem biogeochemical cycling in systems under change.