Throat Bones Get a Head Start on the Mammal Milk Run

A 165-million-year-old fossil from China preserves throat bones that researchers say show the swallowing and suckling machinery of living mammals had already evolved in a Jurassic relative of mammals ^[2][3].

The animal is Megacauda sungei, described in Nature on October 7, 2026 by a team led by Y. Li, P. Li, A. I. Neander, H. Zhang, C.-F. Zhou, T. Martin and Z.-X. Luo ^[2]. It is not a mammal in the strict sense. It was a docodont, a member of an extinct lineage of mammaliaforms, the group of animals that came before modern mammals ^[2][3].

That distinction is the point. If the authors are right, the feeding system we associate with mammals was assembled before mammals themselves.

Suckling is among the defining acts of mammal life, and it is almost impossible to catch in stone. A Jurassic world ruled by dinosaurs left few clues about how the small animals at their feet fed their young. A throat bone is one of the few places such a clue can survive.

What the bones show

In a living mammal, a handful of small bones and muscles in the throat do quiet, constant work. The hyoid bones and related muscles help move food and liquid through the throat, and the soft palate helps keep the passages of the mouth and nose apart. Together they allow safe swallowing and suckling.

Megacauda preserves a therian-like bony structure in the throat passage, along with hyoid bones ^[2]. Therians are the group that includes marsupials and placental mammals. From that anatomy, the researchers infer that the soft-tissue structures behind mammal-like swallowing, drinking and suckling had already evolved in ancestors shared by docodonts and mammals ^[2].

The fossil was examined with CT scanning, which let the team reconstruct the skull and throat structures in detail ^[3]. A report on the study from Phys.org adds a telling detail: the modern-looking bony hooks sit alongside more primitive bony ridges. The researchers read that as a sign that the modern structures did not grow directly out of those ridges, as had been thought ^[1].

What it does not show

Nobody has watched a Jurassic youngster feed. The fossil preserves neither a baby nor the soft tissues, only the bones that anchor them. Suckling is an inference from that anatomy and from comparison with living mammals ^[2][3].

The evidence is a detailed anatomical reading of one fossil, not a direct record of behavior ^[2][3]. It is a strong case, built on a well-preserved specimen, but it cannot show how suckling worked in life.

The finding also says nothing about when milk first appeared. It concerns the evolution of a feeding system, and the paper's own language is cautious about the soft-tissue anatomy and behaviors the fossil suggests ^[2].

A swimmer, and a big one

The fossil comes from the Middle Jurassic of Inner Mongolia ^[2][3][4]. The animal was about 40 to 50 centimeters long and weighed one to two kilograms ^[3][4]. That is small by human standards and large for its group.

The researchers interpret it as semiaquatic. It had a flattened tail, along with hind limbs and a pelvis suited to swimming, and its teeth suggest it ate animal prey ^[2][3]. The University of Chicago's account calls it a Jurassic otter ^[3]. The University of Bonn describes a top predator that measures only half a meter ^[4].

That sits awkwardly with the familiar picture of mammal relatives as uniformly tiny creatures scurrying in the shadow of dinosaurs. Docodonts were more varied than that. Megacauda adds to the evidence that they occupied different ecological niches, including aquatic and tree-dwelling ways of life ^[2][3][4].

Not the first hint

The idea is not new, but the evidence is better. In 2019, another Jurassic docodont, Microdocodon gracilis, was described with mammal-like hyoid bones ^[2][3]. That was a tantalizing sign that the throat of a mammal relative could already look modern.

Megacauda goes further. It preserves better structures beneath the skull and adds evidence about the palate and the pharynx, the passage behind the mouth ^[2][3]. Two animals from the same lineage now point the same way.

That matters for how we tell the story of mammals. It is tempting to treat the arrival of the mammalian way of life as a late event, close to the origin of true mammals.

This fossil suggests a long run-up. The throat of a swimming, meat-eating docodont had already been rebuilt along mammalian lines, long before the animals we would call mammals.

It is also a reminder of how much a single bone can carry. Suckling leaves no trace in rock. The structures that make it possible do, and they are what the researchers read.

What comes next

The case rests on one named specimen, the holotype PMOL-AM00026A/B, held by the Paleontological Museum of Liaoning in Shenyang, China ^[2]. That makes the open questions plain.

How widespread was this throat anatomy among docodonts and other mammaliaforms? Was Megacauda typical, or one branch's own solution? And how precisely can the timing be dated, given that the paper places the anatomy in ancestors shared by docodonts and mammals ^[2]?

The paper does not announce a next research step ^[2]. Answers will depend on more fossils from the same lineage, preserved well enough to show the same bones.