Cactus Hill Archaeological Site

Cactus Hill is particularly important because, prior to the discovery of its earliest components, archaeologists generally concluded that the first human presence in the Americas was represented by the Clovis-age culture, dating to approximately 13,000 years ago. (Clovis-age culture is named for stone and bone projectile points—so-called Clovis points used on spears, darts, arrows, and knives—found near Clovis, New Mexico, in the mid-1930s.) Several other sites, such as Meadowcroft Rockshelter in Pennsylvania and Monte Verde in Chile, had probable and possible pre-Clovis components in good stratigraphic context. As enticing as the data from those sites were, however, they had yet to fully convince the archaeological community of a pre-Clovis human occupation in the Americas. The discoveries at Cactus Hill in the mid-1990s played a major role in changing that dominant perspective, often referred to as the Clovis-first paradigm. In subsequent years, research spread beyond Cactus Hill to the Chesapeake Bay and Middle Atlantic Region in general, and the revived interest in the pre-Clovis question led to new theories about how the first people arrived in the Americas.

The Clovis-first paradigm contended that the first Americans arrived from Siberia via the so-called Bering land bridge, also known as Beringia, approximately 13,500 years ago. The Bering land bridge was created by significantly lower sea levels during the last advance of the Ice Age glaciers, or what is known as the Wisconsin Glacial Episode. When the glaciers retreated at the end of the Wisconsin glaciation, the land bridge submerged and an ice-free corridor opened up between the eastern (Laurentian) and western (Cordilleran) parts of the glacier. Theoretically, this corridor connected modern Alaska with the central Great Plains and allowed people to migrate into North America. The ice-free corridor is hypothesized to have become available for migration between 14,500 and 14,000 years ago.

Habitation at Cactus Hill, however, dates to between 18,000 and 20,000 years ago, and when considered alongside the early dates from Meadowcroft Rockshelter and Monte Verde, it provides strong evidence that modern humans were in the Americas well before the ice-free corridor opened. Similarly old dates from other sites, including in the Chesapeake Bay region and in submerged contexts off Cape Charles and Cape Henry, helped to solidify this theory. Scholars have introduced several theories for how humans came to the Americas during and before the height of the Wisconsin glaciation. They might have used boats to skirt the glacier along the Pacific coast of North America or crossed along the edge of the North Atlantic pack ice that might have extended as far south as northern France. Less popular theories imagine people braving the open ocean, crossing either the central Atlantic or the Pacific. All of these models are difficult to directly investigate because the related sites would have been submerged by the post-Wisconsin rise in sea levels.

But Cactus Hill was instrumental in reviving the idea, first proposed early in the twentieth century, that humans may have come to the Americas across the North Atlantic. In 1998, Dennis Stanford of the Smithsonian Institution and Bruce Bradley of Exeter University in England proposed what they called the Solutrean model, named after an archaeological site near the village of Solutré-Pouilly in eastern France. They hypothesize that the famous Solutrean cave painters, who also made highly sophisticated stone tools, were likewise sophisticated sea hunters, like the Inuit, who lived in the Arctic regions. Solutrean exploitation of the North Atlantic pack ice could have drawn them to the east coast of North America and eventually to Cactus Hill.

The Cactus Hill site is on a stable, loamy sand hill consisting of mostly wind-deposited sand and silt derived from the adjacent Nottoway River floodplain, which is to the north and west. The current river channel runs north to south; however, that appears to be a function of geological processes that took place approximately 12,900–11,500 years ago, during what scientists call the Younger-Dryas cold reversal, or the last major cold period of the Wisconsin glaciation. During the site's earliest human occupations—dating to between 18,000 and 20,000 years ago—the river channel appears to have flowed from west to east immediately adjacent to the older parts of the site, which is currently located approximately 1,000 feet east of the modern river channel. Those early human occupation levels are covered by about three feet of loamy sand.

Although a sandy soil might be thought of as being relatively unstable for the preservation of strata, or layers of materials that archaeologists can date to different periods, the Cactus Hill site has nevertheless produced eleven types of evidence that support such early dating. These include:

Many of the above lines of evidence could have alternative explanations, but the simplest explanation for each taken in isolation and all taken together is that they support Cactus Hill's pre-Clovis occupation, dating 18,000–20,000 years ago. It is statistically unlikely that all or most of the lines of evidence have been misinterpreted by multiple archeologists and other scientists who have evaluated the data.

