Fossils document how dinosaurs gave rise to birds

Everyone knows that dinosaurs are extinct. As children, many of us gazed in awe at the fossils of these magnificent beasts. As adults, a lot of us still do!

Except that dinosaurs aren’t extinct, at least based on the most recent interpretations of the fossil record and analyses of DNA. The collective evidence points to a conclusion that once seemed improbable: birds are dinosaurs. I still recall the awe and wonder that beheld me when I learned this in college, and when I’ve taught it to children (all of whom are bonafide dinosaurs experts), I can see the same amazement on their faces.

So what evidence is there in the fossil record for this supposed ancestor – descendant relationship?

Prior to the Triassic, there were a lot of reptilian looking animals that no longer exist today. One such animal, Protorosaurus, has been found in rocks dating to about 260-251 million years ago (Ma). Typical of its contemporaries, it walked around on four limbs, each of which terminated in five fingers or five toes, possessed a long tail and teeth, and was almost certainly covered in scales. At this point in time, Protorosaurus and its fellow reptiles had seemingly little in common with today’s birds.

Not too much later, about 245 million years ago, animals like Asilisaurus appear in the fossil record. Though this species likely walked on all fours, it had shorter arms, suggestive of an increased ability to walk and/or stand on its hindlimbs.


In another nine million years, we see animals like Marasuchus (236-234 Ma): clearly reptilian in form, very dinosaur like, and notably bipedal, just like birds.


Eodromaeus and its kin are among the earliest true dinosaurs, popping up a mere five million years after Marasuchus (231.4-229 Ma). One of its typical dinosaurian traits is a hip socket with a hole in it (perforate acetabulum). What’s additionally notable about this species and some of its contemporaries is how its fingers have changed. Modern birds do not have fingers, but their wing bones terminate in what appears to be the remnants of three fingers. Starting this trend toward digit reduction, Eodromaeus has five fingers, but the ring and pinky are very reduced in size.

Fast forward about 30 million years, and we have more modern-looking dinosaurs on the scene. Coelophysis (203-196 Ma) is typical of the early carnivorous dinosaurs (theropods), and continues the march towards birdiness. The pinky finger is practically non-existent at this point, and the toes have also reduced in number. Whereas earlier dinosaurs and other reptiles have five toes, Coelophysis has only four, with a tiny remnant of the fifth high up on the foot. Notably, birds have four toes, three in the front and one in the back, so this trait had already appeared at least 200 million years ago.

Whereas Coelophysis had four fingers on its hands, Sinosaurus (201-196 Ma), appearing two million years later, only has three fingers, having lost the ring finger altogether.


While other bird-like traits accumulated over time, perhaps the most significant change is found in rocks that date to 50 million years after dinosaurs like Sinosaurus. Archaeopteryx (150.8-148.5 Ma), the first documented transitional fossil, has a striking mix of bird- and reptile-like traits. Perhaps most significantly is the appearance of feathers in this species.

NGS Picture ID:422890

There is reason to think that feathers appeared prior to Archaeopteryx, however. Soft tissues like feathers don’t typically preserve well as fossils, but over the last 20+ years, a number of exceptionally preserved specimens have demonstrated that plenty of non-flying dinosaurs had feathers. This suggests that these structures didn’t appear for the purpose of flight, but rather for a simpler function, such as thermoregulation. Just like hair keeps us and other mammals warm, feathers provide a similar insulating layer for birds.

Archaeopteryx clearly had wings, though, suggesting some ability to glide or perform flapping flight. It had another bird-like feature, known as the furcula. Also known as the wishbone, the bone frequently broken apart as a Thanksgiving dinner ritual in the United States, this bone formed by the fusion of the two clavicle bones present in earlier species like Coelophysis. This fusion is thought to be important in withstanding the stressors caused by flight, but it is also present in many other carnivorous dinosaurs, suggesting it initially appeared for a different reason.


Despite these innovations, Archaeopteryx is still not as bird-like as you’d think. For one, it still very clearly had three fingers, along with claws, on its hand. It also had a long bony tail, a structure reduced to a nub (pygostyle) in modern birds. Last but not least, Archaeopteryx had teeth, whereas birds have lost their teeth entirely in favor of a beak made of a protein called keratin.

After another 15 million years, other dinosaurs, like Sinornis (135 Ma), were just a few steps away from modern birds. The tail bones had finally become shortened and fused into a pygostyle in Sinornis, providing a structure tail-feather attachment. The sternum also became keeled increasingly keeled, allowing for the attachment of more powerful flight muscles, suggesting that Sinornis was a better flyer than Archaeopteryx. Despite these innovations, Sinornis still retained teeth and three distinct clawed fingers, traits that are not present in modern birds.


