BILLINGS - Long before traffic clogged Main Street and backed up on Airport Road, Alkali Creek was a busy transportation corridor.

It must have been an inviting place with a gentle little stream where children could play without too much supervision and women wouldn't have to walk too far with a heavy burden of water.

Trails to bison herds on the bench above and easy access to the Yellowstone River less than a mile away may have attracted bands of hunter-gathers generation after generation in the millennia before contact with Europeans 500 years ago.

Just how far back human occupation goes in this small corner of the world is still a matter for study.

But Billings archaeologist Stephen Aaberg was surprised with the results of radiocarbon dating tests just completed on samples taken from trenches dug in Alkali Creek last summer.

The sample that dated to most recent times - charcoal picked from a hearth uncovered 6 to 10 inches below the grassy surface - was determined to be 1,050 years old. The oldest, a bison foot bone found near stone artifacts, was dated at 5,300 years old.

About 18 inches below the 5,300-year level, archaeologists working for Aaberg's company, Aaberg Cultural Resource Consulting Service, found a single piece of charcoal. Aaberg isn't sure what to make of it but believes it could be 7,000 to 8,000 years old. Also found at that level was a fragment of a long, narrow projectile point characteristic of that ancient period. (In later periods, spear points took on a more triangular shape.)

"We were surprised by the dates, but not so surprised that we just couldn't accept the data," Aaberg said.

It wasn't that the scientists doubted the presence of early humans in the bountiful Yellowstone River corridor. It was a good place for hunter-gatherers to make a living. But they didn't think the land itself had been stable long enough to protect archaeological features for thousands of years.

Just south of the site between Alkali Creek Road and Airport Road is a steep slope, known to the archaeologists as Skeleton Cliff, that could have shed a lot of debris over the ancient campsites. The creek itself could have been expected to take a toll in flood seasons and channel changes.

But when the workers began removing layers of soil and documenting the "horizons" where topsoil had been built from repeated cycles of plant deposits, they discovered they were looking at the record of thousands of years rather than hundreds, Aaberg said.

Aaberg was hired by the state Department of Transportation to document the area in preparation for rebuilding Airport Road. (All major road projects that use federal money require such archaeological surveys.) In 2005, he walked both sides of the portion of Alkali Creek Road that would be included in the project, looking for signs of early human occupation. He found enough on the surface to warrant digging meter-square test pits.

"We came right down on an old hearth," he said. "The site ended up being really substantially large. It's on both sides of Alkali Creek Road and right up to Airport Road."

Aaberg's report found that the road project would have a significant impact on the archaeological site, which prompted the excavations of 2007 to document the archaeology before the bulldozers started moving earth.

"At that point, we didn't know how old or how deeply stratified it would be," he said.

Crews worked clearing 10 centimeters of soil, or about 4 inches, at a time at likely spots within the perimeter of the site.

"We were running into stuff from the first 6 to 10 centimeters," he said. "From 10 to 40 centimeters, it was really heavy."

They found cutting tools, stone knives, broken projectile points, animal bone and "pebble cores," the characteristically scarred rock left over after flakes were chipped off to make tools.

The first major feature was the hearth that proved to be 1,050 years old. Its builders had dug a little pit, and inside were fire-cracked rock and an abundance of charcoal.

About 4 feet below the first pit, archaeologists digging a trench with a backhoe found a second hearth of almost identical construction. The second hearth was dated to 2,600 years ago.

Evidence suggests that these hearths probably were used to heat rocks that would be dropped into separate cooking pits filled with water for boiling food, Aaberg explained. The two-part cooking process can be identified in the way the stones cracked. Rocks heated directly on a fire show a different breakage pattern from hot rocks that hit water, he said.

About 90 yards from the pit used to heat rock, archaeologists found what they believe is a stone boiling pit where food would have been cooked. The features may not have been constructed during the same occupation, but their differences help define what they were used for. The pits used for heating rock were filled with charcoal and were ringed by orange marks left by mineral oxidation. The boiling pit contained no charcoal and no orange ring.

They found the bones of grouse-sized birds, deer and antelope as well as bison. There weren't enough bison bones to indicate that large numbers of animals were processed there at one time.

But the finds suggest that the Alkali Creek site may have been occupied about the same time that a known bison kill site nearby was in use. That site, now buried under MetraPark Arena, was first excavated in the 1930s, before radiocarbon dating techniques were available.

