Trees colonising formerly open tundra as the climate warms could cause Arctic ecosystems to release vast amounts of stored soil carbon into the atmosphere, a new paper argues. Many climate models have assumed that trees taking over the Arctic, and the enormous increase in plant biomass this would bring, would cause these landscapes to absorb much more carbon than they did before, helping restrain the effects of climate change. But this study suggests that\'s far from certain. In Scandinavia at least, when tundra heath turns into birch woodland it seems it could release much of the carbon stored in the soil into the air. This will more than counterbalance the fact that a forest holds around twice as much carbon in its biomass. So far from holding climate change in check, accelerated tree growth, and colonisation of treeless landscapes, could speed it up. The question matters a great deal, because Arctic soils hold a huge amount of carbon - much more than is in the whole atmosphere at any given time. If all this carbon were released in a short period, the effects on the climate would be profound. And because global warming helps the birch trees get a foothold on tundra heathland, this could turn into a vicious circle whereby warming causes more carbon release, leading in turn to further warming. \'Our study questions the idea that high-latitude environments will be big carbon sinks\' says Dr Iain Hartley of the University of Exeter, lead author of the paper in Nature Climate Change. \'It\'s hard to know the magnitude of these effects at the moment, but this certainly suggests that our understanding of the effects of environmental change in the Arctic needs to change - greater plant biomass doesn\'t always mean greater carbon storage.\' The team examined sites around the forest/tundra boundary (known as an \'ecotone\') of northern Sweden, comparing the stocks of carbon in open tundra plants and soils compared to those in nearby woodlands dominated by mountain birch, a species expected to account for much of the colonisation of the far north of Scandinavia. They found that even though these vigorous trees produce far more biomass than the grasses, lichens and shrubs they will replace, that doesn\'t mean the ecosystem as a whole holds more carbon - instead, the trees\' presence seems to make organic matter in the soil break down more quickly, and this means the stock of carbon built up in the tundra soil could be released into the atmosphere in a short period. To understand the differences between the two habitats, the researchers had to be able to distinguish carbon that\'s cycling quickly through the soil system from carbon that\'s stored there over long periods. To do this, they turned to radiocarbon analysis. The use of the first nuclear weapons in the mid-twentieth century sent a pulse of radioactive carbon into the air that left a permanent signature in the biosphere at the time. So scientists can fairly easily tell things that were alive before then from things that came afterwards. The team analysed the radiocarbon signature of the CO2 being given off by the soil in the two different ecosystems, learning that the trees seem to be stimulating the release of carbon from the soil. Scientists aren\'t sure exactly how this process, known as \'soil priming\', happens. They suspect that in general terms it\'s to do with soil-dwelling microbes, particularly the mycorrhizal fungi, which have a cooperative relationship with tree roots. As the trees grow they provide these organisms with more carbohydrates, which in turn lets them break down organic material in the soil more quickly and efficiently, as well as helping the trees to access nutrients such as phosphorus and nitrogen. The findings suggest that methods used so far to assess the effect of climate change on carbon stocks in the Arctic, such as vegetation models and using satellite data to estimate carbon levels, may need to be reconsidered, as they often assume that tree growth will lead to more carbon being absorbed overall. But Hartley says much more bigger and longer-term studies will be needed to provide the data that climate models will need to represent the effects of soil priming accurately. This could involve planting birches in sections of tundra heath and monitoring the effects on the soil as the years pass. He notes that it\'s not certain how far the findings apply to trees other than birches, and that more work is needed to see if conifers will have a similar effect. Likewise this study looked at well-drained tundra without permafrost; different dynamics could well apply in other habitat types, including, for example, the UK uplands. For the latter, the study indicates that caution will be needed when assessing the consequences of afforestation of heathland ecosystems with peaty soils, for overall carbon fluxes. The research is part of the NERC-funded Arctic Biosphere Atmosphere Coupling at Multiple Scales (ABACUS) project, and involved scientists from the Universities of Exeter, Sheffield and Stirling as well as the NERC Radiocarbon Facility and the James Hutton Institute.