Improved global estimation of seasonal variations in C3 photosynthetic capacity based on eco-evolutionary optimality hypotheses and remote sensing

AbstractThe maximum carboxylation rate of plant leaves (Vcmax) at 25°C (Vcmax25) is a fundamental parameter in terrestrial biosphere models (TBMs) to estimate carbon assimilation of C3 biomes. It has been reported that ignoring the seasonal variations inVcmax25induces considerable uncertainties in TBMs. Recently, a model was developed to estimateVcmax25of C3 biomes mechanistically from climate data based on eco-evolutionary optimality hypotheses, which hypothesized that plants acclimate to the environment to achieve maximum carbon assimilation with minimum related costs. However, uncertainties of this optimality-based model (EEO model) have been found to correlate to leaf nitrogen content, partly due to the lack of parameterization on how the acclimation ofVcmax25is constrained by photosynthetic nitrogen other than that in RuBisCO. This constraint could be parametrized by remote sensing methods globally. In this study, we developed remote sensing methods to estimate leaf absorptance of radiation based on MODIS LCC (leaf chlorophyll content) data and the ratio of the maximum electron transport rate of plant leaves (Jmax) toVcmaxat 25°C (rJV25) based on TROPOMI SIF (solar-induced chlorophyll fluorescence) data (RS-rJV25). These two parameters contain photosynthetic nitrogen information related to light harvesting, electron transport, and carboxylation, and they were then incorporated into the EEO model to constrain howVcmax25acclimates to the environment. The simulatedVcmax25constrained by MODIS LCC and RS-rJV25agreed well with seasonal variations in field-measuredVcmax25at 18 sites (R2 = 0.76, RMSE = 13.40 µmol·m− 2·s− 1), showing better accuracy than the simulation without incorporating leaf absorptance andrJV25(R2 = 0.63, RMSE = 31.59 µmol·m− 2·s− 1). Our results indicated that variations in leaf absorptance andrJV25constrained the acclimation ofVcmax25to the environment and contributed to the variation inVcmax25that cannot be fully captured by environmental factors alone in the EEO model. The remote-sensing-based leaf absorptance andrJV25captured the sensitivity of these two parameters to environmental conditions on the global scale. The influence of leaf absorptance onVcmax25was primarily affected by the irradiance level, whilerJV25was determined by the growing season mean temperature. The simulatedVcmax25had large spatiotemporal variations on the global scale, and the environment drove the variation pattern more greatly than the biome distribution. With reasonably accurate seasonal variations inVcmax25, this study can help improve the global carbon cycle and leaf trait modelling.