• (1) temporally diagnostic artifacts (like points and pottery) consistently in appropriate stratigraphic relationships;
• (2) stone tool types stratigraphically consistent between cultural levels;
• (3) stone raw materials stratigraphically consistent between cultural levels;
• (4) isolated stone hearths confined within cultural levels;
• (5) cross-mended artifacts within cultural levels (i.e., pieces of the same broken artifact that can be fitted back together);
• (6) radiocarbon dates consistent with the artifacts and stratigraphy;
• (7) corroborating soil dates (optically stimulated luminescence or OSL date) also consistent with the artifacts and stratigraphy;
• (8) soil chemistry (phosphorus) content consistent with stratigraphically separated, human occupation levels;
• (9) phytolith (grass "fossil") distributions consistent with discrete, stratigraphic human occupations;
• (10) soil strata consistent with minimal vertical and horizontal disturbance; and
• (11) replication of (1), (2), (3), (4), (5), (6), and (10) in more than one part of the site.

Cactus Hill was discovered by artifact collectors in the mid-1980s and eventually reported to retired state archaeologist Howard A. MacCord, who recorded the site as 44SX202. It was initially tested in 1988 by Joseph M. McAvoy of the Nottoway River Survey (NRS) and systematically excavated from 1993 to 2002 by separate volunteer and professional archaeological teams directed by McAvoy and Michael F. Johnson of the Archeological Society of Virginia (ASV).

The site's earliest component and date were discovered by the NRS in the mid-1990s. The component consisted of charcoal and a number of blade-like flakes several inches below a Clovis-age living surface. The Clovis surface contained two Clovis-like points, one in three pieces, and several Clovis tools. The artifacts were made of stone consistent with the kinds preferred by the Clovis-age people, all indicating an undisturbed living surface over the pre-Clovis surface. The NRS dated one piece of white pine charcoal from the deeper blade-like flake level to approximately 18,000 years ago. The NRS also recovered from another part of the site the first Clovis-age date yet recorded in Virginia—about 12,950 years ago—which is consistent with Clovis-age dates from other parts of North America.

In 1996, those findings were corroborated in a different area of the site by the ASV team. There a line of blade-like flakes and two pieces of the same broken point were recovered in excellent stratigraphic context eight inches below a Clovis level, which contained appropriate tools and a late-stage Clovis point preform.

Additional excavations by the NRS produced two reworked lanceolate/triangular points from pre-Clovis contexts, along with additional tools and dates, all indicating that the pre-Clovis occupation was between 18,000 and 20,000 years ago. Of particular note, the two points recovered by the NRS do not match any point type recovered in any stratified context in or above the Clovis levels on Cactus Hill or any other site yet known to have been excavated in the region.

A detailed monograph on the first three years of the investigation was published in 1997 by the Virginia Department of Historic Resources. That report covered the most significant finds at the site. Subsequent research on the site has been published in professional and nonprofessional articles.

Any archaeological site that is claimed to contain evidence that contradicts the prevailing theories must stand up to professional scrutiny. Cactus Hill was excavated using rigorous professional standards, including three-dimensional piece plotting of individual artifacts, that allow the data to be thoroughly examined by skeptics. In addition, the site was open to inspection during excavation. It was also excavated by two independent teams, working in different parts of the site, using a multidisciplinary approach.

Archaeological data recovered from Cactus Hill document that the site contains a full chronological sequence of human occupation dating from as early as 18,000 to 20,000 years ago up through the latest prehistoric time periods. Early Cactus Hill–like points from elsewhere have been reported to the NRS and ASV, but were always found in disturbed surface contexts, where no potential for deeply buried deposits was present. This fact suggests that the earliest occupants at Cactus Hill were not merely a founding population—or the first-ever in an area—but established hunter-gatherers who had intimate knowledge of the Nottoway River Valley and possibly beyond. The implication is that Cactus Hill is not the oldest human site in the Americas.

The most important unanswered archaeological question in the Western Hemisphere is, "Who got here first, and when?" To that end, Cactus Hill proved to be pivotal in breaking down the Clovis-first paradigm. The site has provided another pre-Clovis site; a chronological niche for those previously unidentified points; and a model for how people may have survived and possibly even prospered in harsh, near-glacial conditions. (It is unlikely that the site was occupied during the winter months, but the earliest inhabitants may have found the exposed location desirable during the summer, with winds off the mile-thick glacier, located less than 700 miles to the north, keeping down biting insects.) As a result, it has helped stimulate further research into that remote period of Virginia's and America's cultural history. Both the NRS and ASV teams have moved on to the task of trying to replicate the Cactus Hill discoveries elsewhere. The theory is that if there is one Cactus Hill, there should be others.

The Cactus Hill Archaeological Site was listed on the Virginia Landmarks Register in 2001.