table of contents
September 2001
[back] [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [next]

The Emergence of Whales, Chp. 15

Synopsis of Chapter 15: Paleobiological Perspectives on Mesonychia, Archaeoceti, and the Origin of Whales

Philip D. Gingerich, Museum of Paleontology, University of Michigan, Ann Arbor, MI 48109


Starts with a discussion of taxonomy: how organisms are grouped together based on similarities and differences. Whales (order Cetacea) are mammals: endothermic, lactating, big brains, and highly active. Specializations: body size, reduced/simple dentition, audio-dominated sensory system, hydrodynamic streamlining with tail propulsion, and many more.

Mysticeti and Odontoceti are usually Cetacea suborders, but some whale specialists have regarded them as distinct orders. Why? CONTEXT! Including an interdependencee of morphology, classification, and evolutionary history.

“When mammals as different as odontocetes and mysticetes were classified in different orders, this was interpreted to reflect a long history of evolutionary independence (the history had to be long because of a general belief that evolution is so slow that differences take a long time to accumulate).”
Now we Odontoceti and Mysticeti go back to the Oligocene. Diverged late Eocene/early Oligocene, not more than 40 mya. Whales from Eocene rivers and oceans = Archaeoceti. No whales known before Eocene. Presumed ancestry to Mesonychia, based on morphology: both will be considered.

1.1 Study of Whale Origins Since most mammals live on land, Cetacea aquatic specializations have “long been viewed” as derived characteristics acquired when they made the transition from land to sea. Initially suggested due to long history of mammals on land, short history of whales in sea.

Considers all the questions: From who? Where? When? HOW? What caused it? What happened next? Three broad objectives of the study:

“1: Identification of the morphologically, geographically, and temporally intermediate stages of change (here the stages by which whales made the transition from land to sea). These intermediates, when known, are direct evidence (and the only direct evidence we have) telling us what happened in evolution.

2. Association of the times of acquisition of distinctive morphological specializations with other changes in morphology within the group of interest (here Cetacea) and with biotic- and physical-environmental changes outside the group of interest. These associations provide a context critical for understanding how any evolutionary transition took place. 3. Evaluation of consequences. What was the effect of any change on the group under study? This can be measured in terms of morphological disparity, taxonomic diversity, or taxon longevity.”

Notes that studing Archaeoceti “in its infancy”! Brief history of Basilosaurus, Richard Owen, Remington Kellogg, and some mixed-up reconstructions.

Protocetus was of interest to Kellogg. Thought that Protocetus was off the Odontoceti/Mysticeti line. George Gaylord Simpson thought the same, i.e., that the archaeocetes weren’t ancestral to modern whales.

Boyden and Gemeroy in 1950s used immunology to look at serum proteins of Cetacea and other orders (one of first attempts to infer phylogenetic relationships with immunology). Artiodactyl relationship was suggested. Essentially confirmed by modern gene methods. Mentions the “cetaceans are highly derived artiodactyls” debate, to whit, genes vs. physical morphology. “conflicting claims … taken together, cast doubt on our ability to reconstruct past evolutionary history from living animals.” Van Valen (1966) concluded from morphology that the archaeocetes (and modern whales) had to be descended either from Mesonychidae or Hyaenodontidae. Quote given. Van Valen looked at Andrewsarchus (similar to Protocetus). Figured whales got into the ocean about mid-to-late Paleocene. (A little early.) Noted the Boyden and Gemeroy results lent credence.

Discussion again of sister-group relationship between Artiodactly- Cetacea vs. “mother group” relationship. Point to be discussed later; most authors work with Mesonychia –> Archaeoceti.

1.2. Diversity and Morphology of Mesonychia

About 20-28 known genera. Table shows who, when, and where. Discussion of possible crossovers and different classifications by different researchers. Time range of occurrence is early- mid Paleocene (63 mya) to Oligocene (33 mya).