40 million years later, when tyrannosaurs and pachycephalosaurs were still roaming the earth, some extremely bird-like animals appear in the fossil record. At first glance, Ichthyornis (93-83.5 Ma) may be difficult to discern from a modern seabird. Many of the limb bones, including the fingers have become fused in Ichthyornis, resulting in a skeleton that was probably well-adapted for powered flight. Nonetheless, as one of the final holdouts of its reptilian past, Ichthyornis still had teeth. Interestingly, however, its jaw tip appears to be covered by an incipient beak, suggesting the transformation is nearly complete.


After most dinosaurs disappeared from the fossil record, coinciding with geological evidence for a disastrous meteor impact and intense volcanism, birds persisted and began to really thrive. Among the earliest species is Waimanu (60 Ma), which appeared just five million years after this mass extinction event. Not only does Waimanu have the appearance of a modern bird, paleontologists think it was the earliest known penguin, suggesting the surviving bird species have already begun to resemble some of their modern forms.


Over a period of about 200 million years, the fossil record appears to document the gradual appearance birds from reptilian forebears. With the advent of bipedalism, the reduction in fingers and toes, derivation of feathers and wings, reduction of the tail, and loss of teeth, fossils seem tell the story that birds are, in fact, dinosaurs. Do not hesitate to remember this the next time you feed a duck or eat a chicken wing!

Question for Creationists

Where did all of these fossil animals go? Could they not fit on Noah’s ark? Wouldn’t species with some capacity for flight, such as Sinornis, Archaeopteryx, and Yi qi, be able to avoid the Flood? Why does the fossil record appear to document a transition between reptilian animals and birds? If Noah’s Flood is responsible for the placement of these fossils, why did they appear in this particular sequence? Is it just a coincidence that the fossil record appears to document birds descending from ancestors with teeth and birds have remnants of tooth genes in their genomes?

Photo credit

Protorosaurus, chickenAsilisaurus, Marasuchus, Eodromaeus, Eodromaeus handCoelophysis, Coelophysis handSinosaurus, Archaeopteryx, Sinosauropteryx, Beipiaosaurus,  Sinornithosaurus, Microraptor, Psittacosaurus, Epidexipteryx, Similicaudipteryx, Anchiornis, Changyuraptor, Yi qi, furcula, Sinornis, Waimanu


Manatees that walked on land

Manatees and the dugong, known collectively as sirenians, are herbivorous, aquatic mammals that are generally restricted to shallow, tropical waters. Like whales, sirenians have reduced the amount of hair on their bodies, have forelimbs shaped into flippers, lack external hindlimbs and have a paddle-like tail.

Despite their overall similarities, the DNA of sirenians is much more similar to that of elephants and the DNA of whales is much more similar to hoofed mammals like cows and giraffes. This suggests that both sirenians and whales independently adapted to living in an aquatic medium. In fact, the fossil record of whales suggests a transition from a land-based habitat to a completely aquatic way of life. Do sirenians show the same pattern?


Indeed they do! One of the earliest sirenian species appearing in the fossil record is Pezosiren, a mammal with an unmistakably sirenian-like skull, but the rest of its skeleton clearly suggests that it had the ability to walk on land. Unlike modern sirenians, this animal had forelimbs that are not shaped like flippers but instead were used to support their weight on land, as well as hindlimbs and a distinct hip bone (pelvis). Pezosiren has been found in rocks estimated to be ~47.8 million years in age, similar to the age that walking ‘whales’ were also found roaming the earth.


More recent sirenian fossils show evidence of fully committing to an aquatic lifestyle, including Halitherium, which is found in rocks estimated to be ~38 million years old. This species and similarly dated fossils had modified their forelimbs into flippers, possessed thick and dense ribs to provide ballast, and reduced their hindlimbs, presumably to minimize drag. Though it looked very much like modern manatees (see below) and dugongs, it had a better-developed, albeit very reduced, pelvis+hindlimb complex.


In addition to their better-developed limbs, early species of sirenians had a number of other traits that manatees and the dugong have lost. Pezosiren and others had multiple vertebrae in the hip region (sacral), much like their land-dwelling elephant relatives, whereas most fossil and all modern sirenians have fused these into a single bone. Modern sirenians also have reduced their teeth to a great extent compared to their fossil forebears. Earlier species had permanent premolars, canines and incisors, whereas modern species lack these teeth (dugongs have only one incisor). See below for a diagram of a fossil sirenian (Protosiren, ~47.8 million years old) showing the positions of the premolars (P), canine (C) and incisors (I).


