"Our findings are essential to understanding how future climate change will affect terrestrial carbon dioxide exchanges with the atmosphere in the 21st century," Chuixiang Yi of City University of New York told environmentalresearchweb. "Our results suggest that the most likely climate-change scenarios would strongly intensify terrestrial carbon dioxide uptake in high latitudes and weaken uptake in low latitudes."

At mid-latitudes, meanwhile, NEE was a function of both temperature and dryness; the team found that the sensitivity of NEE to mean annual temperature broke down at a threshold value of ˜16 °C, above which terrestrial carbon dioxide fluxes were controlled by dryness rather than temperature.

Yi and colleagues from City University of New York; Oak Ridge National Laboratory; Pennsylvania State University; the Swedish University of Agricultural Sciences; Polytechnic Institute of Leiria, Portugal; and Duke University, US, believe that climate modellers should stop using universal control parameters for terrestrial ecosystem carbon dioxide exchange; instead they must use different parameters according to the climatic zone.

"This study is timely because understanding how terrestrial carbon fluxes are controlled by climate change is crucial for the development of successful policy," said Yi. "Management of the terrestrial carbon cycle will almost certainly be a key component of new policies, requiring detailed understanding about the carbon balance of all major ecosystems. Our observational findings would provide much needed evidence for policymaker debates."

Researchers from a large number of institutions provided data from 125 sites over 6 continents for the study. The FLUXNET project uses eddy covariance sensors around the world to measure the movements of carbon dioxide, water vapour and energy.

"Although several synthesis efforts have demonstrated that climate is not a major control of net ecosystem exchange (NEE) across multiple sites on annual or longer time scales (Law et al 2002, Valentini et al 2000, Rechstein et al 2007), I strongly believe that climate control is the most important control of NEE," said Yi, who reckons that earlier studies may have relied too much on precipitation as a measure of whether the soil is dry, when factors such as evaporation and transpiration are also important.

Yi's initial study of NEE data and climate did not show any link. "However, I continued to believe that climate control is the most important control of NEE and realized that the seeming lack of climate control occurs because different groupings of sites have different climate controlling factors," he said. When the team split the dataset into three groups, this clearer pattern emerged.

"I grew up in a small village in North China," said Yi. "I know well how important rain was to farmers for their agricultural harvests and had seen how they desperately prayed for rain. NEE must be related to the strength of ecological activities of ecosystems and hence related to the biome distribution. This biome distribution closely follows major climate classifications, revealing that climate controls are the most important controls for biome formation. This was the basis for our hypothesis: climate factors are the dominant factor in NEE variability."