There are 4 Mesonychidae represented by complete skeletons:
mid-Paleocene Hukoutherium

late Paleocene Sinonyx ** early Eocene Pachyaena ** (to which many archaeocete comparisons have been made; now we know why. Also, Gingerich is the one writing it up!)

mid-Eocene Mesonyx
– Skeleton of early Eocene Pachyaena –

Gingerich makes use of skeletal proportion diagrams, the first two of which are shown here. The skeletal proportions of Pachyaena ossifraga are compared to Canis lupus (wolf), which is skeletally similar. They are bar diagrams expressing the ratio of the measurement to the “mean height of the centrum of the six thoracic vertebrae”. Shown are skull lengths, vertebral length and height, and forelimb and hind limb long bone lengths are shown (a lot of info packed into one diagram). Regular proportions are shown for the skull and limbs; the bar for the spine shows the difference between the centrum length and height, and its position (above and below the mean line) indicates the size. It makes more sense to see than to describe, and I can’t show them. What these are most useful for is to allow comparisons of skeletal changes for animals of somewhat different sizes. They also show the “standing waves” in the spine that indicate locomotion mode, as discussed in Chapter 11.

As for Pachyaena…
Skull resembles Canis in size, about 1 order of magnitude larger than the spinal baseline. Differences due to dentition.

Vertebral column similar to Canis as well (seeing is believing), pattern typical of “cursorial” land mammals. Cervicals and anterior thoracics form anterior arch for skull support. Further explanation of the diagram. [“Cursorial” means related to running]

Posterior thoracic, lumbar, and sacral centra form the 2nd arch, the central arch. Similar in both, with similar connections at shoulder and to hind limb.

One more arch, the posterior. Differences in limbs between Pachyaena and Canis: Radius is shorter than the the humerus in the forelimb, tibia is shorter than the femur in the hindlimb, metacarpals and metatarsals are shorter. This indicates Pachyaena had slightly heavier build and wasn’t quite the runner as Canis.

Details: Pachyaena’s central cheek teeth don’t resemble that of carnivores (including archaeocetes), and aren’t as sharp or as shearing as meat eaters. The lumbar vertebrae have “revolute zygapophyses”, meaning they would run stiff-backed. Terminal phalanges are like ungulates, consistent with non-predatory behavior. Mostly indicates that Pachyaena was not an active pursuit predator, more likely a solitary carrion feeder/scavenger that trotted around and chewed somewhat decomposed, rather than fresh, flesh. (Interesting behavioral suppositions; presumably mesonychians started looking for dead fish on the beach? A little bit more on this in section 1.3.1 when Ambulocetus is briefly considered.)

1.3 Diversity and Morphology of Archaeoceti

As we know, they are now about 25 genera of Archaeoceti, grouped into Ambulocetidae, Basilosauridae, Dorudontidae, Protocetidae, and Remingtoncetidae. (Next table).

Found on the margins of most of the world’s oceans. Six genera known well enough osteologically to do skeletal proportions:

Rodhocetus (Lutetian, i.e., early-mid Eocene)
Dalanistes (Lutetian)
Protocetus (Lutetian)
Dorudon (Bartonian, i.e., latest mid Eocene)
Basilosaurus (later Bartonian, early Priabonian)
Saghacetus (Priabonian)

Notes that Ambulocetus and Pakicetus have important limb bones, but not enough for this type of analysis.

1.3.1. Skeletons of Early Middle Eocene Rodhocetus and Dalanistes Figures are great. I should just type in the caption. I’ll do my best.

From the text:
Skulls different from Pachyaena by: external nares open at position in front of dentary, much larger mandibular foramen — latter necessary for hearing in water.

