Gross primary production (GPP) and red solar induced fluorescence (SIF) respond differently to light and seasonal environmental conditions in a subalpine conifer forest

The phenology of montane conifer forests is likely to shift in response to climate change and altered seasonal dynamics of light, temperature, and moisture. Solar-induced fluorescence (SIF) is expected to provide substantial improvement for mapping temporal changes in evergreen gross primary production (GPP) over greenness-based remote sensing indices. The utility of SIF to monitor seasonal changes in the phenology of conifer photosynthesis depends on the degree to which GPP and SIF respond in synchrony to key environmental drivers. However, to what extent SIF and GPP become decoupled by responding differently to the combined effects of light and other environmental conditions remains unknown. The goal of this study was to characterize the responses of GPP and SIF_red to a suite of environmental drivers at the half-hour time scale and determine how these relationships change across seasons. We analyzed one year of tower-based SIF_red and eddy covariance-derived GPP data from a conifer forest at Niwot Ridge, Colorado. We compared the light responses of GPP and SIF_red across the year, finding that SIF_red increased in response to light earlier in the year than did GPP. The light response of GPP had a positive temperature dependence in spring, and this dependency reversed in summer due to increased evaporative demand, while the light response of SIF_red was less temperature dependent. Using artificial neural network ensemble analysis, we found that from spring to summer, SIF_red did not exhibit a parallel response to the seasonally dynamic temperature and moisture controls on GPP. In summer SIF_red was not correlated with canopy conductance, suggesting that SIF is less sensitive to stomatal control than GPP. Our results suggest that, in conifers, photosystems begin to activate in spring prior to when water becomes available for photosynthesis, presenting a challenge for the use of SIF as a phenological indicator in conifer forests.