"Significant uncertainties on the response of ecosystems to climate change remain to be addressed to advance the progress of the Intergovernmental Panel on Climate Change. We propose to investigate several of these uncertainties by improving regional-scale dynamic vegetation modeling for montane systems vulnerable to climate and land-use change. Our approach will modify a state-of-the-art dynamic global vegetation model (LPJ-DGVM) to include a new disturbance module for grazing, a key driver of alpine treeline dynamics. High-spatial resolution data for soils and climate (20th century and climate projections), essential for regional modeling, will be available from the Host institute and international collaborators based in the United States and China. The direct effects of grazing on seedling mortality and on light and water competition with grasses will be incorporated to the framework of LPJ. Model output, including net ecosystem exchange and species biogeography, will be evaluated for the Swiss Alps against databases from Swiss FluxNet, phenology observations from MeteoSwiss, and net primary productivity modeled from tree rings. Physiological parameters for plant functional types found in mountain habitat will be evaluated and adjusted following this assessment. The modified dynamic vegetation model, LPJ ‘Grazing’, will be applied with various climate scenarios to a biodiversity hotspot in the Hengduan Mountains in the Yunnan Province of China where little field-data exist and where the vulnerability to climate and land-use change is highly uncertain. The proposed research combines expertise and datasets of researchers from three continents to reduce uncertainties related to regional-scale climate attribution, impacts, and vulnerability. Our modifications to the LPJ-DGVM will have a significant impact globally for mountain biogeography and biogeochemical modeling because the model provides the basis for vegetation dynamics in many coupled global climate models."