Duke Studies Show U.S. Central Plains Vulnerable To Global Warming

In a report prepared for a meeting of the Ecological Society of America (ESA), the Duke scientists said the region responded readily to climate changes lasting decades or centuries during the mid-Holocene period -- 8,000 to 4,000 years ago -- by flip-flopping between grassland and forest. During this period, immediately after the last ice age, the region underwent many such short-term cycles of warming and cooling.

"We're finding that this system is really responsive, with the grasslands expanding eastward into forests and an increase in burning of this tallgrass prairie," said lead author James Clark, a professor of botany. Other co-authors of the paper were graduate students Eric Grimm and Jason Lynch.

In a variety of other Duke papers prepared for the meeting, Clark and his colleagues also reported more advances in understanding the ecological impact of global warming, including effects of climate change on carbon release from peatlands and on seed germination. This work was sponsored by the National Science Foundation and the Department of Energy.

Also at the meeting, the Duke ecologists reported other basic ecological studies that included methods to predict tree mortality, the effects of seed-eating predators on the production of seedlings and how seed dispersal affects biodiversity. This work was sponsored by the National Science Foundation.

In their studies of climate changes in the Central Plains, Clark and his colleagues analyzed pollen and fire-produced charcoal laid down in sediments in lakes in western Montana, the Dakotas, Nebraska, Minnesota and Wisconsin. Their analyses revealed the plants that dominated the region at various times during the period, as well as the incidence of fires fed by tallgrass prairies.

Their conclusion from the study -- that the region rapidly changed with even short-term climate fluctuations -- was not surprising, Clark said.

"One reason this region is so susceptible to climate change is that it lies at the boundaries between air masses that have very different climates associated with them," he said. "The center of the continent is dry because it is dominated by Pacific air that has passed over the mountains and lost its moisture. At that point, however, it loses out to tropical air from the Gulf of Mexico, which provides the eastern states with moisture. In addition, there's an Arctic air mass that plays a climatic role.

"Subtle changes in the factors that affect atmospheric circulation can translate into shifts in the boundaries where these air masses meet. And small changes in those boundaries can change the climate from a prairie climate to a forest climate."

Climatologists expect that the "greenhouse effect" -- atmospheric heating caused by increases in solar-energy-trapping carbon dioxide from fossil fuel burning -- will significantly increase temperatures worldwide over the next century. The past climate changes studied by Clark and his colleagues were not caused by increased carbon dioxide, but rather by changes in solar heating due to a wobble and tilt in Earth's spin and changes in its orbit around the sun.

In another paper prepared for the ESA, Clark and graduate student Philip Camill reported on the expected effects of global warming on peatlands in permafrost regions of Earth's northern latitudes. Scientists have theorized that such warming may speed decomposition of the vast peat deposits, increasing even further the release of carbon dioxide into the air.

However, Camill's research suggests that global warming might actually produce a sequestration of carbon dioxide in the peat deposits over the short run.

Peatlands are marked by an ever-changing mosaic of collapsed regions called scars, in which the peat has thawed and abruptly released large amounts of carbon dioxide before reforming.

In attempting to understand how this mosaic will change as global temperatures rise, Camill studied the behavior of the peatlands along a north-to-south line from colder to warmer regions. The colder regions feature more or less permanent permafrost, while the warmer ones show considerable scarring.

Camill discovered that the local dynamics of peat collapse and reformation dominate over the north-to-south temperature gradient, said Clark.

"This finding means that, even though it gets warmer as you go north to south, the overall balance of carbon uptake and release in the scars, and the scars' re-growth, is not dominated by the regional climate, but by the local processes. These processes include the growth of trees in these scars, which encourage the re-growth of permafrost." Camill's finding has implications for global warming effects, said Clark.

"These findings say that as the climate warms, you would not get a step-wise retreat of peat lands. But you would begin to speed up this local process which, rather than a net release of carbon dioxide, might cause a transient uptake as these scars reform into peat. The transient dynamics on the decades to centuries scale could be very different than the equilibrium change after a thousand years of warmer global temperatures.

Because climate change may disproportionately affect trees during their vulnerable early stages as seedlings, graduate student Janneke and Clark reported to the ESA meeting studies of how climate changes might affect seed germination. These early studies hint that under global warming some tree species may suffer reduced germination and seedling growth, a process botanists call "recruitment."

Another ESA meeting paper by graduate student Peter Wyckoff and Clark reported the latest advances in an "epidemiological" method Wyckoff developed to predict the death of trees from their growth rates.

"It's usually very difficult, if not impossible, to know why a tree dies," explained Clark. "You can't do an autopsy to figure it out." Nevertheless, he said, predicting tree mortality is extremely important to ecologists studying the complex dynamics of forests. Measuring such mortality is a short-term observation that helps ecologists extrapolate the dynamics of forest tree populations that occur over decades to centuries.

In his paper, Wyckoff reported new techniques of his mathematical method of gauging the probability of a tree's death according to its growth rate, with a lower growth rate indicating a higher risk.

"There are a lot of reasons why a tree can die, ranging from insects to insufficient carbon, but the effects of all of them are summarized by its growth rate," said Clark. "Peter's model is an elegant approach, because you don't need to know what all those things are."