Most process-based ecosystem models use fixed root depths within 2 m, regardless rooting depths of trees and shrubs in semiarid areas are often greater than 4 or 5 m. In this study we developed a model with vertical growth of root and canopy adapting to environmental variations. The vertical growth was controlled by a maximum leaf density parameter for canopy, and a maximum fine root density for root. A modified pipe model was used to simulate vascular growth. The model also used acclimated photosynthesis and respiration so that carboxylase activity strongly depended on available nitrogen, light, and temperature. Respiration was modeled as 4 components of nitrogen turnover, synthesis of structural and functional materials, root uptake of nutrients, and phloem loading. Hydraulic redistribution was implemented in the model to facilitate the vertical water exchange and the downward growth of roots. The model was applied to three ecosystems in China: the south subtropical broadleaved forest at Heshan, the warm temperate deciduous broadleaved forest at Donglingshan, and the semiarid shrub ecosystem at Maowusu, to simulate the carbon and nitrogen dynamics over the past and future 50 years, in comparison with the corresponding model without adaptive mechanism.

Results/Conclusions

The results showed that the carbon and nitrogen fluxes simulated by the model with adaptive mechanism had less temporal variation than those by the model without adaptive mechanism, indicating that the model with adaptation is more resistant to environmental change. Deep root is especially important for the semiarid shrub ecosystem which enabled them to survive several droughts. The results have important implications for predicting future ecosystem carbon budget and ecosystem resilience to various environmental stresses.