Changes in labile carbon (LC) pools and microbial communities are the primary factors controlling soil heterotrophic respiration (R_h ) in warming experiments. Warming is expected to initially increase R_h but studies show this increase may not be continuous or sustained. Specifically, LC and soil microbiome have been shown to contribute to the effect of extended warming on R_h . However, their relative contribution is unclear and this gap in knowledge causes considerable uncertainty in the prediction of carbon cycle feedbacks to climate change. In this study, we used a two-step incubation approach to reveal the relative contribution of LC limitation and soil microbial community responses in attenuating the effect that extended warming has on R_h . Soil samples from three Tibetan ecosystems-an alpine meadow (AM), alpine steppe (AS), and desert steppe (DS)-were exposed to a temperature gradient of 5-25°C. After an initial incubation period, soils were processed in one of two methods: (a) soils were sterilized then inoculated with parent soil microbes to assess the LC limitation effects, while controlling for microbial community responses; or (b) soil microbes from the incubations were used to inoculate sterilized parent soils to assess the microbial community effects, while controlling for LC limitation. We found both LC limitation and microbial community responses led to significant declines in R_h by 37% and 30%, respectively, but their relative contributions were ecosystem specific. LC limitation alone caused a greater R_h decrease for DS soils than AMs or ASs. Our study demonstrates that soil carbon loss due to R_h in Tibetan alpine soils-especially in copiotrophic soils-will be weakened by microbial community responses under short-term warming.