ABSTRACT Climate change significantly influences natural and semi‐natural ecosystems at global scale. This impact is further amplified in alpine environments, owing to a set of climate feedback mechanisms (i.e., elevation‐dependent warming). Due to the short snow‐free season and the consequent adaptation of alpine plants to fast growth, high‐elevation grasslands mainly act as carbon sinks during the vegetative season, with primary productivity usually exceeding ecosystem respiration. We combined timeseries of field data (CO 2 fluxes and environmental variables), with weather, vegetation and remotely‐sensed land‐cover data from the Nivolet area (Gran Paradiso National Park, western Italian Alps), to identify the factors controlling diurnal ecosystem respiration and primary productivity in such environments. We found that growing‐degree‐days‐since‐snowmelt (GDD 0 —a measure of cumulative heat available for plant growth since the snow has melted) exerted a strong control on both respiration and primary productivity. The effect of GDD 0 on productivity is both indirect, mediated by the seasonal trend in greening (proxied by vegetation indices), and direct, possibly reflecting the importance of cumulated heat on vegetation height. While greening had limited effect on respiration, GDD 0 had a strong effect, supporting the view that ecosystem respiration was mainly temperature‐related and microbial‐driven. After accounting for the effect of greening, we found that scarce winter cumulative precipitation had positive impacts on productivity and respiration, suggesting a possible role of reduced snowpack on the cycling of organic matter. Other variables (i.e., solar irradiance, air temperature, soil moisture, vegetation composition and topography) either supported our expectations or showed no significant effect. Our results contribute to identify the causal relationships between climate and diurnal carbon cycling, improving our understanding of how climate change affects alpine ecosystems in these high‐elevation environments.
Marta et al. (Mon,) studied this question.