Phenological responses of Alpine snowbed communities to advancing snowmelt

In recent decades climate warming accelerated substantially and consequences were especially pronounced in alpine ecosystems. Among others, alpine grasslands characterized by long lasting snow cover e.g., snowsbeds, will be especially threatened, as they are subjected to substantial shifts in phenological timing and development due to earlier snowmelt. To assess the impact of advancing snowmelt on the phenology of alpine snowbed communities, we have monitored two study sites in the Italian Alps for 3-years. While both study sites are characterized by the same vegetation community (Salicetum herbaceae), altitude, and topography, they differed in the amount of winter precipitation and as a consequence in the timing of snowmelt. We monitored the phenological development of five shared plant species between both study sites by deriving the “green chromatic coordinate” (gcc) from images of the Phenocams, installed at both study sites and in-situ phenological assessment, following a standardized protocol. We then compared results between the early and the late snowmelt site and related the results of the community level to those of the species level. The gcc identified the start of growing season (SOS), peak of growing season (POS) and end of growing season (EOS) at both study sites but failed to grasp the interspecific and inter-site differences on the species level. We found that a three-week earlier snowmelt at the early snowmelt site (DOY 176 vs. 197) did not extend (25 vs. 26 days) but only advance the POS by 20 days and resulted in a approx. 10% lower gcc-values at the early snowmelt site. At the same time, the in-situ monitoring highlighted species-specific responses of the study species. Within the study period we could identify changes in the speed of the phenological development for each study species in at least one year, with the interannual differences being greater than the differences between the study sites. Nevertheless, when compared to the late snowmelt site, the phenological development at the early snowmelt site was slower for some species (e.g., Poa alpina -24.2%) and faster for others (e.g., Euphrasia minima +70.4%). This indicates the ability of these species to modulate their phenology in response to shifting snowmelt dates. Relating the results of both approaches to each other led to mixed results, as the species-based approach only partially supported the results of the community-based approach and vice versa, with R²-values ranging from 0.1 (Euphrasia minima) up to 0.61 (Veronica alpina). These results highlight the importance of multi-level approaches when trying to identify the effects of climate change on alpine vegetation communities, as species specific results in most cases don’t represent the results on a community level and results on community level mostly fail to address the single species’ responses in these complex and heterogeneous habitats.