Evolutionary Treasures Locked in the Teeth of Early Whales

Whales are highly-modified, once-hoofed mammals which are entirely aquatic. This is arguably one of the greatest of evolutionary punchlines. We just didn’t get the joke until relatively recently.

Upon viewing a whale skeleton, the fact that the animal once had terrestrial ancestors is not too difficult to detect – hints that their forerunners walked the prehistoric shoreline can be seen in the arrangement of the long, altered fingers which provide the framework for whale flippers, as well as the minute vestiges of hindlimbs and hips which are deeply embedded in whale bodies. Naturalists understood the relevance of these features to whale evolution long ago. The trouble was that the fossil record of whales was so poor that no one could be sure exactly how cetaceans had evolved. Perhaps whales were modified from some otter-like creature, as proposed by Thomas Henry Huxley, or, as William Henry Flower speculated, maybe whale ancestors were omnivorous, hoofed, hippo-like animals which dwelt in prehistoric marshes.

The deep ancestry of whales remained a virtually insoluble problem for decades. Whales must have had land-lubber forebears, but the fossils necessary to flesh out the details of the transition were entirely missing. All of the known fossil whales were fully-aquatic. No whale with limbs or other transitional features had been found. Things were so frustrating that in 1976 paleontologists Jere Lipps and Edward Mitchell wondered if any fossils of the transformation would be found at all – perhaps the changes happened so fast that the chances of creatures with transitional features being preserved was low.

Nevertheless, by the late 1960’s paleontologists had identified a possible parent stock for whales on the basis of anatomical evidence. In 1966 Leigh Van Valen proposed that whale ancestry could be drawn back to an odd group of hoofed, carnivorous mammals called mesonychids. Some of these creatures, Van Valen hypothesized, “were mollusk eaters that caught an occasional fish, the broadened phalanges [finger and toe bones] aiding them on damp surfaces”, and the resemblance between the teeth of the earliest known whales and mesonychids appeared to support this connection. Even better, in 1981 paleontologists Philip Gingerich and Donald Russell found what they thought was a confirmation of Van Valen’s suggestion in the remains of one of the earliest whales, Pakicetus. The small collection of teeth attributed to the animal – a creature primarily represented by the back portion of a skull – corresponded to the mesonychid type.

The big question of where whales had come from appeared to be solved, but then the course of the research took a left turn. Biomolecular and genetic studies repeatedly found that whales were most closely related to the group of hoofed mammals called artiodactyls, particularly hippos. (This major group of mammals is distinguished, among other things, by the possession of an even number of hoofed toes on each foot, and encompasses animals such as camels, pigs, deer, antelope, and many others.) The fossil record and the evidence from the living animals was in apparent discord, and resolution was difficult to achieve. Which was more reliable in drawing out the origins of leviathan – fossil teeth or genes?

As it turned out, whales were not only more closely-related to artiodactyls than to the mesonychids. They actually are artiodactyls. (Today the group containing whales plus all the traditionally-recognized artiodactyls is called the Cetartiodactyla.) In 2001 two different teams of paleontologists described the rare ankle bones from three different early whales – Hans Thewissen and colleagues reported on Pakicetus and Ichthyolestes, and a team led by Philip Gingerich described Rodhocetus. (These whales were not like their living cousins – each was an amphibious species which still retained functional limbs. It is difficult to find stranger creatures in the fossil record.) While a single part of the skeleton may seem like too little to influence a major debate, though, the three early whales were found to have a particular ankle bone – known as the astragalus – with a distinctive “double-pulley” shape which is characteristic of artiodactyls. Mesonychids do not share this same shape, and so the genetic and fossil data were finally brought into accord.

A more recent discovery bolstered the new view of whale ancestry. In 2007 Thewissen and colleagues presented the small fossil artiodactyl Indohyus as a very close relative of early whales. Though not directly ancestral to walking whales such as Pakicetus, the animal exhibited a suite of transitional features which hint that whales got their start as semi-aquatic omnivores or herbivores which later switched over to being specialized fish-hunters. This picture was different from what had previously been supposed. Under the discarded mesonychid hypothesis, whale ancestors were already committed carnivores that simply slid into the water, but the new picture indicated that whales underwent some major changes in diet and dentition during their early evolution.

Thewiseen, Jennifer Sensor, Mark Clementz, and Sunil Bajpai have just published an overview of early whale diet in Paleobiology. Whales, they point out, originated from a group that was either herbivorous or omnivorous, and so the formidable, flesh-shearing teeth of early whales such as Pakicetus were a specialization which evolved sometime during the early history of whales. In order to gauge the timing of this major event – as well as investigate how early whale diets changed – the researchers carried out an in-depth dental exam on several of the creatures.

Obviously, we can’t directly observe how prehistoric whales fed or what they ate. In order to investigate these mysteries, scientists turn to indirect evidence preserved on and inside the teeth themselves. Since mammals only get two sets of teeth – a milk set, which is shed, and a permanent adult set – their feeding choices and strategies leave distinctive wear patterns on their teeth which can be examined for clues. Additionally, chemical isotopes of oxygen and carbon are locked away inside the teeth. These chemical signatures are influenced by environment and behavior and can be used to construct a general picture of an animal’s ecology. Whereas carbon isotopes relate to diet, oxygen isotopes have been used to examine details of the surrounding environment (whether an animal lives in a freshwater or saltwater habitat, for example).

