NEW YORK, NY.- Spinosaurus was one of the largest carnivorous dinosaurs, and it ate fish. That much paleontologists agree on.

But did it just wade into rivers and snatch them out of the water like a grizzly bear? Or did it dive after its prey like a penguin or a sea lion?

This has become a question of prickly contention among dinosaur experts.

One group is increasingly convinced that Spinosaurus was a rarity among dinosaurs: one that stuck its head underwater and swam beneath the surface. Others say no way.

The latest salvo, published Wednesday in the journal PLOS One, comes from the Spinosaurus-couldn’t-swim contingent to counter a pro-swimming paper published a couple of years ago. The earlier work, published in the journal Nature, claimed that in general, animals that spend much of their time in the water, like penguins, have denser bones that provide ballast and make it easier to dive. Spinosaurus also had dense bones, and therefore was most likely a swimmer, the Nature paper concluded.

But that bone density analysis was “statistically absurd,” said Nathan Myhrvold, a former chief technology officer at Microsoft and an amateur paleontologist who led the new research with Paul Sereno, a paleontologist at the University of Chicago.

Myhrvold and Sereno have also argued that the ungainly body shape of Spinosaurus would have made it a poor swimmer, if it could swim at all. The weight distribution of the dinosaur would have made it top-heavy and unstable, Myhrvold said.

“It’s obvious as to why it can’t swim,” he said.

The giant sail on its back would make it difficult for a swimming Spinosaurus to stay upright, Myhrvold said. “If it tips even the slightest amount, it’ll keep tipping.”

In other words, Spinosaurus would capsize and struggle to pull its sail out of the water.

In this dispute, there are points of agreement. Spinosaurus was perhaps longer and heavier than Tyrannosaurus rex. It lived about 95 million years ago in what is now the Western Sahara but was then a lush environment with deep-flowing rivers. It was also an odd-looking dinosaur, with elongated vertebrae forming a huge sail on its back.

There has been a burst of interest in Spinosaurus in the past decade after a new fossil was uncovered in Morocco by Nizar Ibrahim, who was also an author of the earlier bone density study and is now a senior lecturer at the University of Portsmouth in England. The only other fossil was found by Ernst Stromer, a German paleontologist, in 1915, and destroyed in an aerial bombing of Munich in 1944.

In the latest study, Myhrvold and colleagues argue that the paleontologists who made the bone density claims employed a sophisticated statistical technique without understanding its limitations.

“It’s totally misapplied here,” Myhrvold said. “Unfortunately, when you have something that involves lots of dense statistics, most paleontologists’ eyes glaze over.”

Myhrvold is not a traditional academic. Since departing Microsoft in 1999, he is perhaps best known for leading the development of the encyclopedic Modernist Cuisine cookbooks. But he has stirred up esoteric statistical scuffles before, criticizing findings about the growth rate of dinosaurs and claiming that a NASA trove of asteroid data is flawed and unreliable.

Earlier work by other researchers had found that diving mammals tended to possess denser bones than mammals that stayed on land. But other mammals also have dense bones for other reasons. Elephants, for one, need stronger bones to support their weight.

In 2022, researchers led by Matteo Fabbri, now a postdoctoral researcher at the University of Chicago, argued in their paper that bone density was a reliable predictor of whether an animal lived in the water or on land for a much broader swath of creatures, including extinct species.

“We thought, Oh, is this just mammals and or is this also reptiles?” Fabbri said in an interview. “And if this is true, can we infer ecology in extinct animals, including weird-looking dinosaurs like Spinosaurus?”

Fabbri said the analysis showed that “very high bone density is correlated with the probability of going underwater.”

Spinosaurus and a Baryonyx, a relative of Spinosaurus, did dive, while another related dinosaur, Suchomimus, did not go underwater, the team of scientists concluded.

However, Myhrvold argues that bone density does not neatly divide into two groups. There are many aquatic animals with bones less dense than many land animals and vice versa. “If the two distributions are close together, you can’t get a valid conclusion, or at least one that has any statistical strength,” he said.

He gives an example: in humans, men are generally heavier than women, but not every man is heavier than every woman. Thus, if someone told you that a person weighs 135 pounds, you could not reliably deduce whether that person is male or female.

Although Myhrvold and Sereno are now at odds with Fabbri and Ibrahim, they were all once on the same side as co-authors of the 2014 paper that described the Spinosaurus uncovered in Morocco.

“We split intellectually,” Sereno said.

Fabbri is in the same department as Sereno, although he will become a professor at Johns Hopkins University this summer.

“We say hi in this corridor,” Fabbri said. “It’s OK. We are not killing each other, obviously.”

Ibrahim, who is in Morocco conducting additional studies, said further findings would make an even more convincing case that Spinosaurus was aquatic.

He also dismissed Myhrvold’s biomechanical arguments for why Spinosaurus could not swim, saying much is still unknown. He compared Myhrvold’s findings to paleontologists who argued that Tyrannosaurs must have been scavengers because they could not run fast enough to catch small, quick prey. But Tyrannosaurs did not have to be fast to pull down a big, slow-moving Triceratops.

Similarly, the prehistoric African rivers were filled with giant, slow-moving fish, Ibrahim said. Spinosaurus would not have had to be a proficient swimmer to catch them.

“I can’t reveal too much,” he said. “But we have new material. We have several very exciting ongoing projects.”

This article originally appeared in The New York Times.