Beginning in the 8th century and continuing for 150 years, the great Mayan civilization of the Peten rain forest in present-day Guatemala fell apart. Cities were abandoned, people fled and wars raged across the encroaching wilderness.

This prolonged event -- known traditionally as the Maya "collapse" -- is one of the enduring mysteries of pre-Columbian America and a subject of continued debate. How did it happen?

In research reported yesterday, a German-led team of earth scientists offered new evidence that a 200-year dry spell, punctuated by three periods of serious drought, may have played an important role.

"There's competition for food, there are wars, there's deforestation, and the climate is drier," said paleo-oceanographer Gerald Haug of Potsdam's Geoscience Center. "These were problems you could cope with to a certain degree -- but then you had the extremes. It's a subtle catalyst."

By measuring the undisturbed sediments of Venezuela's Cariaco Basin on the Caribbean coast, Haug's team was able to identify a significant decline in regional rainfall beginning around A.D. 750, with drought spikes starting at 810, 860 and 910.

The sequence corresponds fairly closely to protracted Maya upheavals that began in the western Peten in the late 7th century, and in the central Peten lowlands in the 9th century. By A.D. 930, some archaeologists calculate that the Maya heartland had lost 95 percent of its population.

For more than a century, this diaspora bewildered archaeologists even as it cemented the popular vision of a "lost civilization" of spectacular pyramids and monuments overtaken by jungle in a trackless tropical wilderness.

Much more is known today, and archaeologists are much less likely to accept overarching theories for the "collapse," a term that is losing cachet as evidence accumulates that the Maya did not "disappear," but simply moved: north to Yucatan in Mexico, eastward to Belize and to highland settlements on the edges of the rain forest.

"It's not a question of whether there was a drought or an invasion. There wasn't some big, single anything that happened at some big, single time," said Vanderbilt University archaeologist Arthur A. Demarest, who is editing a book on the period. "This kind of theory doesn't have a place anymore, given the detail of cultural history we have."

More sympathetic was the University of Pennsylvania's Robert J. Sharer, author of a classic text on the Maya, who noted that "climate changes, including drought, have always been part of the mix," and "the argument has been strengthened" over the last 10 years. "But what everybody wants is a pat answer," Sharer continued, "and we're still not at that point, and probably never will be."

The new research, reported yesterday in the journal Science, was sponsored by the Ocean Drilling Program, a multinational initiative led by the National Science Foundation. Haug's team studied the topmost layers of a 170-meter Cariaco Basin core sample.

The basin in Venezuela is about 1,800 miles east of the Peten, but both places lie on the "Intertropical Convergence Zone," also known as the doldrums, a band that encircles the Earth where the northern and southern trade winds meet to create a region of almost perpetual thunderstorms. When it rains in the basin, it is raining in the Peten.

"The Cariaco Basin is the best climatological archive in the tropics, and since the

Maya region is clearly affected by the same climate, it was perfect for us," Haug said.

Just as important, the deeper waters of the Cariaco Basin have been oxygen-free for 14,600 years, since rising sea levels breached natural barriers and filled what had been a lake, displacing fresh water, which subsequently returned to cover the surface like a suffocating blanket. The sterile environment allows sediment to fall unimpeded to the bottom of the basin, where it rests undisturbed.

During the rainy season, runoff from the surrounding area deposits dark sediments on the basin bottom. When the convergence zone migrates, the dry season sets in, and the sediments are lighter, composed principally of plankton from the basin's oxygenated top layer.

Each year has a dark and a light layer, and Haug's team found that the dark layers beginning around A.D. 750 became thinner, and then became much thinner during the three spiking periods of three to 10 years each.

"No one archaeological model is likely to capture completely a phenomenon as complex as the Maya decline," the authors wrote in Science. "Nevertheless, the Cariaco Basin sediment record provides support for the hypothesis that regional drought played an important role."

Demarest said, however, that the western Peten was receiving 100 inches of rainfall per year during the latter half of the 8th century, when warfare ravaged the region and destroyed the culture. "It's a very varied picture," Demarest said, and no single theory fits everywhere.

Besides drought and war, scholars over the years have placed varying degrees of blame for the Maya decline on pestilence, overpopulation, environmental degradation and class warfare, with varying degrees of evidence.

"But that kind of theory doesn't work anymore, because we have too many details," Demarest said. "The only approach that works is to look at it in detail, region by region, and hook up the cultural histories, and we are just doing that now. These changes in the Maya took a long time -- like the decline and fall of the Roman Empire."