A variety of different fossil mammals were included in the study. Thewissen and co-authors selected first and second molars from early artiodactyls (Bunophorus, Khirtharia, Anthracobunodon), the anthracothere Elomeryx, the whale-cousin Indohyus, and a number of early whales spanning the transition from the water’s edge to the sea (Pakicetus, Rodhocetus, Babiacetus, and Zygorhiza). By looking across this swath of fossil mammals, the researchers hoped to detect how drastic the shift in early whale diet was, whether it coincided with the adaptation of early whales to an aquatic lifestyle, and if all early whales shared the same carnivorous lifestyle.

In terms of tooth wear, the early whales included in the study were clearly different from the land-dwelling artiodactyls such as Bunophorus. The reason why is attributable to the way the whales were chewing. Thewissen and colleagues were looking at three different kinds of tooth wear which indicate different jaw motions – apical wear (caused by crushing and puncturing of food by the tips of the tooth cusps), Phase I wear (created by initial contact of the teeth), and Phase II wear (minute damage done when the teeth grind food against the upper and lower teeth). Whereas creatures such as Bunophorus exhibited all three wear types, the molars of most early whales were dominated by Phase I tooth wear. The kind of wear created by puncturing prey or grinding food was almost entirely absent, and, as was expected on the basis of evolutionary relationships, Indohyus turned out to be more like early whales in this regard than other artiodactyls. The only, strange exception was the whale Babiacetus – this whale had a significant amount of apical wear on its teeth, including teeth which had been broken and then polished by continued use.

Why the early whales clustered together in the tooth wear diagram is probably a result of their tooth shape. All of them had high molars which would have created a kind of cutting shear to slice through their fishy prey. This type of molar shape appears to have evolved early on in whale history and remained intact through the evolution of fully-aquatic whales such as Basilosaurus and Zygorhiza, and, by combining the tooth wear diagram with our present understanding of whale evolution, it appears that this transition happened sometime between the evolution of Pakicetus and its Indohyus-like forerunners. Just what Babiacetus was doing, though, is a mystery. Perhaps the whale was tackling larger prey with bigger, harder bones, or perhaps the difference is an indication that the whale had a more varied diet then its cousins. Thewissen and co-authors do not put forward any firm conclusion on this point – only one individual of Babiacetus was included in the study, and more comparison with more specimens are necessary to see whether the tooth wear patterns are common are peculiar to the one individual. Nevertheless, the different pattern of tooth wear in the one individual and the odd tooth shapes of some other early whales – such as Andrewsiphius – hints that these animals had a wider array of feeding strategies than we presently understand.

The data gleaned from the carbon isotopes indicated a curious paradox, however. Even though Indohyus exhibited tooth wear that was similar to that seen among the earliest whales, it was not consuming the same kind of food. Indohyus appears to have had a diet similar to that of the land-dwelling Khirtharia – thought to feed primarily on terrestrial plants – and there is no indication that Indohyus was carnivorous. What this means, Thewissen and collaborators hypothesize, is that Indohyus was chewing in the same way that the early whales did despite having a significantly different diet. Exactly why this should be so is left unresolved – maybe Indohyus ate plants which required a similar chewing strategy, or maybe the way Indohyus chewed was constrained by the way its other teeth interlocked.

A change in the way whale ancestors chewed may have preceded the transition to a carnivorous diet. Further studies will be required to reinforce this hypothesis, particularly since Indohyus is acting as the sole keystone between early whales and their more ancient, artiodactyl ancestors. Still, the emerging picture is that the transition from land-to-water occurred in a mosaic fashion. If we take Indohyus as a model for the sort of creature at the root of the whale family tree, then the forerunners of early whales primarily ate terrestrial plants, spent most of their time on land, but would sometimes swim in fresh bodies of water and chewed in specialized, shearing manner which would later be co-opted by their descendants to cut flesh. Just as the initial emergence of vertebrates onto land over 375 was not a linear march onto land, the origin of whales was not a directed dive into the sea in which all the important evolutionary changes occurred in the water. Co-option of existing structures and behaviors had their role to play, and made it possible for whales to eventually walk into the sea.

For more on our changing understanding of whale evolution, check out the chapter “As Monstrous as a Whale” in my first book Written in Stone.

Top Image: A cast of the reconstructed skeleton of the early whale Pakicetus at the Natural History Museum of Los Angeles County. Photo by the author.

References:

Gingerich, P. (2001). Origin of Whales from Early Artiodactyls: Hands and Feet of Eocene Protocetidae from Pakistan Science, 293 (5538), 2239-2242 DOI: 10.1126/science.1063902

Gingerich, P.D., D.E. Russell. 1981. Pakicetus inachus, a new archaeocete (Mammalia, Cetacea) from the early-middle Eocene Kuldana Formation of Kohat (Pakistan). Contributions from the Museum of Paleontology: The University of Michigan. 25: 235-246

Lipps, Jere, and Edward Mitchell. 1976. Model for the adaptive radiations and extinctions of pelagic marine mammals. Paleobiology 2: 147-155

Thewissen, J., Sensor, J., Clementz, M., & Bajpai, S. (2011). Evolution of dental wear and diet during the origin of whales Paleobiology, 37 (4), 655-669 DOI: 10.1666/10038.1

Thewissen, J., Williams, E., Roe, L., & Hussain, S. (2001). Skeletons of terrestrial cetaceans and the relationship of whales to artiodactyls Nature, 413 (6853), 277-281 DOI: 10.1038/35095005

Van Valen, Leigh. 1966. Deltatheridia, a new order of mammals. American Museum of Natural History Bulletin 132:1-126