Questions for Creationists

Where did the dozens of other sirenian species go? Given their aquatic adaptations, wouldn’t they have survived Noah’s flood? Is it just a coincidence that molecular phylogenetics shows sirenians as being genetically similar to land mammals like elephants and there are sirenian fossils that appear to document a transition from land to water? Is it coincidence that we see parallel evidence of a land to water transition in whales?


1. Springer, M. S., Signore, A. V., Paijmans, J. L., Vélez-Juarbe, J., Domning, D. P., Bauer, C. E., … & Meredith, R. W. (2015). Interordinal gene capture, the phylogenetic position of Steller’s sea cow based on molecular and morphological data, and the macroevolutionary history of Sirenia. Molecular phylogenetics and evolution91, 178-193.

Photo credit

Manatees, dugong, Pezosiren, Halitherium, manatee skeleton, Protosiren

The earliest platypuses: bigger and toothier


Platypuses are strange creatures, even to evolutionary biologists. They have hair, webbed feet, a beaver-like tail, and duck-like bills with which they can sense electrical currents, they lay eggs, and the males possess venom glands on their hind limbs. When a drawing and a pelt of the animal were first sent to British scientists at the end of the 18th century, the incredulous naturalists assumed it was a hoax. In college, I had a t-shirt that made light of their bizarre features:


Given the idiosyncrasies of this animal, you might imagine that they’re a bit of an enigma to evolutionary biologists. In fact, I recall a creationist acquaintance of mine once declaring that they must be an “evolutionist’s worst nightmare”. To be sure, there are a lot of gaps in our knowledge regarding how they might have evolved. What the evidence points to pretty clearly is this: 1) they’re mammals, as they have hair, produce milk for their young, and their DNA is much more similar to other mammals than it is to, say, birds or lizards, and 2) they split from other mammals a long, long time ago, somewhere on the order of 220 million years. The fossil record of their relatives, known as monotremes, is extremely spotty, with the earliest dating to over one hundred million years ago. Unfortunately, the specimens are extremely fragmentary, frequently consisting of only part of a jaw or arm, preventing researchers from making many firm conclusions about their evolution.

However, at least one fossil animal gives some insights into their ancestry. Obdurodon includes several species of platypus-like monotremes that are estimated to have lived ~28.1 to 5.3 million years ago. Below you can see a comparison between Obdurodon dicksoni (left; 23-11.6 million years ago) and the modern platypus (right).


You’ll notice differences between the two, but overall they look extremely similar. Clearly the duck-like bill has been around for a while! However, at least one feature  points to Obdurodon being a transitional fossil, namely the fact that it has teeth. Modern platypuses have teeth when they are very young, but shed these by adulthood and replace them with horny pads (see below).


Considering these horny pads are a very uncommon feature in mammals, and vertebrates as a whole, biologists interpret it as an evolutionary novelty. With Obdurodon, we have support for this hypothesis, with evidence that a toothed ancestor preceded the modern platypus dental condition.

Questions for creationists

Where did the other platypus species (i.e., Obdurodon) go? Could Noah not fit them or their eggs on his ark? Is it a coincidence that there are fossil platypuses that had teeth as adults, much like typical mammals, but modern platypuses replace them during development with horny pads?


1. Musser, A. M., & Archer, M. (1998). New information about the skull and dentary of the Miocene platypus Obdurodon dicksoni, and a discussion of ornithorhynchid relationships. Philosophical Transactions of the Royal Society of London B: Biological Sciences353(1372), 1063-1079.

Photo credit

platypus, t-shirt image, platypus Obdurodon comparison, horny pads

How the ankylosaurs got their club-like tails

Ankylosaurs were a group of herbivorous dinosaurs that were covered with bony armor composed of osteoderms (bones formed in their skin), much like the mammalian glyptodonts. Many of the ankylosaurs had osteoderms flanking the tips of their tails, making a club-like structure that may have been used to ward off predators or compete for females.


A recent paper [1] by ankylosaur expert Victoria Arbour showed that these ornate creatures appear to have evolved these club-like tails in a two-step transition.

Scelidosaurus 199-191 million years ago*


Before talking about ankylosaurs, it’s first important to mention Scelidosaurus, a precursor of sorts to the ankylosaurs. The earliest dinosaurs were bipedal, but Scelidosaurus was quadrupedal, walking on all four legs. Also unlike earlier dinosaurs, Scelidosaurus possessed osteoderms, although not the large, fused types characteristic of ankylosaurs.