"Projectile points found at the kill site are almost identical to one found at one of the occupation sites here," Aaberg said.

The occupation site contemporary with the Metra kill site points was at the 2,600-year level - approximately 600 B.C.E.

"We found a fair amount of bison bone at first and since it is less than a half-mile from the Metra bison kill site, we thought maybe it was a processing site," Aaberg said. "But excavation showed more variety.

"Now we're thinking it might not have been used with the Metra kill site."

Researchers did find some obsidian that may have been brought from the area of Yellowstone National Park, but Aaberg suggests that it may have been acquired through trade rather than travel. The obsidian will be submitted for tests that will show exactly where it originated.

During the survey, archaeologists collected hundreds of stone flakes, most of them less than a quarter-inch long. Flakes constituted about 95 percent of the artifact total.

"Our models for lithic technology went out the window," he said. "There just weren't a lot of large or intermediate flakes."

Ordinarily, an absence of larger flakes would have meant that people at the site weren't producing tools - just finishing and maintaining them. But Aaberg said that his team found pebble cores around the site that were byproducts of actual manufacturing.

A check of the rock material on Skeleton Cliff showed that most of the cores were native stone. The flakes were small because the rocks they were working with were small, not the chunks of chert or obsidian that would have had to come from someplace else.

"It's a bit of a mystery how these people worked little pebble cores into tools," he said. "It's something we need to spend more time on."

Since the occupation sites dated hundreds or thousands of years before Spanish explorers introduced horses, people who made camp on Alkali Creek probably didn't move in a wide circuit, Aaberg theorizes. "Foot people" transporting entire households and heavy hide tepees weren't likely to make epic journeys every few days.

"I think 100-mile radius would be pushing it," he said. "They probably didn't range beyond 50 or 75 miles radius. There really wouldn't be any need to unless there was some kind of catastrophic change."

The occupation site at Alkali Creek was likely a seasonal camp, although it's hard to tell what season. The fact that most tools were made from local rock may indicate that people camped there during the winter, when they were less mobile, he speculated.

There is no reliable way to estimate the size of the population at any given time, Aaberg said.

"What we can say is that particularly in the period beginning about 5,300 to 5,200 years ago, there are archaeological sites and artifacts all over Montana, especially east of the Divide," he said. "My guess is that there was a heck of a population. The game populations were high enough to support a pretty large human population."

The people who warmed themselves at the 5,300-year-old hearth - that's about the year 3,300 B.C.E. - probably didn't live much differently from the people who built a similar fire 2,700 years later, Aaberg said.

"Lifestyles didn't change much on the Northern Plains," he said. "Basically they were hunter-gatherers, and it appears hunter-gatherers up here on the Northern Plains had remarkably stable cultures."

He said there was not a lot of evidence of warfare, which displaces people and alters cultures suddenly.

These people probably lived in animal hide tepees and hunted efficiently with the same type of spear throwers for thousands of years. The bow and arrow, a major technological advance, didn't come into use in this region until 1,700 or 1,500 years ago. The horse didn't make it this far north until about 300 years ago.

It's harder to say what life would have been like 7,000 to 8,000 years ago, when the earliest evidence of human occupation was deposited near Alkali Creek. A severe, extended drought that stretched from 8,200 years ago to 5,500 years ago, called the Altithermal, had probably culled the human and animal population in the American West. But it's possible that the deep springs that feed Alkali Creek could have been flowing when others failed, providing a water source for game and a few hardy people who continued to hunt on the Northern Plains.

Aaberg notes that the density of artifacts at the Alkali Creek site begins to increase about 200 years after the Altithermal ends.

The Alkali Creek site presented archaeologists with a rare treasure - up to 15 acres of pristine ground, rich in artifacts, in the middle of a metropolitan area that has never been touched by a plow or residential or commercial development. Other cities have turned such resources into parks and cultural attractions, Aaberg said.

The Alkali Creek project adds new chapters to both the geology and archaeology of the Billings area, Aaberg said. A geo-archaeologist working with Aaberg estimated that the 11-foot-deep sediments on the site could be as old as 43,000 years - far more ancient than expected.

Alkali Creek could serve as a model for further study in the Billings region.

"There has not been a lot of archaeology in this part of the Yellowstone Valley," he said. "Very little has been done from Livingston on down."