Asynchronous phenological dynamics in a deciduous plantation and their implications for the seasonal and annual carbon dynamics

Forest phenological dynamics shape the underlying biogeophysical processes and impact the carbon balance from the seasonal to inter-annual time scales. Disentangling the phenological phases of the forest components (e.g., overstory and understory), could provide novel insights on ecosystem response to climate change. This quantitative description is particularly important not only for natural ecosystems but could also assist in the design of restoration and reclamation projects. Here, focusing on a deciduous plantation (black locust, Robinia pseudoacacia L.) in a degraded land of Northern Greece and combining multiyear field observations with detailed ecohydrological modeling, we assessed the ecosystem-level carbon dynamics and its individual components from seasonal to decadal time scales. Site-level long-term (>10 yr) biophysical processes were characterized with eddy covariance measurements together with detailed meteorological and soil data. In addition, ecosystem-level phenological dynamics were quantified with timelapse imagery available at the site and satellite remote sensing. These observations were used to parameterize and validate the ecohydrological model T&C which was then used for numerical experiments. Numerical simulations allowed us to disentangle the contribution of the overstory and understory to the overall carbon dynamics at the site, a separation hard to be done by field measurements alone. The phenological phases of the understory (perennial grass) and the canopy (black locust) were found to be asynchronous, with the former reaching its peak in late winter and the latter in late summer. Ground shading by black locust together with drying of the upper soil layer during the summer months lead to the observed mismatch, with grass activity only in winter and early spring. Yet, the asynchrony in the phenological phases of understory and canopy vegetation results in overall ecosystem-dynamics that are non-negligible over winter, despite the deciduous phenology of black locust. Quantitative description of the interplay between phenological cycles of the forest components enhances our process understanding including their interactions and intra- and inter-annual dynamics. Moreover, for species widely used in forest restoration projects, like the black locust, quantifying such interplays, where the forest is more than the tree, it is important for robust carbon balance estimations.