Gastonia 127-124 million years ago


After Scelidosaurus disappeared, the ankylosaurs appeared in the fossil record. They had much more elaborate osteoderms that had fused into plates and spines. The earliest fossils, such as Gastonia here, had flexible tails, often flanked with spikes, but lacked the mallet shape characteristic of later species.

Gobisaurus 93-90 million years ago


Gobisaurus differs from earlier ankylosaurs in that it had a very stiff tail, achieved by having highly overlapping vertebral joints. These fossils have not been discovered with the tail clubs characteristic of the latest ankylosaurs, such as Euoplocephalus (78-77 million years ago) shown below:


This suggests that a stiffened tail, which would have been strong enough for striking predators or opponents, predated the clubs of later ankylosaurs. The clubs then appeared at a later stage, presumably to amplify the power of this putative weapon.

To further illustrate this transition, here is a diagram created by Victoria Arbour**:


*Dates are estimates from fossil distributions reported in figure 4 of Arbour et al. [2015]

**Note that she highlights different ages than I, due to her summarizing of multiple fossils

Questions for Creationists

What happened to the ankylosaurs? Why does the fossil record appear to record a transition from dinosaurs that were bipedal and lacked osteoderms, to quadrupedal species with elaborate osteoderms and club-like tails? Is it a coincidence that these species appear in this order?


1. Arbour, V. M., & Currie, P. J. (2015). Ankylosaurid dinosaur tail clubs evolved through stepwise acquisition of key features. Journal of anatomy227(4), 514-523.

Photo credit

Ankylosaurus by Raul Martin, Scelidosaurus 1 by FunkMonk, Scelidosaurus 2, Gobisaurus by Sydney Mohr, Euoplocephalus, Ankylosaur transition by Victoria Arbour

How the turtle got its shell

Turtles are unique among reptiles in that they have a large shell, composed of a carapace on their back and a plastron on their belly. This shell largely develops from fused ribs and bones derived from the skin. Since turtles presumably evolved from a lizard-like ancestor to become the distinctively shelled creatures we know of today, we might expect that there would be transitional fossils that illustrate this link.

Pappochelys 242-237 million years ago


At first glance, Pappochelys [1] might not look very turtle-like, as it had no shell. However, it did have what appears to be the beginnings of a shell. Specifically, it had very broad t-shaped ribs and little bones in its belly region known as gastralia. Beyond these, Pappochelys had various other features that it shares with other turtles, such as a process coming off of the pubis, a femur with an offset head, and a modified shoulder blade.


The overhead view of the reconstruction above gives the impression of a lizard-like animal with the beginnings of a shell.

Odontochelys 237-227 million years ago


Not too long after Pappochelys appears in the fossil record, we find Odontochelys [2]. This extinct species apparently fused those belly bones (gastralia), making up a plastron that is very similar to that of modern turtles. In other words, this turtle only had half of a shell (the bottom half). In addition to this new feature, Odontochelys had a shorter tail than Pappochelys, a larger forelimb to hindlimb ratio, new bones called neurals, and it lost two holes in the sides of its skull that formerly allowed jaw muscles to bulge out during chewing.

Proganochelys 227-208.5 million years ago


Next to appear in the fossil record is Proganochelys, a species very close to modern turtles in overall form. It has a complete shell, with both the lower plastron and upper carapace. Additionally, whereas Pappochelys and Odontochelys had typical reptilian teeth on the margins of its mouth, Proganochelys lost these and probably replaced them with a keratinous beak. However, Proganochelys did retain teeth on its palate, as well as other traits related to the skull, shell, shoulder and pelvic girdles that are not found in turtles today.

Notably, this overall pattern of tooth reduction in turtle history (i.e., marginal teeth in Pappochelys -> marginal+palatine teeth in Odontochelys -> palatine teeth in Proganochelys -> no teeth in modern turtles) is paralleled by genetic data. Specifically, turtles retain remnants of tooth genes, suggesting that they formerly possessed teeth.

Questions for Creationists

Where did the turtles without shells and the turtles with half shells go? Or even turtles like Proganochelys, which had palatine teeth? Some of these species, such as Odontochelys, were likely aquatic, so should they not have survived Noah’s flood? Is it a coincidence that the species found in lower (=older) rocks are less like modern turtles than the species in higher (=younger) rocks? Is it coincidence that the reduction in teeth in the turtle fossil record mimics the loss of tooth genes in modern turtles?


