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. 2

Chapter 2: Middle to Late Eocene Basilosaurines and Dorudontines

Mark D. Uhen Cranbrook Institute of Science Bloomfield Hills, MI

1. Introduction

How it began: with Basilosaurus, described by Harlan in 1834, recognized as mammalian and cetacean by Owen, 1842. Properly named Zeuglodon, but it was a junior synonym, so Basilosaurus survives. Dorudon serratus described in 1845 by Gibbes. Basilosaurus kochii, now Zygorhiza, described by True from Alabama, 1908.

There are a lot of these guys.


* Basilosaurus cetoides, Harlan 1834
Bartonian – early Priabonian
Ouachita Parish, Arkansas (Numerous specimens, virtually the whole skeleton is known)

*Basilosaurus isis, Andrews 1906
Bartonian – early Priabonian
Birket el-Qarun beds and Qasr el-Sagha formation, Fayum, Egypt

*Basilosaurus drazindai, Gingerich 1997
middle to late Bartonian
Bar Nadi W. of Satta Post, Punjab, Pakistan
based on vertebrae only

*Basiloterus hussaini, Gingerich 1997
middle to late Bartonian
Bar Nadi W. of Satta Post, Punjab, Pakistan


*Dorudon serratus, Gibbes 1845
Cooper River, South Carolina (Harleyville Formation?)
partial skeleton, mostly jaw, cranial, and spinal

*Dorudon atrox, Andrews 1906 (!!) a species to keep in mind
Bartonian to early Priabonian
Gehannam and Birket Qarun Formations, Fayum, Egypt
several previously named species now classified as D. atrox; almost the entire skeleton is known

*Pontogeneus brachyspondylus, Muller 1849
USA and Egypt sites yielding other basilosaurines

*Zygorhiza kochii, Reichenbach 1847
Bartonian-Priabonian of N. America
Ocala Limestone, Clark County, Alabama (Moodys Branch formation)
most of the skeleton has been described in detail

*Saghacetus osiris, Dames 1894
Bartonian – early Priabonian
Qasr el-Sagha Formation, Zeuglodon Valley, Egypt
reasonably well-known skeletally

* Ancalacetus simonsi, Gingerich and Uhen 1996
Bartonian – early Priabonian
Birket Qarun formation
cranium, jaws, vertebrae, forelimbs

2. PHYLOGENETIC RELATIONSHIPS of BASILOSAURIDS (this is where it gets interesting!)

It’s hard for me to interpret his phylogenetic diagrams, so I have to summarize what he says. Much of this is based on Uhen’s Ph.D. thesis. The method he used was “stratocladistics”, defined as a method that “uses morphologic data in the same way that cladistic analysis does, but it allows the relative stratigraphic occurrence of the fossil taxa to be included as data in the analysis.”

Result: Basilosauridae is derived from the paraphyletic Protocetidae. The subfamily Basilosaurinae is derived from within the Dorudontinae, as is the Odontoceti + Mysticeti.

2.2 Relationships of Basilosauridae to Earlier Cetaceans

Mesonychid mesonychians are the sister taxon to Cetacea, but not sure which ones (mesonychids, hapalodectids, or andrewsarchids).

Primitive characters of early cetaceans (Pakicetidae, Ambulocetidae, and Remingtoncetidae) = large hind limbs, sacra with multiple fused vertebra, poorly developed aquatic auditory systems. Pakicetids are the most primitive, the others are derived.

Protocetidae: Rems, Prots, and Ambs share a common ancestor: therefore Ambulocetus is NOT the direct ancestor of all later cetaceans. Protocetids share some characters with basilosaurids: reduced sacra, smaller hind limbs, more posteriorly placed narial openings. (!!) The protocetid most closely related to the basilosaurids is either B. indicus, E. schweinfurthi, or G. vogtlensis.

2.3 Relationships within Basilosauridae and Later Cetaceans

The relationships within Basilosaurinae are somewhat trivial; they’re obviously related.

All dorudontines share characters with Odontoceti + Mysticeti, like loss of the sacrum, rotation of the pelvis, reduction of the pelvic girdle and hind limb, increase in size of the infraspinous fossa of the scaula (sic) [scapula?], radius and ulna articulate with the humerus in a common trochlea, and lack of saddle-shaped articular surfaces on the carpals. Almost any dorudontine could be the sister taxon to the Odontoceti + Mysticeti because no one species shares an overwhelming number of characters with O+M.

