Earth system models (ESMs) are process-based models that integrate across disciplines to simulate biogeophysical and biogeochemical feedbacks. While ESMs are primarily used for global scale simulations, recent work has applied them to ecologically relevant questions offering transferrable insight into abiotic and biotic responses to climate change. Here, we used single-point Community Land Model (CLM, the terrestrial component of the Community Earth System Model) simulations paired with the representative hillslope hydrology model to characterize a topographically heterogeneous alpine environment at the Niwot Ridge Long Term Ecological Research (LTER) site in Colorado, USA. We simulated three distinct alpine tundra vegetation communities along a snow accumulation and resulting soil moisture/temperature gradient using meteorological and eddy covariance data from 2008-2020 as well as plant functional trait datasets. Based on these simulations, we assessed 1) whether CLM paired with hillslope hydrology can reproduce observed water, carbon, and energy fluxes at Niwot Ridge; 2) how much the inclusion of Niwot Ridge-specific plant functional traits improves the representation of plant communities compared to default CLM Arctic C3 grass traits; and 3) the impacts of future warming scenarios on alpine tundra communities.
Our simulations showed changes in snow accumulation, moisture, and vegetation productivity across the alpine tundra hillslope gradient at Niwot Ridge that were consistent with observations and corresponded to the effects of fine-scale topographic variation. We found that dry meadow productivity was principally limited by soil temperature and moisture stress, whereas moist meadow productivity was limited by the length and timing of the snow-free period. Wet meadow vegetation received water subsidies from adjacent moist meadow areas that allowed soils to retain greater moisture levels during the growing season. Incorporating community-specific foliar traits such as specific leaf area and C:N ratio significantly improved model productivity estimates compared to the default CLM Arctic C3 grass parameterization. In both dry and moist meadows, experimental air temperature increases of 2 and 4 degrees C resulted in increased aboveground productivity compared to the control simulation; warming-induced productivity was stimulated to a lesser degree in the dry meadow, which was likely constrained by water limitation. Our findings demonstrate that CLM can be manipulated to represent distinct vegetation communities across environmental gradients, for the purpose of informing future observational and experimental work. These tools can be leveraged to forecast climate change impacts on ecosystem function at landscape scales.