Cervical vertebrae shorter than Pachyaena. Cervicals of Rodhocetus “noteworthy” with centra shorter than high, found in ALL later archaeocetes and modern whales. Indicates neck shortening. Rodhocetus shows only a hint of the central arch. Dalanistes shows it a bit more — in both cases much less pronounced than Pachyaena. Indications they could use forelimbs to lift body weight on land (long neural spines on thoracics). Rodhocetus has 4-centrum sacrum unfused, Dalanistes has 4-centrum fused sacrum.

No forelimbs available: hind limbs and pelves consistent with land capability. Different skull proportions: Rodhocetus has normal archaeocete proportions, Dalanistes is long and narrow. Indicates feeding differences: skeletal characters indicate Rodhocetus was the better swimmer and aquatic pursuit predator, while Dalanistes was slower and —

“probably an aquatic ambush predator like earlier Ambulocetus”.
[Synopsizer’s comment: Back to those behavioral suppositions for a minute. I found it interesting that it was supposed that the Mesonychidae were scavengers, due to the way alligators behave. Gators will lie in wait in the water until they can grab something, which they drown. In a lot of cases they stash the dead prey and let the water and decomposition soften it a bit, then they come back to feed. One time a gator got a diver at Wakulla Springs, and the glass-bottomed boats happened on Mr. Gator taking the body to his stash spot. Initially they thought the croc had snagged a deer, and when they looked closer they got a nasty surprise. So I guess Ambulocetus and Dalanistes are presumed to behave something like modern-day alligators. Interestingly enough, the way gators and crocs swim is a lot like what’s proposed for Basilosaurus. So was Basilosaurus a shallow-sea ambush predator? Hmmm.]

The caption essentially summarizes the text, but I have to quote this sentence:

“Dalanistes and Rodhocetus are intermediate in skeletal proportions and vertebral profile between land mammals like Pachyaena and aquatic archaeocetes like Dorudon.” Gotta love it.
1.3.2. Skeletons of Middle to Late Eocene Dorudon and Basilosaurus

Of course there’s a figure. Nares open farther back on the rostrum. Cervical centra are further shortened, no sacrum, only one long vertebral arch is now present. “… different in detail but similar in overall functional conformation to vertebral arches of extant odontocetes and mysticetes”. Forelimbs well-known, with humerus much longer than radius. Elbow joint still mobile. Hind limbs reduced in size. No sacrum, free pelvis = couldn’t support weight on land, so fully aquatic.

Principal differences between D. and B.: relative sizes and shapes of posterior thoracics, lumbars, and anterior caudals. (You can sure see this in the figure!) Makes Basilosaurus “anguilloform” and definitely affected swimming. Buoyancy of vertebrae means that Basilosaurus was a surface swimmer, while Dorudon was a “three- dimensional diving swimmer”. Consistent with Slijper (1946) that Basilosaurus would have moved “partly by lateral undulation”. Figure 6 shows skeletal proportions of two extant cetaceans: Feresa attenuata (globicephalid odontocete, also known as the pygmy killer whale) and Balaenoptera acutorostrata (balaenopterid mysticete, AKA minke whale). These are provided for subsequent comparative purposes.

Useful Order Cetacea page:

1.4 Body Mass of Mesonychids and Archaeocetes

Mesonychians are typical land mammals, so body mass can be figured in the usual ways. Hapalodectidae: 1-8 kg. Mesonychidae: 10-250 kg (large mammals).

Archaeocetes aren’t typical land mammals. So he developed a body mass estimation method using selected vertebrae, based on eight cetaceans ranging from 150-23,500 kg, and 5 pinnipeds ranging from 85-1210 kg. Est. body masses are given in the figures; here they are again (95% confidence interval in parentheses):

Pachyaena: 160 kg (110-220) 65 kg predicted from long bones
Canis: 90 kg (30-260) actual 30 kg (so I guess we can see how
accurate this technique is!)
Rodhocetus: 590 kg (410-850)
Dalanistes: 750 kg (440-1280)
Dorudon atrox: 1140 kg (740 – 1770)
Basilosaurus cetoides: 5840 kg (2910 – 11,720)

1.5 Brain Mass and Relative Brain Size of Mesonychids and Archaeocetes

Brain mass now known for mesonychids Mesonyx and Pachyaena. Gives background on brain mass/body mass estimates and calculation of encephalization quotients (EQ). I’ll summarize results.

