Constraining Light-Driven Plasticity in Leaf Traits With Observations Improves the Prediction of Tropical Forest Demography, Structure, and Biomass Dynamics

Predicting tropical tree demography is a key challenge in understanding the future dynamics of tropical forests. Although demographic processes are known to be regulated by leaf trait diversity, only the effect of inter-specific trait variation has been evaluated, and it remains unclear as to what degree the intra-specific trait plasticity across light gradients (hereafter light plasticity) regulates tree demography, and how this will further shape long-term community and ecosystem dynamics. By combining in situ trait measurements and forest census data with a terrestrial biosphere model, we evaluated the impact of observation-constrained light plasticity on demography, forest structure, and biomass dynamics in a Panamanian tropical moist forest. Modeled leaf physiological traits vary across and within plant functional types (PFT), which represent the inter-specific trait variation and the intra-specific light plasticity, respectively. The simulation using three non-plastic PFTs underestimated 20-year average understory growth rates by 41%, leading to a biased forest size structure and leaf area profile, and a 44% underestimate in long-term biomass. The simulation using three plastic PFTs generated accurate understory growth rates, resulting in a realistic forest structure and a smaller biomass underestimate of 15%. Expanding simulated trait diversity using 18 nonplastic PFTs similarly improved the prediction of demography and biomass. However, only the plasticity-enabled model predicted realistic long-term PFT composition and within-canopy trait profiles. Our results highlight the distinct role of light plasticity in regulating forest dynamics that cannot be replaced by inter-specific trait diversity. Accurately representing light plasticity is thus crucial for trait-based prediction of tropical forest dynamics.