A sinkhole in Florida’s Aucilla River has become one of the most heavily studied archaeological sites in the Americas, with evidence indicating humans were present there roughly 14,550 years ago, more than a thousand years before the culture long credited as the first in North America.
The site, known as Page-Ladson, sits in the Big Bend region of Florida where the Aucilla River runs dark with tannins. During the last ice age, when sea levels were lower and the climate drier, the location was a spring-fed watering hole. Mastodons, camels, horses, and bison came there to drink, and evidence now shows humans did too.
For most of the 20th century, archaeologists agreed on a simple story. The Clovis culture, identified by distinctive fluted stone points found in New Mexico in the 1920s and 1930s, was believed to be the first to reach the Americas, arriving roughly 13,000 years ago across a land bridge from Siberia. The theory held that these people moved through an ice-free corridor and spread rapidly across two continents. Anything suggesting an earlier arrival was dismissed as contamination or flawed methodology, and challenging the model carried professional risk.
In the early 1980s, a former Navy SEAL exploring the river by scuba discovered ancient bones at the sinkhole. That led to a decade-long project beginning in 1983, run by paleontologist S. David Webb and archaeologist James Dunbar. Working in near-total darkness, they recovered stone tools associated with butchered mastodon bone and dated the layer to roughly 14,400 years ago. In 1993, they found a mastodon tusk with markings consistent with human cutting. The findings were published, but the wider field largely ignored them. The dates were older than Clovis, and the site was underwater, making independent verification difficult.
In 2015, two respected archaeologists, Donald Grayson and David Meltzer, published a formal critique. They did not dismiss the site entirely, but noted the earlier publications lacked clear photographs and descriptions of the cut marks. It was a procedural challenge, demanding better evidence.
That challenge prompted a return to the site. In 2012, a new team led by Jesse Halligan, then a doctoral student at Texas A&M, and Michael Waters, director of the university’s Center for the Study of the First Americans, returned to the sinkhole. They were joined by Dunbar, paleontologist Daniel Fisher, who studies how ancient people butchered large animals, radiocarbon specialist Thomas Stafford Jr., and researchers from several other universities.
The team worked in conditions that Halligan described as literally diving into iced tea, with tannins blocking nearly all light. Divers worked by touch, mapping their excavation grid by memory and fixed survey points. They found a small stone biface, a knife flaked sharp on both sides, made from local chert, meaning it was made on site by people familiar with the landscape. It did not resemble Clovis technology. They also recovered five more stone artifacts alongside the bones of mastodon, ancient camel, and bison.
The original tusk from 1993 was reassembled and examined again by Fisher. He found more than a dozen deep parallel grooves across the base of the tusk, in the spot that had been embedded in the mastodon skull. The cuts ran perpendicular to the tusk’s length and were located exactly where a person would need to cut through a ligament to remove the tusk. Fisher, who has personally butchered mammoth and mastodon carcasses to study how stone tools interact with bone and ivory, ruled out trampling or scavenging as explanations.
The removal of the tusk would have made practical sense. A tusk of that size holds more than 15 pounds of soft, nutritious tissue in its pulp cavity, and the ivory itself could be used for tools.
The evidence does not confirm how the mastodon died. No spear point was found lodged in its ribs. The cut marks confirm human hands were on the tusk after the animal died, but not whether it was hunted or scavenged.
The radiocarbon record was the most important part of the case. The team generated 71 radiocarbon dates, almost entirely from short-lived plant material like twigs, bark, and seeds, chosen specifically to avoid the old wood problem where dating ancient tree trunks can make layers look artificially old. The sediment layers themselves were undisturbed. A continuous shell-rich bed sealed the layer containing the stone tools, dated to roughly 14,400 years ago. A vertical column of 24 radiocarbon samples, some more than a meter above the artifact and some well below, all returned nearly the same age, roughly 14,550 years for the seven samples immediately surrounding the biface. A separate stone flake buried under more than a meter of undisturbed sediment dated to the same window.
