Physiological and environmental drivers of temporal variability in solar-induced fluorescence

Background/Question/Methods

The relationship between solar-induced fluorescence (SIF) and gross primary productivity (GPP) is linear at the satellite scale, but becomes nonlinear at finer spatiotemporal resolutions. This nonlinearity is driven in part by leaf-level light energy partitioning, the effects of which become less apparent when averaging over many observations, many angles, and areas of vegetation with differential illumination. The resulting limitations to linking satellite observations to plant physiology on the ground indicate that a greater knowledge of energy partitioning dynamics is needed to best use satellite SIF observations to constrain models of GPP. We used the Cameron SIF-measuring system to persistently monitor vegetated targets in an urban area during the 2020 growing season (August 10 to October 31, 2020), gathering data on SIF and the photochemical reflectance index (PRI), which can be used to track changes in the xanthophyll cycle pigments that regulate thermal energy dissipation at the leaf level. Throughout the study period, we directly quantified the pigment composition of small samples of leaf material, measured sapflow from tree stems and branches near Cameron fields of view, and monitored leaf-level gas exchange and pulse-amplitude modulated (PAM) fluorescence using a LICOR LI-6800 photosynthesis system.

Results/Conclusions

To investigate energy partitioning dynamics at fine spatial and temporal scales, our analyses focused on SIF-PRI relationships in addition to SIF-GPP relationships. The slope of the SIF-PRI relationship showed a phenological trend, becoming more strongly negative as the growing season progressed, and corresponded with a decline in leaf chlorophyll concentration. This trend persisted even after controlling for maximum photosynthetically active radiation (PAR) and daylength, indicating that this was not solely an effect of illumination changing with time of year. In addition to these seasonal-scale trends, we observed diurnal-scale variability in steady-state PAM fluorescence and photochemical yield, which were consistent with conceptual models of leaf-level energy partitioning dynamics across illumination levels. Finally, we identified several high temporal frequency effects that explain additional variability in the SIF-PRI relationship, including changes in cloud cover and sky condition, and wind-mediated changes in leaf orientation and reflectance. Our data make a strong case for persistent monitoring of SIF and PRI to track changes in energy partitioning under real-world conditions and inform remote sensing-driven models of primary productivity.