By Mark Kinver
Science and environment reporter, BBC News
Ecologists have seen a sharp increase in the number of trees dying
Old growth trees in western parts of the US are probably being killed as a result of regional changes to the climate, a study has suggested.
Analysis of undisturbed forests showed that the trees' mortality rate had doubled since 1955, researchers said.
They warned that the loss of old growth trees could have implications for the areas' ecology and for the amount of carbon that the forests could store.
The findings have been published in the journal Science.
"Data from unmanaged old forests in the western US showed that background mortality rates have increased rapidly in recent decades," the team of US and Canadian scientists wrote.
"Because mortality increased in small trees, the overall increase in mortality rates cannot be attributed to ageing of large trees," they added.
"Regional warming and consequent increases in water deficits are likely contributors to the increase in tree mortality rates."
After ruling out a variety of other possible factors, including insect attacks and air pollution, the researchers concluded that regional warming was the dominant contributor.
Scientists fear the die-back could be a signal of worse to come
"From the 1970s to 2006, the mean annual temperature of the western US increased at a rate of 0.3C to 0.4C per decade, even approached 0.5C," they observed.
"This regional warming has contributed to widespread hydrological changes, such as declining fraction of precipitation falling as snow, declining snowpack water content, earlier spring snowmelt and a consequent lengthening of summer drought."
The team, led by the US Geological Survey (USGS), examined data from 76 temperate forest stands older than 200 years, which contained almost 59,000 trees.
Over the study period, which stretched back to 1955, more than 11,000 trees died.
The researchers reported that the increased mortality rate affected a range of species, different sized trees, and all elevations.
"The same way that in any group of people, a small number will die each year; in any forest, a small number of trees will die," explained co-author Phil van Mantgem, a USGS ecologist.
"But our long-term monitoring shows that tree mortality has been climbing, while the establishment of replacement trees has not."
The change in the forests' dynamics, the team noted, was going to have an impact on the forests' ecology and carbon storage capabilities.
"We may only be talking about an annual tree mortality rate changing from 1% a year to 2%, but over time a lot of small numbers add up," said co-author Professor Mark Harmon from Oregon State University.
He feared that the die-back was the first sign of a "feedback loop" developing.
As regional warming caused an increased number of trees to die, there would be less living trees to absorb carbon dioxide from the atmosphere.
Yet there would be an increased proportion of decaying trees, releasing the carbon that had been locked away inside the trees' wood.
Warmer temperatures might also increase the number and prevalence of insects and diseases that attack trees, the team added.
They used the example of recent outbreaks of tree-killing bark beetles in the US, which have been linked to a rise in temperatures.
Another member of the team, Dr Nate Stephenson, said increasing tree deaths could indicate a forest that was vulnerable to sudden, widespread die-back.
"That may be our biggest concern," he warned.
"Is the trend we're seeing a prelude to bigger, more abrupt changes to our forests."