1. Schoch, R. R., & Sues, H. D. (2015). A Middle Triassic stem-turtle and the evolution of the turtle body plan. Nature.

2. Li, C., Wu, X. C., Rieppel, O., Wang, L. T., & Zhao, L. J. (2008). An ancestral turtle from the Late Triassic of southwestern China. Nature456(7221), 497-501.

Photo credit

Rainer Schoch 1Rainer Schoch 2,  Li et al. 2008, Claire Houck 

Way back when snakes had legs

As I recently discussed, snakes are genetically nested within lizards. One of the major evolutionary implications of this fact is that snakes used to have legs. Just this week, researchers [1] heralded the discovery of a major fossil, helping to bridge the putative transition from legged-lizard ancestors to modern legless snakes.

Tetrapodophis 125-113 million years ago

Tetrapodophis whole

A casual glance at Tetrapodophis reveals that it has a very snake-like body, possessing a typical snake-like count of over 150 vertebrae before the tail. To a trained morphologist, there are also a number of other features that are notably serpentine including recurved teeth and an intramandibular joint to allow for the widening of the gape, among others.

The most noteworthy feature, however, is the retention of all four limbs.

Tetrapodophis forelimb

Above is a forelimb

Tetrapodophis hind

and here are the hindlimbs.

Eupodophis 101-94 million years ago

After the appearance of Tetrapodophis, there are at least four known species of snakes that show evidence of a major modification compared to their predecessor: they have completely lost their forelimbs, but still retain hindlimbs. In Eupodophis, the example species I highlight here, the hindlimbs are further reduced, eliminating the foot bones altogether.

Eupodophis limb

By contrast, modern snakes of course do not have legs. Some species, such as pythons and boas, have pelvic spurs, apparent remnants of the pelvis and femur, but the transition from legged to legless lizards appears to be mostly complete.

Questions for Creationists

Is it possible that God created snakes with legs? Where did they go? Is it a coincidence that the species with four legs appears in rocks that are older than the species with two legs, and before any snake species that have no legs? Is it also a coincidence that snakes are genetically nested within legged lizards?


1. Martill, D. M., Tischlinger, H., & Longrich, N. R. (2015). A four-legged snake from the Early Cretaceous of Gondwana. Science349(6246), 416-419.

A bat with claws

Bats are unique among mammals in that they are capable of powered flight. While we do not have evidence of flightless bats, we do have fossils of bats with primitive features. The oldest known species* is Onychonycteris finneyi, which Nancy Simmons and colleagues discovered in Wyoming rocks that date to approximately 52.5 million years ago [1].


If you’ve ever looked at a bat’s wings in detail, you’ll notice that the scaffold for the wing membrane is actually a hand, with very long fingers extending to the edge.


Despite these being fingers, bats only have claws on their thumbs, with the exception of some old world fruit bats which also have claws on their index fingers. Onychonycteris is more similar to earlier, non-flying mammals in that it possessed claws on every one of its fingers, hence it’s name: “clawed” (Onycho) “bat” (nycteris).

Another transitional feature involves the relative sizes of its limb bones. Bats are typically characterized by having particularly long forearms and legs that are short relative to their arms, features thought to be adaptations for flight. Just by comparing the photos above, it should be apparent that Onychonycteris is not proportioned the same as modern bats. Simmons et al. [1] compared its limb proportions to modern bats and various non-flying, tree-dwelling species, and they discovered that Onychonycteris was proportioned much more like non-bats than bats (see below). nature06549-f3.2

Though Onychonycteris was almost certainly capable of powered flight, it may have been less adept at flying than modern bats and perhaps spent ample time climbing trees as opposed to just taking to the air.

In short, Onychonycteris, the earliest known bat, had some traits that are more similar to non-flying mammals than bats, consistent with the hypothesis that bats evolved from non-flying species.

*Another bat, Icaronycteris, is approximately the same age, but it has more modern features than Onychonycteris.

Questions for Creationists

Why would God create most bats with only a claw on the thumb, but give Onychonycteris claws on every finger? Is it just a coincidence that this fossil was found in rocks that correlate with 52.5 million years ago and it has some features that are more similar to other mammals than bats? If it could fly, shouldn’t it have survived Noah’s flood?


1. Simmons, N. B., Seymour, K. L., Habersetzer, J., & Gunnell, G. F. (2008). Primitive Early Eocene bat from Wyoming and the evolution of flight and echolocation. Nature451(7180), 818-821.

