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. 6

Synopsis of Chapter 6, “Relationships of Cetacea to Terrestrial Ungulates and the Evolution of Cranial Vasculature in Cete”

Jonathan H. Geisler, Dept. of Vertebrate Paleontology, AMNH, NY, NY

Zhexi Luo, Section of Vertebrate Paleontology, Carnegie Museum of Natural History, Pittsburgh, PA

This is an immense chapter. 49 pages (including 6 full pages of references). Superb and detailed anatomical drawings. Step-by-step descriptions of how each part of the cranial vasculature in terrestrial ungulates has evolved into the cranial vasculature of whales. In short, wow. And it’s fairly hopeless to summarize this in detail, but I’ll do my best.

But first of all, on the first page the authors describe something that I never knew about whales. Rather than the “standard” plumbing of terrestrial mammals, where blood is supplied to the brain via big arteries (carotid and vertebral), whales don’t have them. What they have instead are networks of small arteries and veins, called retia mirabile (“wonderful nets” Cetacean cranial circulation has a series of these.

How come? They’re hypothesized to be an aquatic adaptation. One idea is that they modulate and dampen fluctuations in the flow of blood to the nervous system. Blood pressure, which could rise drastically during diving and damage other tissues, could be prevented by the nets. They could also act as a reservoir of oxygenated blood. Might also be tied into the acoustic system.

They summarize the vascular pattern of cetaceans (8 points to ponder). They summarize the differences in the rete mirabile in odontocetes and mysticetes and the placement of the retes in most cetaceans. They note that fossil bones can reveal quite a bit about the circulatory system, and therefore, they can describe “the transformation of arterial and venous cranial circulation patterns through the terrestrial ungulate – cetacean transition”.

And that’s just the Introduction!

Next step: the “Unresolved Phylogenetic Issues”

Well, it’s the Mesonychians vs. the Artiodactyls again. A much-referenced 1966 paper by Van Valen is apparently the foundation of the Mesonychian ancestry for cetaceans plan. Problems with Andrewsarchus. Van Valen noted that the Mesonychids were “weird” Carnivora and moved them to Condylartha. That got fixed in 1975, when everybody was put into Ungulata. In 1997, to sort this mess out, Cetacea was reduced to a “subordinal” level and they, with Mesonychidae, Triisodontinae, and Hapalodectidae, form Order Cete. Class Acreodi excludes Cetacea. This classification scheme was devised to help minimize paraphyletic groups (which apparently bother both cladists and creationists).

Molecular evidence indicated that the Cetacea sister group is within Ungulata. Evidence to that effect has appeared in previous chapters. The authors summarize and give profuse references. It boils down to this:

“The possibility of Cetacea originating within Artiodactyla has been viewed with skepticism by most paleontologists because of the clear morphological evidence for the monophyly of Artiodactyla to the exclusion of Cetacea.” (refs) “It should be noted that most of these characters occur on the astragalus. Although they appear to be independent, it is possible that they are related and may be overweighted by morphologists.” (Reflective of the comments at the end of Chapter 4.)
Morphological analysis tends to find Cetacea closer to Perissodactyla than Artiodactyla. So obviously questions remain to be answered.

Onto the Material and Methods:

The authors are going to do vascular reconstructions for lots of different specimens. (9 fossil collections are represented.) They are going to address two questions:

1. Are mesonychids the sister group to cetaceans?
2. Are artiodactyls or perissodactyls more closely related to cetaceans?

The authors note “Resolving cetacean phylogeny is not the primary goal of this study; therefore, only taxa that document the main morphological transitions were included: pakicetids, Ambulocetus, Remingtonocetidae, the Cross Whale*, Protocetus, and Basilosauridae. Balaenoptera and Tursiops were used for modern Mysticeti and Odontoceti. In total, 42 different organisms were examined.

*The Cross Whale is described in Geisler, 1996, “A new protecetid cetacean from the Eocene of South Carolina, USA; phylogenetic and biogeographic implications.” in J.E. Repetski, Ed., “6th North American Pal. Conv. Abstracts of Papers, Paleontol. Soc. Spec. Pap. 8, 139. (It’s interesting to me that some exciting finds, such as this and _Georgiacetus vogtlensis_, have come from the USA Piedmont, besides the ubiquitous basilosaurids.) It is a partial skull and postcranium from the Cross Quarry near Charleston, SC.

