The world changed forever some three million years ago, scientists have thought. At that time, for the first time, a ribbon of dry land connected North and South America, as the Isthmus of Panama shook free of the water around it. The new land bridge allowed plants and animals free travel between the two continents, colonizing new worlds. It also changed ocean currents and ushered in an ice age.

Now this textbook date is being challenged. The south-north connection happened around 15 million years ago and perhaps even earlier, not a paltry three million, argues Camilo Montes, a geologist at the University of the Andes in Bogotá. In a series of studies over the last four years Montes and a small but vocal group of evolutionary biologists and geologists have proposed this older date based largely on the age of rocks and fossils unearthed during the recent widening of the Panama Canal. “The work supporting this new date is beautiful. It’s just an airtight story,” says Peter Molnar, a geologist at the University of Colorado Boulder.

The older age is scientifically earthshaking. If the great ice age around three million years ago was not triggered by the isthmus, that means another, unknown factor played a major role in determining this dramatic natural climate swing. It also makes it harder to explain why some of the Americas’ most charismatic species—camel-like animals such as alpacas and llamas, adorable sloths and extinct giant terror birds—did not move between the continents until about three million years ago.

Because of the implications, Montes expected pushback. What he has gotten, however, is a fierce scientific battle. Anthony Coates, a paleobiologist at the Smithsonian Tropical Research Institute in Panama and defender of the previously accepted date—he did much of the research behind it—scoffs at the idea of a 15-million-year-old connection, calling the analysis “sloppy” and “shoddy.” He and a team of 30 other scientists threw down the gauntlet in a 2016 Science Advances paper, arguing the upstart researchers had paltry data to support their new theory. Montes, for his part, says Coates has yet to patch up holes in the old model opened by the new results.

Here’s what scientists agree on: When the great southern landmass known as Gondwanaland broke up around 180 million years ago, the plate of Earth’s crust carrying South America began its long drift from modern-day Africa to its current location. Fifty million years ago Central and South America were separated by hundreds of miles and a stretch of ocean at a depth of two kilometers. This deep Central American Seaway (CAS) created a sharp demarcation not only between the North and South American landmasses but also between their species. By 15 million to 20 million years ago, the CAS had become much shallower but still kept land apart. Geologic and fossil data Coates had gathered in the 1960s and ‘70s indicated Panama did not completely bridge the gap until about three million years before today.

In 2008 Montes received a $1-million research grant from the Panama Canal Authority as a newer, deeper canal was being built to allow the passage of ever-larger cargo ships. Montes wanted to refine the dates of the rocks in the area using magnetic and chemical analyses to provide more precise dates for fossils. He fully expected the dates to converge around the accepted age of the isthmus closure—except they did not. Instead, he found the rocks were 10 million to 15 million years of age.

Puzzled at the discrepancy, Montes turned to a type of crystal called a zircon to resolve the issue. Zircons often trap and preserve small amounts of radioactive uranium at their cores when they are formed. This uranium decays at a steady rate, like the ticking of a clock, making it perfect to date old rocks. Using samples from the San Blas Mountains, which stretch along Panama’s northeastern coast and into Colombia, Montes showed this volcanic mountain range began forming some 70 million years ago and had helped to close the isthmus no later than around 10 million years ago. He presented his initial findings that the formation of the isthmus was older and far more complicated than previously thought at The Geological Society of America in 2012.

New fossil data provided additional support for Montes’s hypothesis. Evolutionary biologists Christine Bacon at the University of Gothenburg in Sweden and Carlos Jaramillo at the Smithsonian Tropical Research Institute analyzed the evolutionary trees of 3,589 species found in both North and South America that crossed the isthmus. The data, published in Proceedings of the National Academy of Sciences in May 2015, indicated several major movements across the isthmus, including one as far back as 23 million years ago. To Bacon, species traveling from north to south contributed to South America’s tremendous biodiversity, and if organisms began traveling earlier than expected, this biodiversity may have taken much longer to arise and evolve than previously thought.

Coates does not believe Montes’s zircon data are sufficient to rewrite history. Neither are Bacon’s fossil data, he says. Species cross open ocean all the time. It’s how islands like Hawaii and those in the Caribbean were populated. And even if there was an early trickle of migration, Coates’s fossil data indicates large mammals like sloths and llama ancestors only began crossing the isthmus en masse around three million years ago. To gainsay Montes and Bacon, Coates has pointed to marine fossils showing identical small creatures such as foraminifera (tiny unicellular, shelled organisms) can be found in both the eastern Pacific and the Caribbean until about three million years ago. This indicated a stretch of open ocean existed until that date, allowing the foraminifera to pass unimpeded. “Plenty of plants and animals can cross open ocean, but marine life can’t just walk across dry land,” Coates says.

Coates and some other geologists believe what is now modern-day Panama may have begun to peek above the waves early on, but it remained a series of islands until about three million years ago, when the isthmus finally closed. Coates compares this process with what is currently happening in Indonesia. The Australian Plate currently abuts the Eurasian Plate, but the two landmasses were not always so close and apparently kept animal species apart. The plates of crust are still moving, and scientists predict the islands dotting part of Indonesia will eventually form a stretch of dry land, much like the Isthmus of Panama did several million years previously. It is the Indonesian model that Coates and 35 other authors cite in their August 2016 Science Advances rebuttal to the work of Montes and colleagues.

The Montes model does leave some questions unanswered, says Andrew Crawford, a biologist at the University of the Andes, although he says he is “leaning towards” the older date. Crawford has proposed a hypothesis in a 2016 Geology study, co-authored by Bacon, that could account for the both the early isthmus and the late arrival of many animals. Before the ice age what is now Central American jungle may have been much drier savanna, which may have interfered with the movement of large animals because they were unable to adapt to the dry conditions and find suitable food. “Frogs can physically climb mountains, but they won’t because of environmental barriers like rain and cold,” Crawford says, arguing environmental barriers may have interfered with species exchange.

Then there is the mystery of the three-million-year-old ice age. Researchers have generally believed the closure of the CAS shifted ocean currents, which changed weather patterns by bringing moist tropical air to the Arctic via the Gulf Stream. This air dropped its moisture in the frigid northern latitudes, which became the basis of massive sheets of ice that ultimately covered much of the Northern Hemisphere. But if the seaway between Central and South America was blocked several million years before this ice age started—as Montes argues—scientists have to find another trigger. Molnar says geologists and oceanographers have proposed a variety of hypotheses, although none are considered a definitive explanation.

“Instead of disagreeing, we should be out there gathering more data,” Bacon says. The continents she studies did eventually merge into a unified mass, but it is anyone’s guess as to whether, or when, scientific understanding of the isthmus will have the same outcome.