When whales walked: Fossils document the rise of the largest animals on earth

Whales provide some of the best evidence that evolution has occurred, with DNA, development, biogeography, and the fossil record all telling the same story: that whales once walked on land.

Now before you start imaging a humpback whale or a bottlenose dolphin with legs, let me back up and clarify. Whales, as we now know them today, are extremely well-adapted to living in an aquatic environment. So much so, that if you took them out of the water for too long, they would certainly die. But scientists think that they had ancestors that could leave the water, much like seals, and even more ancient ancestors that lived completely on the land.

Whales are, after all, mammals, not fishes. They produce milk for their young and have hair (albeit very little). Since the vast majority of mammals are land dwelling, it stands to reason that if evolution is true, then these denizens of the sea likely descended from mammals that walked on land. In fact, when scientists compare their DNA to that of other animals, we find that they’re genetically most similar to hoofed mammals, such as cows, deer, hippos and llamas (Cetartiodactyla).

So are there any fossils that document this supposed transition from a hoofed mammal to the massive blue whale? As illustrated in the figure below [1], there most certainly are.


Let’s unpack this illustration. At the bottom is an animal called Indohyus, a small, plant-eating, hoofed mammal found in rocks as early as 56–47.8 million years ago [2]. Interestingly, Indohyus possessed dense ear bones which likely facilitated underwater hearing. Sound does not travel underwater the same way it does through air, so an animal that spends significant amounts of time in the water presumably would adapt to spending time in such an environment.


But why on earth would a land-dwelling animal be spending so much time in the water? One possibility can be found in the modern day water chevrotain. These hoofed mammals will dive into water to avoid predation. Such a behavior may have been the beginning of facilitating the transition to water.

Next is Pakicetus, another hoofed mammal found in rocks of the same age as Indohyus. While very similar to Indohyus, one key difference is its teeth. Whereas Indohyus, like nearly all hoofed mammals today, had teeth that were well-suited to eat plants, Pakicetus has sharp pointed teeth that are good for eating fishes. Perhaps an animal like Indohyus only forayed into water to avoid predators, whereas Pakicetus likely sought water to be a predator.


Pakicetus also had a more robust tail than Indohyus, suggesting that it may have used its tail to aid in swimming. Modern whales primarily use their tails for swimming, suggesting the beginnings of a new method of locomotion.

Perhaps Indohyus being related to cows, deer and camels doesn’t seem like too much of a stretch to you, but I anticipate that you might find Pakicetus so dissimilar to these herbivores that you might roll your eyes in disbelief. Well, one of the key features that suggests that Pakicetus is related to this group is a so-called double-pulley astragalus. Basically, this is an ankle bone with an indentation on the bottom, and it is one of the defining characteristics of the group that includes camels, pigs and deer (“artiodactyls”). Indohyus, Pakicetus, and some of the other species discussed below (Artiocetus, Rodhocetus) have this special ankle bone, further supporting the case that these are aquatic or semi-aquatic hoofed mammals.

Astragalus of Pakicetus compared to other hoofed mammals
Astragalus of Pakicetus compared to hoofed mammals

Now whereas Indohyus and Pakicetus are approximately the same age, Ambulocetus, the next transitional form, is found in younger rocks dating to 47.8–40.3 million years ago. Whereas the former two animals clearly were land-dwelling, it should be immediately apparent that Ambulocetus was better adapted for life in the water. Looking at the length of its arms relative to the rest of the body, this creature was probably not the best on land.


Beyond its limbs, Ambulocetus shows evidence of having a fat pad in its lower jaw. These pockets of fat, believe it or not, are used by modern whales to aid in underwater hearing. This suggests that Ambulocetus had a better capacity for underwater audition than either Indohyus or Pakicetus.

Ambulocetus is found in the same aged rocks as Remingtonocetus, and both display another feature indicative of a transition to an aquatic realm. Both fossils lack the telltale signs of a vomeronasal organ.


The vomeronasal organ is a sensory structure used to detect chemical signals, similar to your nose, but its found at the roof of the mouse. If you’ve ever seen a cat close its eyes, open its mouth and flare its lips, its using its vomeronasal organ.


For a lot of animals, chemical signaling is not quite as useful underwater, so the loss of the vomeronasal organ suggests a further commitment to living in the water. Interestingly, modern whales have remnants of vomeronasal organ genes in their genomes, providing further evidence of this transition.

