The carbon fertilization effect (CFE) is in principle simple: the larger amount of carbon dioxide in the atmosphere that has resulted from rising anthropogenic emissions should help the growth of plants, which use carbon dioxide during photosynthesis. The effect ought to increase crop yields – and that is some good news for farmers, amid the overwhelmingly gloomy forecasts for other aspects of climate change.

In detail, however, carbon fertilization is not so simple. It tends to have a greater effect on plants when soil moisture is low, and it affects some plants more than others: those with "C4" photosynthesis systems, which can concentrate carbon dioxide onto reaction sites, are affected less than plants with "C3" photosynthetic systems, which cannot. But while many vegetation models have taken into account this broad C4/C3variation, there has been no provision for variation between species. This means, for instance, that the C3 species wheat and potatoes are treated the same, as are the C4 species corn and sugarcane.

Such a detailed provision is important, said David Lobell of Stanford University and Justin McGrath at the University of Illinois, because the characteristics of certain species can strongly affect their receptivity to the carbon-fertilization effect. Tuberous C3 species (including potatoes), for instance, are known to be more receptive than non-tuberous species, such as wheat.

Lobell and McGrath have now taken into account this species variation to provide preliminary estimates of how the carbon-fertilization effect ought to vary in the future on a regional level. They began with historic regional data linking water availability to crop yields. They then combined this with published estimates of how the carbon-fertilization effect varies with water availability for major staple crops, to predict how yield ought to vary with the carbon fertilization effect, given the 100 parts per million increase in atmospheric carbon dioxide expected by 2040.

The researchers found that the change in crop yields due to the carbon-fertilization effect was as much as 70% of the variability due to changes in climate, such as temperature and precipitation. "Experiments had shown that the response should be variable, but we did not expect the degree of variability to be so large," said McGrath.

McGrath stressed that he and Lobell cannot yet accurately predict future crop yields, which should also take into account interactions with temperature and nitrogen availability. Despite that caveat, he said that the regions that will get the biggest benefit from the carbon-fertilization effect are north Africa, Kazakhstan and Russia, while the regions set to receive little to no benefit are west Asia and Egypt.

The researchers are now planning to study how yield responds to temperature and nitrogen, in addition to the carbon-fertilization effect. "Those factors are likely [to be] equally as important as water availability for regional yield estimates," he said.