If you have spent any time wandering in alpine areas, you no doubt have heard the familiar high-pitched "eek" of the pika. These small mammals -- also known as coneys and rock rabbits -- live on the edges of rock slopes next to alpine meadows. They spend their summers busily collecting vegetation of all kinds, storing it in haypiles among the rocks for use the following winter.

But the last two winters have not been kind to the pikas -- at least to some of the ones living in the Ruby Range near Kluane Lake. Some pika populations there have dropped by 80 to 90 percent, and one scientist thinks that this decline may be an early warning sign for climate change.

David Hik, a professor at the University of Alberta in Edmonton, has been studying plant and animal communities living above treeline in the Ruby Range. He thinks that alpine species might be more sensitive to changes in climate than those found in valley bottoms, and in 1995 he set up several study areas in the Kluane area so that changes could be tracked if they did occur.

Hik says that when he started this project, he never dreamed that he would see such dramatic effects so quickly. But then again, the last two winters have been anything but normal in the Kluane area.

The weather station in Haines Junction registered a high of almost 17 degrees Centigrade one day last winter, and it rained several times during the Christmas holidays. Humans living in the valley found these conditions unsettling enough, but for the pikas -- it was probably devastating.

Pikas do not hibernate, and they do not put on much fat during the summer. But under normal conditions, long intense winters are not a problem for these little herbivores. They live underneath the snow, and spend their winters munching on vegetation in the haypiles that they built the previous summer.

The snow insulates them from the colder air temperatures, but dry snow is definitely the preferred option as it insulates better than soggy snow. Also, if ice forms in the snowpack, it is harder for pikas to forage in what is called the subnivean layer.

For the first four years of Hik's study, he found between 80 and 100 pikas living in the study area. But after the warm winter of 1998-99, the number of pikas surviving until the spring had dropped by about 80 percent.

However, the numbers were almost back to normal by the end of the summer, and Hik assumes that enough baby pikas survived or migrated in from other areas to partly replace those individuals who had died.

Then came the second disastrous winter, and pikas died in large numbers once again. In the spring of 2000, Hik and his students could find only 11 pikas in the whole study area, and there were none in the core area. And this time the population did not bounce right back; by the end of the season only four pikas had been spotted.

"We never anticipated that we would have a system flip from one state to another so quickly," says Hik. "The regional population obviously took a hit as well. They will recover, but the question is how quickly that will happen."

Hik says that not much is known about alpine ecosystems, and particularly about the climate at higher elevations. Weather stations are located on valley floors, where people live, and the weather can be radically different at higher elevations in winter.

In most of the Yukon, strong temperature inversions in winter mean that conditions are warmer on mountain ridges than on the valley floors below them. Hik has not yet looked at all his data from the previous winter, but he has started to find evidence that these inversions are not as pronounced during unseasonably warm winters.

A long-term change of this sort could have huge implications for animals and plants both in alpine areas and in the boreal forest. Some researchers think that these winter inversions drive the whole pattern of predation in the boreal forest. Since cold air holds less moisture than warm air, inversions help reduce the amount of snow falling in the valleys.

The thinner snowpack makes it easier for predators like wolves and lynx to hunt in winter. If snowfall increases in the boreal forest, predator-prey cycles there could be changed.

Hik thinks that mammals in alpine ecosystems could prove to be more useful indicators for climate change since predation does not play as big of a role there as it does in the boreal forest. And, while the warmer winters have been devastating for pikas, other alpine mammals appear to be thriving.

Researchers are also tracking marmots in the study areas, and their numbers have been almost doubling every year. "The difference is that they are social hibernators," says Hik. "They sleep together in family groups. So that might help them deal with climate variability. Also they are much larger animals than pikas."

Ground squirrels also hibernate and store fat for the winter, and so far they seem to be faring better than the pikas. Although 80 percent of the ground squirrels died the first warm winter, their numbers bounced back a little bit this year.

A short growing season also makes alpine areas more sensitive to changes in climate, and last summer, the season was even shorter than normal. Normally the snow has melted on vegetation plots in Hik's alpine study sites by about June 15. But, after a big dump of snow in May, the snow lingered there until July 23.

That delay meant less food for herbivores like pikas, and less time for collecting whatever vegetation did eventually grow. Hik points out that if the snow lingers too long in summer, some plants might not grow at all.

Hik says that it is too early to say what the future might hold for the pikas and other mammals living above treeline, as so many factors come into play. And while it has been fascinating to watch the changes that can happen in a short period of time, he is hoping that conditions return to normal sooner rather than later.

"We are really hoping for a good year for the pikas because they are just dropping out of the system," he says. "A more typical winter should let their numbers increase and also let us see if our ideas about the consequences of climate variation on their numbers are correct."

David Hik can be contacted at david.hik@ualberta.ca.