Influence of atmospheric CO₂ on vegetation: evidence from past climates

Changing levels of the greenhouse gas carbon dioxide have a direct influence on vegetation. Plants depend on CO₂ for growth and development. Thus, when the atmosphere contains low levels of CO₂, as during ice ages, forests are reduced. Taking this effect into account makes climate models more reliable. This is the conclusion of a recent article in EGU's open access journal Climate of the Past by Colin Prentice and Sandy Harrison of the University of Bristol, UK.

In their paper Prentice and Harrison summarize evidence that atmospheric CO₂ has directly affected both plant growth and ecosystem composition in the past, and therefore will likely continue to do so in the future. These direct effects are included in dynamic global vegetation models (DGVMs). DGVMs predict, correctly, that land ecosystems are taking up about a quarter of anthropogenic CO₂ emissions. This is similar to the total amount taken up by the world’s oceans. According to DGVMs, direct effects of CO₂ on plants are the main cause of the current land uptake of CO₂, and also of the encroachment of trees which has been observed in savanna ecosystems all over the world. These attributions are supported by evidence from the past.

There is more CO₂ in the air today than there has been for 20-30 million years. But there were natural variations, long before humans started to clear forests and burn fossil fuels. Low CO₂ levels kept the world cold in ice ages; higher levels kept climate warmer at other times.

CO₂ is unique among greenhouse gases because it provides the carbon that plants need to grow. Ice ages were hard for plants; low CO₂ nearly starved the trees! When climate and vegetation models are given this information, they simulate a realistic ice-age world, in which forests were greatly reduced. Without this information, the simulated reduction of forest area is less than half what it should be (according to fossil pollen data, which record the past distribution of forests). So when reconstructing past climates, we need to use models that allow CO₂ to affect plant growth.

This insight explains why for a long time climate models were thought to underestimate ice-age cooling. The Mediterranean region is a case in point. For a long time it was thought that ice-age temperatures in the Mediterranean region were perhaps as much as 20 degrees lower than today, based on pollen records showing vegetation similar to central Asian steppes. But this analogy proved false. Steppe-like vegetation is also favoured by low carbon dioxide levels. When this effect is factored in, it appears that a more moderate cooling of 5 to 10 degrees is all that is needed to explain the ice-age vegetation. Unlike the previous reconstructions of very cold conditions, the moderate cooling suggested by new, model-based reconstruction methods agrees with climate-model predictions of the ice-age climate.
There was also about 10-30% less biomass and soil carbon on land during the last ice age. The numbers are rough, but they point the same way. Climate change alone cannot explain why the land stored less carbon. The key is the effect of low carbon dioxide levels on plant growth.

Large-scale field experiments in forests have confirmed that raising CO₂ concentration does increase plant growth, but there is still no consensus on whether this effect is persistent, or whether it really applies at a global scale. Similarly, there is no consensus as to whether tree encroachment in savannas is really an effect of rising CO₂. The evidence from the past speaks to both issues. It suggests that changing CO₂ concentration does produce long-lasting effects, both on global land carbon storage and on the competitive success of trees, that are consistent with the findings of DGVMs.

Ecosystem effects of CO₂ concentration: evidence from past climates by I. C. Prentice and S. P. Harrison http://www.clim-past.net/5/297/2009/ .

Information for editors
Author contact:
I.C. Prentice This e-mail address is being protected from spam bots, you need JavaScript enabled to view it
QUEST, Department of Earth Sciences
University of Bristol, Wills Memorial Building
Bristol BS8 1RJ, UK