Mesonyx: endocranial volume 80 cm^3 (typo in the text had “2” instead of 3, which is the brain of a Flatland species), mass 80 g.
EQsubTC (TC stands for Terrestrial, Class-level) is 0.51 – 0.40.
Pachyaena: volume 40 cm^3, mass 40 g, EQsubTC is 0.18.
Dorudon: some confusion with earlier attempts, including a “Dorudon” now called Saghacetus osiris. So finally we have

Saghacetus osiris: 485 cm^3
Dorudon atrox: 1200 cm^3
Basilosaurus isis: 2800 cm^3. Problems with exact conversion of volume to mass, due to the presence of the intercranial rete mirabile. About 20% est. in Dorudon atrox. So brain masses are about 388, 960, and 2250 g.

And there’s also uncertainty in body mass estimates. But they come up with 350 kg for Saghacetus, 1140 for Dorudon, and 6840 for Basilosaurus (a little bigger than the previous estimate. EQsubTC:

S. osiris: 0.49
D. atrox: 0.51
B. isis: 0.37

Avg. for mid-to-late Eocene archaocetes is 0.46, which means that their brains are about 46% as large as that of an average terrestrial mammal of the same body mass living today. Question he doesn’t answer here, but addresses later: how do EQs compare for modern terrestrial mammals and modern cetaceans?)

Rodhocetus: 0.27
Dalanistes: 0.29 So I guess they weren’t heavy thinkers. Someone else might care to comment on evolutionary “expectations” regarding EQs.


[ If you’ve been reading these, and have found them mildly amusing and somewhat informative but haven’t really dug in and paid attention, now’s the time to PAY ATTENTION. I’m going to be quoting liberally. ] Starting off:

“Eocene Archaeoceti are important for understanding the land-mammal ancestry of whales because they are intermediate (!!) in time, space, and form (!!) between slightly earlier Paleocene and early Eocene Mesonychidae and slightly later Oligocene Mysticeti and Odontoceti.”
Cites “morphological intermediacy” of archaeocetes, which can be appreciated by comparing the skeletal proportion diagrams. (The main things seen are the “flattening” of the terrestrial arch structure and the development of the long hind-torso and tail section, which we’ve heard about before). Also, nasals are moving back on the skull.

3 trends in the vertebral column to note:

  1. Length of cervicals decreasing, while cervical height stays about the same: significant changes in cervical shape.
  2. The central and posterior vertebral arches “merge” into a single arch with loss of the sacrum as in intermediate support point.
  3. Vertebrae in the new single arch assume a more equidimensional shape as their functions become more uniform/uniformly shared.
“I have quantified several other changes in the transition from mesonychids to cetaceans in Figs. 7 and 8.” (Body size, trophic specialization based on size of largest tooth, auditory specialization based on bulla length, and hydrodynamic streamlining/limb reduction based on femur length).

[Each will be discussed subsequently. Gingerich does two nice things here: The Y-axis for each figure is the age of the species. The X-scale varies depending on the quantity. Fig. 7 is direct measurement (log length scale), while Fig. 8 is the same parameter size-normalized based on body mass. These are superb trend figures, because they illustrate what happened when! Body mass obviously shows increase through time. Tooth size is relatively uninteresting because it’s pretty much directly related to body size. Bulla length shows a more rapid increase early than late. Femur length reduction is pronounced when the size-normalized trend is examined. I.e., as they got substantially bigger, the rate of hind limb reduction accelerated.]

2.1 Body Size

Yes, they got bigger in general, but relatively small mass archaeocetes like Saghacetus are also found in the late Eocene.