PAUP 3.1.1. was used to find the most parsimonious trees. 26 ingroup and 3 outgroup taxa were scored for 80 (yes, eighty!) characters. More on phylogenetic methods is provided, but nothing of particular excitement for the layperson.

The next section is of interest, but I’ll only describe what it’s about. It’s about how soft-tissue structures are inferred from fossils (primarily bones). There are lots of marks and some structures that are utilized by the vascular system. I.e., soft tissue features have “osteological correlates”. Again, amazing what can be learned from fossils.

Soft-tissue reconstruction for extinct taxa uses the “extant phylogenetic bracket” method, which uses “… a known phylogeny of extinct and extant taxa to reconstruct soft tissues in extinct taxa. Extant taxa are first surveyed to determine osteological structures that always co-occur with a specific soft-tissue feature. A vascular or soft tissue structure can be reconstructed in an extinct taxon with a high degree of confidence if: (1) its unique osteological correlate is present, and (2) its two closest extant relatives have the soft tissue structure and osteological correlate”. This makes me go “Hmmmm” and again amazes me with how much can truly be learned from fossils, such as how some soft-tissue structures appeared as organisms evolved.

Results of the phylogenetic analysis:

  1. Cetacean monophyly strongly supported
  2. Mesonychidae should be defined as all taxa more closely related to Mesonyx obtusidens than to the crown group of cetaceans (this excludes Andrewsarchus and Hapalodectidae)
  3. Clade Acreodi is defined as most recent common ancestor of M. obtusidens and extant cetaceans, plus all of its descendants. (Similar to Prothero et al., 1988). Acreodi contains two monophyletic groups, Mesonychidae and Cetacea.
  4. Vascular reconstruction “in basal members of Cete” requires a Cetacea+Artiodactyla clade, termed “Paraxonia” = the most recent common ancestor of Artiodactyl and Cetacea and all of its descendants.
  5. Unlike molecular studies (!!) Artiodactyla is found to be monophyletic, with four unequivocal synapomorphies. (Only one of the four is not apparently based on astragalus, however.)

Still some work to be done.


So far we’ve covered pgs. 163-174. Since we still have to get to pg. 212, there’s still some ground to cover. BUT, much of this is detailed diagrams with even more detailed captions. And most of it is TOO detailed for me to summarize. So what I’ll do is describe each feature that is being considered, and try to find interesting evolutionary implications.

Arteries 1. Proximal stapedial artery

Requires a stapedial foramen

“A patent stapedial foramen occurs in basilosaurid archaeocetes, the sister group to the cetacean crown group. Therefore, a reduction of the stapedial foramen in extant cetaceans is interpreted as an apomorphy of these taxa and not the primitive condition of Cetacea.”
The proximal stapedial groove is also absent in extant and extinct cetaceans.

2. Superior Ramus of Stapedial Artery

Absent in all extant cetaceans. Could have been lost at any time between the origination of the cetacean crown group and the first appearance of the earliest cetaceans.

3. Ramus anastomoticus

Found in some odontocetes, possibly in mysticetes, suggesting that it should be found in archaic cetaceans.

4. Spinal meningeal arteries

Neomorphic vessels in Cetacea that are embedded in the rete mirabile of the vertebral canal. The authors compare odontocetes and mysticetes; this is difficult to follow because they keep switching between species names and odontocete/mysticete designations. The Cross Whale does not necessarily show spinal meningeal arteries. The Cross Whale is “Indocetus-grade”.

5. Branches of Occipital Artery

The location of osteological structures in Pakicetus, artiodactyls, and Equus is in three different locations. The “arteria diploetica magna” could supply blood to the brain, and could be an evolutionary intermediate between cranial supply via the carotid and vertebral arteries and the cetacean “derived” condition, via spinal meningeal arteries and a spinal rete. There is some similarity to Hippopotamus amphibius in this regard.


