How Hot Was Titanoboa's Swamp? Scientists Still Argue

PALEODEX · 28 August 2026 · 6 min read

A one-ton fossil snake was turned into a thermometer for a world 60 million years gone. The number it produced started a real fight in the pages of Nature — and nobody has fully won it.

In 2009, a team of paleontologists pulled vertebrae out of a Colombian coal mine, described the largest snake ever found in the fossil record, and then did something bolder: they used its size to calculate the temperature of the world it lived in. The number they got was hot — hotter than that same patch of Colombia runs today. The same year, in the same journal, two separate teams told them their method couldn't support that conclusion. Neither side backed down, and the argument is still open.

A snake as a fossil thermometer

Titanoboa cerrejonensis was described from vertebrae recovered at the Cerrejón open-pit coal mine in La Guajira, Colombia — about 13 metres long and roughly 1,135 kilograms, living in the late Paleocene, some 58 to 60 million years ago, not long after the asteroid impact that ended the non-avian dinosaurs (Head et al., Nature, 2009). That alone made it a record holder. What made it a bigger story is what the authors did next.

Snakes are ectotherms — their body temperature, and therefore how large they can safely grow, is bounded by the temperature of their environment. Run too hot metabolically for too cold a world, and a snake that size simply couldn't survive. So the team took Titanoboa's estimated mass, ran it through a body-size-to-temperature scaling model calibrated against the largest living boid alive today, the reticulated python, and calculated a minimum mean annual temperature for the Cerrejón tropics of 30 to 34°C.

Same year, same journal, two rebuttals

The estimate did not survive unchallenged — and not eventually, but immediately. Nature published two independent "Brief Communications Arising" responses that same year, each attacking a different weak point in the model.

Denny, Lockwood & Somero (2009) argued that a snake as large as Titanoboa could hold onto its own body heat through sheer thermal inertia, largely independent of the surrounding air. If that's true, the real Paleocene air temperature could have run several degrees cooler than 30–34°C without the snake overheating — the model, in their view, mistook the snake's internal temperature for the world's.

Sniderman (2009) came at it from a different angle: he applied the exact same size-to-temperature method to a different fossil giant, the Pleistocene lizard Varanus (Megalania) prisca, and found it predicted a body length inconsistent with that animal's well-documented climate. If the method over-predicts in a case we can already check, that's a reason to distrust what it says about a case — Titanoboa's world — we can't check any other way.

The reply, and a second, independent method

Head's original team didn't concede. In their own published reply, they defended the scaling model against both critiques, arguing the thermal-inertia effects the critics raised were already accounted for and that the critics had mischaracterised how the model worked. No retraction followed. No side conceded. The 30–34°C figure was never withdrawn, and the rebuttals were never fully answered away — a genuinely live, unresolved argument, not a closed one dressed up as open for drama.

A few years later, a completely different line of evidence weighed in — not on the snake's side or against it, but independently. Wing et al. (2009, PNAS) studied fossil leaves from that same Cerrejón Formation using leaf-margin physiognomy, a method that reads leaf shape as a temperature proxy with no relationship at all to reptile body-size scaling. Their estimate: roughly 24–31°C, depending on regression method — lower and wider than the snake-derived figure, but pointed in the same direction.

What the argument is actually about

It's worth being precise about what's contested here, because it isn't "was the Paleocene hot." Nobody disputes that — the snake-size method and the leaf method agree on direction, and so does essentially every other line of paleoclimate evidence for this period. What's genuinely unresolved is a narrower, more technical question: does a giant ectotherm's maximum size track ambient air temperature closely enough to back-calculate a precise number, or does thermal inertia at extreme body size break that assumption? That's a live methodological dispute, and it hasn't been settled — this is exactly the kind of "if it isn't provably true and fair, reframe it" honesty PaleoDex tries to hold data to across the whole catalogue.

It's also a useful reminder about how paleoclimate numbers actually get made. A "the world was X degrees" headline usually compresses a chain of assumptions, a calibration choice, and often a live argument other scientists are still having — see how differently that plays out when scientists try to time recovery after the extinction that came before Titanoboa's world existed. Read how PaleoDex sources every figure for the same reason: a number is only as good as the method that produced it, and the honest version of the story includes the argument, not just the answer.

Sources

Every figure above traces to a peer-reviewed paper — sources below.

• Head, J.J. et al. (2009), "Giant boid snake from the Palaeocene neotropics reveals hotter past equatorial temperatures," Nature 457
• Denny, M.W., Lockwood, B.L. & Somero, G.N. (2009), "Can the giant snake predict palaeoclimate?", Nature 460
• Sniderman, J.M.K. (2009), "Biased reptilian palaeothermometer?", Nature 460
• Head, J.J. et al. (2009), "Head et al. reply," Nature 460
• Wing, S.L. et al. (2009), "Late Paleocene fossils from the Cerrejón Formation, Colombia, are the earliest record of Neotropical rainforest," PNAS 106