2.2. Trophic Specialization

The trend is explainable larger due to body mass increase, but the actual “largest tooth” is different in mesonychids vs. archaeocetes. Pakicetus had “delicate and pointed” incisors. “Both of these differences from mesonychids imply that the change to a characteristically archaeocete dentition and trophic specialization was achieved very early in archaeocete evolution.” “Experimentation” noted in ambulocetids and remingtoncetids, “presumably related to feeding”. Later trophic specializations related to pursuit swimming ability.

2.3 Auditory Specialization

Larger separation between Pakicetus and later archaeocetes. Looking at body mass effects, variance is larger than for tooth size. Basilosaurus had the smallest bullae, Saghacetus much larger for its size. Due to acoustic characteristics of seawater and body size combined? (Mysticetes have small auditory bullae, which lends support, he says.)

2.4 Hydrodynamic Streamlining and Hind Limb Reduction

“One of the most interesting changes in archaeocete evolution is reduction of the hind limb from a limb size typical of land mammals to the small limb of Basilosaurus and Dorudon. Reduced limbs in late archaeocetes probably retained a functional role in reproduction, but these are clearly too small to support the body on land.”
Cites disarticulation of the innominate bones of the pelvis from the sacrum as part of the same trend. Strong trend relative to body mass, as noted in caption. Coincedent with cervical shortening, “… both are part of the hydrodynamic streamlining necessary for efficient swimming.” Remainder is about Dorudon and Basilosaurus and has been said before, except that they must have had “precocial” (“capable of a high degree of independent activity”) birth.

2.5 Encephalization

Ah, he does me a favor: he looks at the EQs of modern mysticetes and odontocetes, too. There’s a nice trend of increasing brain size early-to-late archaeocetes, but the brains are still smaller than for an equivalent body mass modern mammal.

BUT: “When we compare the relative brain size of extant whales, they span virtually the full range of sizes seen in terrestrial mammals living today.” Mysticetes smaller than expected for body mass, but they have BIG body mass! Odontocetes larger than expected, even with large body mass. Size-normalized, odontocetes “as a group would undoubtedly prove to have the largest brains of all mammals.”


First paragraph quoted verbatim:

“The most important thing that can be said about Eocene Archaeoceti is that they are beginning to fill the temporal, geographical, and morphological gap between Paleocene land mammals and Oligocene and later whales. The temporal and geographic distributions of Mesonychia and Archaeoceti in Figs. 1 and 3 [actually tables] support this. Size-adjusted comparisons of morphological characteristics of the skull, vertebral column, forelimb, and hind limb in Figs. 2, 4-6 show the general pattern of change from a wolflike mesonychid model ancestor through primitive remingtoncetid and protocetid archaeocetes to more advanced dorudontid and basilosaurid archaeocetes, to modern mysticetes and odontocetes. Characteristics of archaeocetes such as body mass, tooth size, auditory bulla size, femur length, and relative brain size are documented in Figs. 7-9, which are related to progressive trophic, auditory, locomotor, cognitive, and life history adaptation to life in the sea.”
Data now refutes the idea of a long separation between Archaeoceti and Mysticeti/Odontoceti. Associations with the Artiodactyla and mesonychids in the ancestry of archaeocetes are now widely accepted. Regarding the three goals stated in the Introduction, goal 1 is being addressed well, more to be learned about the “context” (goal 2), and it’s too early to say much about goal 3.

Readers are invited to research the connection between the terms “adaptation” and “natural selection”. The following URLs are provided to assist this research:

(Note the section “The Principle of Natural Selection”)

We’re almost to the finish line: The final chapter, 16, is entitled “Cetacean Origins: Evolutionary Turmoil during the Invasion of the Oceans” and is written by the editor of the book, J.G.M. Thewissen.

[back] [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [next]

This page was last updated September 1, 2001.
It was reformatted and moved August 6, 2007
Copyright © 2001 by James Acker

table of contents
September 2001