IMPROVED REPRESENTATION OF PHOTOSYNTHETIC TEMPERATURE RESPONSE IN TERRESTRIAL BIOSPHERE MODELS

Abstract

Photosynthesis, a key process of the terrestrial carbon cycle, is crucial for predicting ecosystem responses to climate change. Terrestrial biosphere models (TBMs) commonly use the biochemical model of C3 photosynthesis, which relies on two primary parameters: Vcmax (maximum Rubisco activity) and Jmax (maximum electron transport rate). The temperature responses of these parameters are often modeled using the peaked Arrhenius function, relying on K25 (process rate at 25°C), Ea (activation energy), Hd (deactivation energy), and ΔS (entropy). To avoid overfitting, most studies assume a fixed Hd; however, the impact of this assumption remains untested. In this study, using a global scale dataset of plant photosynthetic temperature response, we compared the temperature response of Vcmax and Jmax by fitting with and without assuming a fixed Hd. For Vcmax, found a strong relationship between the fixed and variable Hd models for K25 (R²=0.85, p<0.001), suggesting that assuming a fixed Hd moderately captures the observed variability in K25. However, the relationships were weak for Ea (R²=0.04, p=0.22) and ΔS (R²=0.16, p=0.045), indicating limited suitability of the fixed Hd assumption for these parameters. For Jmax, K25 similarly showed a moderate correlation between models (R²=0.66, p<0.001), whereas Ea (R²= 0.14, p=0.015) and ΔS (R²=0.04, p=0.19) displayed weak to negligible relationships, highlighting limited applicability of the fixed Hd approach. These findings suggest that while a fixed Hd assumption may be reasonable for K25 in both Vcmax and Jmax models, it does not adequately capture the temperature dependence of Ea and ΔS, which could lead to inaccuracies in TBMs. Further investigation into the variable Hd model’s benefits for temperature response accuracy is warranted.