Magnetic susceptibility readings and micromorphological studies showed no sign that the layers had been stirred or redeposited. Waters said that if people did not believe the site, they would not believe anything, citing four meters of undisturbed sediment and 71 agreeing radiocarbon dates.
The 2016 paper, published in Science Advances, drew strong support from peer reviewers. One called the case unassailable. A follow-up paper in 2025 added further analysis.
The site also raised a question about megafauna extinction. Microscopic spores from a fungus that grows almost exclusively in the dung of large herbivores showed humans and mastodons coexisted at the site for roughly 2,000 years before the mastodons vanished around 12,600 years ago. That timeline is a problem for the theory that human hunters wiped out American megafauna quickly after arrival. A fast kill-off does not fit a two-thousand-year coexistence.
Page-Ladson did not remain an isolated finding. Once its dating held up, it joined a list of other sites independently pointing to earlier human presence. In Chile, Monte Verde, excavated since the late 1970s, yielded wooden structures and tools dated to roughly 14,500 years ago. It was treated with suspicion for 20 years until a verification panel inspected the site in person in 1997 and endorsed the dates. Cooper’s Ferry in Idaho produced stemmed projectile points dating to between 15 and 16,000 years old. The Friedkin site in Texas, part of the Buttermilk Creek complex, yielded more than 15,000 artifacts beneath Clovis material, dated using optically stimulated luminescence. The Paisley Caves in Oregon preserved human waste dated to roughly 14,300 years ago. White Sands in New Mexico had fossilized human footprints dated to between 21,000 and 23,000 years, though that site has its own ongoing debate with no stone tools found alongside the footprints and questions about dating methods.
Individually, any one of these findings could be argued away. Together they form a pattern suggesting multiple groups arrived through multiple routes over different time periods. If the ice-free corridor only became usable around 14 to 15,000 years ago, sites older than that require another path, likely a coastal migration route along the Pacific, sometimes called the kelp highway.
In 2026, archaeologist Todd Surovell of the University of Wyoming published a redating study challenging Monte Verde itself, arguing the layer long dated to roughly 14,500 years is actually far younger, between 4,200 and 8,200 years old. Surovell acknowledged that even if Monte Verde falls, enough other pre-Clovis sites exist independently that the broader conclusion survives. He insisted each site must be judged on its own evidence.
Page-Ladson differs from Monte Verde in the density of its evidence. It rests on 71 radiocarbon dates, a sealed shell layer, and cut marks confirmed by forensic-level analysis.
For years, popular retellings circulated a dramatic claim that a drainage project threatened to destroy the sinkhole. That claim is not supported by the record. The actual situation is the opposite. In December 2025, a conservation easement covering 1,834 acres of the Ladson family’s land, including the sinkhole, was finalized, permanently protecting it from development. The easement is held by Tall Timbers Research Station and Land Conservancy, funded through Florida’s springs restoration program with contributions from the Knoblock Family Foundation and the Ladson family. The protected land includes 2.5 miles of the Aucilla River and nearly two miles of the Wacissa River. A related springs restoration project worth roughly $1.6 million was announced in March 2025 as part of a larger statewide effort to protect Florida’s spring systems.
The real pressure on the Aucilla is less dramatic but more chronic. Regional groundwater withdrawal, agricultural demand, and drought have lowered the water table in the region. The Water Management District monitors minimum flow levels to prevent significant harm to the aquifer.
The original tusk and stone artifacts now sit inside the Florida Museum of Natural History in Gainesville, part of an exhibit called Florida Fossils. Halligan has argued for a systematic survey of the wider Aucilla basin and other submerged Florida sites, noting that much of the earliest record of human life in the American Southeast is still underwater and unexamined.
The deepest unresolved question remains whether the people at the sinkhole hunted the mastodon or found it already dead. No scan or laboratory result will answer that if the evidence was never there. The tusk carries grooves that show human hands removed it, but the decision behind that act, whether a calculated hunt or an opportunistic use of a carcass, happened on one afternoon 14,550 years ago and cannot be recovered.