In that same group of rocks that Ambulocetus and Remingtonocetus are found in, we find several additional species that display further commitments to the realm of the sea. Rodhocetus, Artiocetus and Protocetus have some distinctive changes to their noses, having shifted their nostrils backwards from the tip of their snouts to higher on their faces.

This is an important transition towards a more whale-like form, if you consider where a whale’s nose is. If you look at pictures of dolphins or humpback whales, you won’t find any nostrils on their snouts. So where are they? They’re on top of their heads! Blowholes are just nostrils, but by being on their noggins, they likely facilitate breathing when the animals surface. These fossils show the beginning of that transition from forward facing nostrils to nostrils on the top of their head.

Southern right whale blowhole
Southern right whale blowhole

These fossils also have anatomical evidence of the development of tail flukes, a distinctively whale-like horizontal fin that aids in aquatic propulsion. Despite having fully intact limbs, these animals were probably already swimming much like whales do.

Yet another fossil found in these rocks is Georgiacetus, which has the distinctive trait of having a pelvis detached from its spine. In humans and most other land vertebrates, the pelvis is attached to the spine to allow for maximum stability during running and walking. Georgiacetus, however, was likely spending so much time in the water that it’s hind limbs no longer needed to support its body on land.


Jumping ahead in time, Dorudon and Basilosaurus arrive on the scene some 41.3–38 million years ago, post-dating the previous eight species I’ve described. It isn’t difficult to see that these are the most whale-y looking fossils so far, with perhaps the most obvious new feature being the reduction of the hind limbs into tiny vestiges. After animals like Georgiacetus detached their hips from their spine, their limbs may have simply gotten in the way. Reducing limbs is an excellent way to streamline one’s body, allowing for increased swimming efficiency.

However, if you look closely, you’ll see that these reduced hindlimbs aren’t tiny indistinguishable bones. In fact, they still have a thigh bone (femur), a kneecap (patella), shinbones (tibia, fibula), and toe bones (phalanges). In modern whales, you can see the remnants of a pelvis and occasionally a femur, but nothing as complex as the hindlimbs of Dorudon and Basilosaurus. Interestingly, fetal whales show evidence of this transition, initiating the development of hindlimbs before they disappear entirely.

After the origination of Dorudon and Basilosaurus, the fossil record begins to show evidence of more modern whales. Based on DNA and anatomy, scientists believe that there are two general groups of whales, toothed whales (Odontoceti) and baleen whales (Mysticeti). Toothed whales include dolphins, sperm whales, belugas, porpoises and others, all of which have teeth, whereas baleen whales are characterized by their typically enormous size and having sheets of baleen in place of teeth. Baleen is made of keratin, a type of protein, and it is used like a filter to capture millions of tiny crustaceans for food.


Though baleen whales lack teeth, all of the earliest whale-like fossils described so far had teeth. Indeed, scientists think even some of their more recent predecessors had a mouth full of pearly whites, such as Aetiocetus, which post-dates Basilosaurus and Dorudon at 33.9–28.1 million years old. What’s amazing about Aetiocetus, however, is that it had distinctive features suggesting it had baleen too. Nutrient foramina, blood vessels that nourish the baleen, can be found on the upper palate of Aetiocetus, indicating that this animal had teeth and baleen side by side!


Just a little bit later in the geological record, we begin to find fossils that resemble modern baleen whales. Eomysticetus, dating to 28.1–23.03 million years ago, is still certainly primitive in some respects, such as the lack of bowed mandibles (lower jaw bones) and a stouter skull, but it distinctively lacks teeth, having shifted entirely to baleen. Notably, the genomes of modern whales have genetic remnants of tooth genes, providing further evidence of this transition from toothed ancestors to baleen-bearing descendants.


Together these fossils provide step-by-step transitions from land-dwelling to aquatic mammals, all of which share features with modern whales. Given that whales are charismatic organisms that generate much interest, keep an eye out for additional fossil discoveries providing evidence of the evolution of these impressive animals.

Questions for Creationists

Where did all of the proto-whales go? Since many of them were aquatic, shouldn’t they have been able to survive Noah’s flood? If the flood jumbled up animals randomly, why do we find these fossils in a sequence that is suggestive of evolutionary transitions? Why do we never find bones of modern whales, such as bottlenose dolphins and blue whales, alongside fossils such as Ambulocetus, Dorudon or Aetiocetus? Is it just a coincidence that the transitions inferred by the fossil record are also supported by genetics and development?


1. McGowen, M. R., Gatesy, J., & Wildman, D. E. (2014). Molecular evolution tracks macroevolutionary transitions in Cetacea. Trends in ecology & evolution. 

