Can the eco-evolutionary optimality concept predict steady-state vegetation? An evaluation and comparison of four models

Abstract

The Eco-Evolutionary Optimality (EEO) theory posits that vegetation adopts specific growth strategies, co-evolving with the environment to achieve a steady state. The EEO models, by capturing the mechanistic interactions between vegetation and the environment while maintaining simplicity, hold promise in simulating vegetation at steady states. In this study, four EEO models (the Eagleson model, the Yang–Medlyn model, the VOM, and the P model) were selected for evaluation and comparison of their performance across 44 undisturbed flux sites globally. Overall, all four models effectively reproduced key variables such as fraction of vegetation cover, evapotranspiration, and gross primary production across most sites, with the Yang–Medlyn and P models demonstrating superior performance. Variability in model performance across different plant functional types was observed, with poorer performance generally noted at shrub sites, while forest and savanna sites exhibited better performance. Analysis across precipitation and temperature gradients revealed better model performance under wetter or warmer conditions. Furthermore, variations in model sensitivity to climate factors were evident, with outputs generally exhibiting higher sensitivity to precipitation and atmospheric CO₂ concentration compared to temperature and vapor pressure deficit. Sensitivity tended to be higher in arid regions compared to relatively humid regions. These findings underscore the capability of EEO models to simulate steady-state vegetation with minimal or no parameter calibration, demonstrating satisfactory performance across diverse environmental conditions.