B106-05
Variability in the start and end of the photosynthetically active and dormant periods in seasonally snow-covered forests around the world

Tuesday, 15 December 2020: 20:46
Virtual
Kenneth Smith, University of Utah, Salt Lake City, UT, United States, Michael Bahn, University of Innsbruck, Institute of Ecology, Innsbruck, Austria and David R Bowling, University of Utah, School of Biological Sciences, Salt Lake City, UT, United States
Abstract:
Climate warming is altering the timing and magnitude of net carbon (C) uptake by forest ecosystems via phenological shifts that extend the length of the photosynthetically active season. For deciduous broadleaf forests (DBF), the seasonal dynamics of C exchange are tightly linked to indices of greenness, where the start of the growing season (SOS) is marked by the presence of new foliage; however, for evergreen needleleaf forests (ENF) that retain their foliage year-round, the transition from winter dormancy to increased metabolic activity is much more difficult to discern using remote sensing techniques alone. In this study, we analyzed eddy covariance data from 27 FLUXNET2015 and AmeriFlux sites (spanning 366 site-years in seasonally snow-covered ecosystems) to develop a physiology-based method to characterize the start/end of the photosynthetically active season. Here, we constructed time series of light-saturated photosynthetic capacity (GPPsat) derived from gap-filled estimates of gross primary productivity (GPP); we then fit a smoothing spline to the time series data to examine seasonal inflection points that correspond with changes in phenology. From these time series, we identified the dates at which GPPsat reached 25% of its maximum value for the year (SOS25), a metric that coincides with the start of the growing season.

We found that SOS25 in DBF and MF sites occurred 25-30 days later on average compared to ENF sites (SOS25, ENF: 103; MF: 128; DBF: 133). MAT at the ENF sites was 0.5-2.1°C cooler compared with MF/DBF (MAT, ENF: 2.6°; MF: 3.2°; DBF: 5.1°C); likewise, ENF sites had fewer accumulated degree days (ADD, >5°C) by the date at which SOS25 occurred (ADD, ENF: 5; MF: 22; DBF: 35). After binning our site-level data into discrete MAT categories, we found that SOS25 occurred much earlier in the calendar year at higher temperature regimes (SOS25[-4.5, -1.5°C]: 129; SOS25[7.5, 10.5°C]: 78); we also found that regardless of MAT regime, the linear relationship between SOS25 and ADD within a given year was consistent, exhibiting a strong upward trend. Together these results highlight important differences among plant functional types, climate, and the dates at which these forested ecosystems transition from C source to C sink.