2. all fossil ranges derived from the paleobiology database


Transitional fossils: Are there any, really?

Perhaps the most damning criticism of Charles Darwin’s theory of evolution by natural selection was based on fossils. If organisms evolve such that their descendants adopt radically different forms than their ancestors, why don’t we see evidence of this in the fossil record? As if on cue, two years after Darwin outlined his theory in On the Origin of Species, scientists uncovered Archaeopteryx, a beautifully preserved animal seeming to have mixture of reptilian and bird-like characteristics. And so the first ‘missing link’, or transitional fossil, was discovered.


150 years later, skeptics of evolutionary theory will frequently say that Archaeopteryx does not actually qualify as a transitional fossil [1], that there are practically no other putative transitional fossils to speak of [2], fossils are always found fully formed with no realistic progenitors [3], and the fossil record is complete, so we should not expect to find any other alleged missing links [2].

I think that the first point is the most important and is really worth considering: what constitutes a convincing transitional fossil? The first quality such a specimen must absolutely possess is the presence of characteristics intermediate between two other fossil forms. Archaeopteryx most certainly qualifies on this point. For example, birds have feathers and wings, whereas modern reptiles do not, and modern reptiles have tails, teeth and claws, whereas modern birds do not. Archaeopteryx has all of these traits, proving that it is transitional in form.

But having an intermediate form does not mean that Archaeopteryx descended from reptiles and gave rise to birds. Take for instance, the platypus. The platypus has hair and produces milk for its young, just like other mammals, but it lays eggs like a lizard. While biologists think that this provides evidence that mammals descended from reptile-like ancestors, nobody would argue that our modern day platypuses gave rise to the mammals that don’t lay eggs.

This is where the next piece of critical information comes in: the dimension of time. It isn’t enough to look like a medley of organisms. Rather there has to be a logical chronological sequence to their appearances. Geologists use a variety of information, including the clock-like radioactive decay of atoms, to estimate the ages of rocks. If Archaeopteryx appears in rocks that are more recent than those of the earliest reptiles, but also pre-dates the appearance of the earliest birds, then it constitutes a specimen consistent with it being a transitional fossil. (spoiler alert: it does)

And this is a testable, refutable hypothesis. If tomorrow scientists provide insurmountable evidence that modern bird fossils predated or lived alongside Archaeopteryx,  we can no longer say that Archaeopteryx meets the transitional fossil criteria.

But why aren’t there more fossils like Archaeopteryx? There are, but certainly not the millions and millions that must have existed if evolutionary theory is correct. So why haven’t we found them?

The simple answer is bias. There is a whole field of research (taphonomy) that seeks to understand when, where and how fossils are formed, and to be blunt, we think we’re missing a lot of information. The general idea is that you have to die in the right place, at the right time, with the right conditions in order to preserve information to last millions of years. The fossils that researchers tend to find are almost always extremely fragmentary and limited. For instance, fossil teeth are extremely abundant. Fossil brains and skin? Not so much. Whole organisms? We wish!

Then there’s a more human problem: paleontologists tend to focus on the best fossil-bearing rocks that they’ve discovered so far, yet even those are limited. Not only do the fossils have to actually land in the right place and the right time, but the earth has to shift and erode in such a way to expose the fossils. How many fossils might be under your feet right now, but you’ll never know because you’re never going to dig under your house? How many fossils are buried in the sea? Under antarctic ice? What are we missing out on that we could find only if we could CT scan the entire earth?

Taphonomic bias

So the fact that there are only so many fossils we can find, and we only find them at a certain rate, and we usually only get to look at parts of bones and not the squishy parts of the bodies that presumably also were evolving, how much information are we missing out on here?

Which brings us to the third common point: if the fossil record is complete, how could we ever find any other transitional fossils? I’m not sure who first conveyed this idea, but boy were they wrong. As just one simple example, consider the fossil record of Mesozoic mammals, i.e. mammals that are found in rocks that date between 251 and 66 million years ago. Between 1830 and 1979, researchers found 116 general kinds of mammals that date to this era. Between 1979 and 2007, that number climbed to ~310 [4]. Far from being complete, it appears that paleontologists are just getting started!

You can bet that as paleontologists keep digging, researchers will uncover more putative transitional fossils, many of which I document here in this blog. Keep an eye out for them, and ask yourself: do these fossils meet the criteria for being transitional? And if so, what does this mean for Creationism?





4. Luo, Z. X. (2007). Transformation and diversification in early mammal evolution. Nature450(7172), 1011.