The roles of vegetation structure and composition in terrestrial ecosystem responses to climate change

Vegetation structure, species compositions, and interactions among individuals change with the development of ecosystems. The sensitivity of ecosystems to climate change is therefore differed. However, current dynamic global vegetation models treat vegetation as a couple of compartmentalized pools and neglect the details of vegetation dynamics with only explicitly representations of physiological and biogeochemical processes. These may lead to bias in the predictions of ecosystem dynamics in response to climate change. We incorporated a vegetation structure model, perfect plasticity approximation model (PPA), into a land model, LM3V, the land component of GFDL ESM, and explored the sensitivity and uncertainty induced by the structure of forest ecosystems in Northeastern USA, where the PPA model explicitly describes the competition of light and resources from soil (water or nitrogen) among plant species and the resulted vegetation structure with different environmental conditions and LM3V is a biogeochemical model describing carbon and water fluxes and storage in ecosystems and their feedback to atmosphere. We employed the rules of PPA to represent the vegetation dynamics (species composition, competition for light and water, and vegetation structure) in LM3V. The simulations showed that vegetation structure and the interactions between species could substantially affect the sensitivity of ecosystem carbon cycle to multi-factorial changes of climate. With changes in vegetation compositions and structure, ecosystem responses were less sensitive than those when only plant physiological and biogeochemical responses were considered. Our study shows that the processes at plant community level play a key role in mid- to long term ecosystem responses to climate change.