Long-term CO2 rise has increased photosynthetic efficiency and water use efficiency but did not stimulate diameter growth of tropical trees

Tropical forests are a crucial component of the global carbon cycle, and their responses to atmospheric changes may shift carbon cycling and climate systems. Dynamic Global Vegetation Models (DGVMs) are the major tools to simulate tropical forest responses to climate change. One of the main determinants of these simulated responses is the effect of CO₂ on tropical tree physiology and growth, the 'CO₂ fertilization effect'. The paucity of CO₂ enrichment experiments in the tropics importantly limits insights into the CO₂ fertilization effect as well as the validation of DGVMs. However, use can be made of the 40% rise in atmospheric CO₂ concentration since the onset of the Industrial Revolution. The effects of the historical CO₂ rise on tree physiology and growth can be obtained from stable isotopes, isotopomers and tree diameter increments obtained in tree-ring studies. We studied the physiological and growth responses of 12 tree species in Bolivia, Cameroon and Thailand to 150 years of CO₂ enrichment.

Analyses of ¹³C of wood cellulose revealed strong, long-term increases in leaf intercellular CO₂ concentrations for all study species and a marked improvement of intrinsic water use efficiency (iWUE). For a subset of one species per site, we studied the Deuterium isotopomers (isomers with isotopic atoms) of glucose in wood to obtain a direct estimate of the photorespiration-to-photosynthesis ratio. We found that this ratio consistently and strongly decreased over the past century, thus increasing the effeciency and rate of photosynthesis. In spite of these strong physiological responses to increased CO₂levels, we did not find evidence for increased tree diameter growth for any of the sites, or for sites combined. Possible reasons for the lack of a growth stimulation include increased (leaf) temperature, insufficient availability of nutrients or a shift in biomass investment in trees.

Our results suggest that the strong CO₂ fertilization of tropical tree growth often assumed in DGVMs does not hold and that these models may overestimate future biomass production in tropical forests. Empirical information on responses of tropical trees to historical CO₂rise as presented here can be used to validate and possibly adapt (components of) DGVMs and improve the projections of tropical forest structure under climate change.