(!!) It’s important to leave here with: the dorudontines are the ancestors of modern cetaceans. Dorudon atrox or Zygorhiza kochii have been suggested as “the” ancestor.


This section (8 pages) is chock-a-block with pictures of archaeocete fossils (in case someone thought this was all a hoax). I can’t possibly go into detail, so I’ll try to hit some of the high points.

3.1 Forelimb characters (!!) “The forelimbs of basilosaurid archaeocetes are intermediate in form between the forelimbs of mesonychids and those of modern cetaceans.” He said it, I retyped it. The forelimb of Ambulocetus has both some primitive features of mesonychids and advanced features of basilosaurids. (List follows)

Next paragraphs discuss scapula, humerus, radius, ulna, and carpals of mesonychids, basilosaurids, and modern cetaceans. The humerus becomes progressively more robust (as an example). The large fan-shaped scapula found in basilosaurids and modern cetaceans indicates a fully aquatic existence for basilosaurids, as it is not found in terrestrial mammals.

3.2 Vertebral characters

Cervical vertebrae become shorter.

General uniformity of vertebral length from the posterior thoracic to the anterior caudal region. This is a diagnostic tool: Mesonychians (like other terrestrial mammals), have vertebrae that vary in length down the column. (!!) Rodhocetus has vertebrae that vary in length less than mesonychids but more than basilosaurids (data for basilosaurids is plotted).

Summary: “These vertebral characters all suggest that basilosaurids were even more aquatic than any earlier cetaceans.” Why: “The vertebral columns of terrestrial mammals must support load placed on them at the cranial and caudal ends and at the limbs. … Cetaceans that are always supported by water and never have to venture out on land, such as basilosaurids, do not have to have vertebrae of different sizes to distribute the load of the head and limbs. (!!) Lastly, the changes seen in the caudal region strongly suggest that basilosaurids had a tail fluke like modern cetaceans, and had thus adopted the cetacean mode of propulsion from a tail fluke, which would be almost useless on land.”

3.3 Hindlimb characters

Quite a bit of detail on what is known of mesonychids and the earlier cetaceans. As we might guess, “The hind limbs of the basilosaurids are greatly reduced relative to those of the earlier cetaceans and mesonychians.” The diagram of the Basilosaurus hindlimb from _Science_ (Gingerich et al.) is shown. “The orientation [of the innominate] is unusual because the pubis is anterior to the obturator foramen, which is anterior to the acetabulum. For this to be the case, the innominate would have to be rotated caudally in conjunction with the elongation of the pubis.” (Right.)

Dorudon atrox has hindlimbs similar to Basilosaurus isis.

Summary: “The extremely reduced hind limbs were not large enough to support the bodies of even the smallest basilosaurids on land.” (And they aren’t connected to the vertebral column by any bony connection, either.)


Basilosaurids are the most advanced archaeocetes. Speaking cranially, they have extensive peribullary air sinuses, loss of M^3 and presence of accessory denticles on the premolars and molars (incipient in G. vogtlensis). Refer back to the other parts of the skeleton, too.

Aquatic features are either very similar, or taken further, in modern cetaceans.

Crania in modern cetaceans are different, due to telescoping of the occipital region over the braincase (mysticetes) or extension of the bones of the rostrum and face backwards over the braincase (odontocetes). This is not seen in basilosaurids.

“The divergence of mysticete and odontocete dentitions (!!) following their origin from the dorudontines is profound and probably related to different styles of feeding”. … “basilosaurids retain cheek teeth that are capable of shearing prey into smaller pieces and wear facets indicate that they chewed their food”.

(Note the rather clear certainty that modern whales originated with the dorudontines. Which is also seen in the final paragraph:)

“All of these further changes in the cranium and dentition of mysticetes and odontocetes were possible after basilosaurids became fully aquatic. The changes in odontocete skulls are almost certainly related to the development of their highly developed echolocation system, which would be ineffective on land. The development of baleen was also tied directly to the achievement of a fully aquatic existence in basilosaurids as mysticetes usually travel long distances in the water to find high-density food sources that are worth their effort to feed on. Additional study of basilosaurids, particularly dorudontines, is needed to further explore (!!) the details of the transition from archaeocetes to mysticetes and odontocetes.”
Prepare yourself for Chapter 3:
“Molecular Evidence for the Phylogenetic Affinities of Cetacea” (a very difficult chapter for a non-specialist to evaluate)

[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