Sorry, you just have to see the diagrams (the Cross Whale is great). Unless you want me to try and describe the capsuloparietal emissary vein (the CEV). Note, however, that the CEV is greatly reduced in stem cetaceans, and therefore likely reduced in the most recent common ancestor of mesonychids and cetaceans. That’s because the CEV has disappeared in odontocetes and _adult_ mysticetes, and the connection to the transverse sinus is shifted posteriorly compared to basilosaurids.

Endocranial Retia Mirabile and Related Vessels

This is somewhat more interesting, so I’ll try to do it justice. Only 9 pages to go, some with very illustrative diagrams, so I can do better here, I think.

Again, what are retia mirabile? — complex nets of small anastomosing vessels/ arteries, veins, or both; endocranial are intradural within the cranial cavity, may be continuous with extracranial retia

The rostral rete mirabile occurs in all extant cetaceans

The caudal arterial rete is present in mysticetes

Some ungulates have a ventrally confined caudal arterial rete

How is it identified in fossils?

“A hypertrophied endocranial rete can be identified in an endocast or skull by the presence of extraneous endocranial space that cannot be accounted for by the brain, nerve ganglia, or the spinal cord”
(ASSUMING parts of the central nervous system weren’t appreciably larger back then compared to now).

The caudal rete mirabile is an obvious autapomorphy of extant cetaceans, traceable back to protocetids. It was already developed in basilosaurids to the same extent as mysticetes (this includes _Dorudon atrox_).

Page 194 has a diagram (Fig. 10) comparing Mesonyx, Indocetus, and Basilosaurus. In Mesonyx, only a small rostra rete mirabile is located next to the olfactory lobe below the cerebrum. The vertebral artery runs along the spinal cord. In Indocetus, a large caudal rete mirabile is next to the cerebrum and merges with the rostra rete mirabile, which is slightly larger, and a small spinal rete mirabile has formed parallel to the vertebral artery. In Basilosaurus, the caudal rete mirabile has expanded (now larger than the cerebrum behind the cerebellum), as has the spinal rete mirabile, and all of the circulation is routed through the ramus anastomoticus to the common carotid artery. Essentially the vertebral and internal carotid artery have disappeared in Basilosaurus (the internal carotid is uncertain in Indocetus).

Cool. And progressive (dare I say gradual?)


6 pages to go

The first thing that the authors do is describe in excruciating detail the “hypothetical ancestral cranial vasculature” of the Paraxonia (remember, that’s Cetacea+Artiodactyla). They do this by referring to Figure 2, a diagram of Diacodexis (an artiodactyl).

– the internal carotid artery is patent and a major supplier of blood to the brain
– the internal carotid artery and the ramus anastomoticus supply the rostral arterial rete (checking Fig. 10, yes, that’s what it shows)
– the “circle of Willis” (the immediate blood supply to the brain) recieves arterial blood from (1) a branch of the internal carotid emerging from the rostral rete and (2) from the basilar artery (linked to the vertebral artery and branches of the occipital artery)
– veins and sinuses are also discussed

Now for Acreodi: (most recent common ancestor of M. obtusidens and extant cetaceans, plus all of its descendants.)
– internal carotid in same position, though fainter sulcus suggests decrease in girth (“sulcus” is the groove for the artery. Do you realize how much better at Scrabble I’m going to be after this book?)
– reduction in the internal carotid compensated by one or any combination of 3 vessels: “the arteria diploetica magna via an anastomosis [joining] homologous to the anterior end of the spinal meningeal arteries, the ramus anastomoticus from the external carotid, and the basilar artery via the vertebral artery and branches of the occipital artery.” I think I can see that in 10b (which doesn’t show occipital arteries)

Now for Cetacea.

Based on pakicetid cetaceans:
– size of internal carotid is uncertain. Groove is poorly defined in pakicetids, absent in other cetaceans
– alisphenoid canal is absent; the maxillary artery is extracranial.
– appears that the arteria diploetica magna was anastomosed to (and supplied by) the rostra rete mirabile
– circle of Willis supplied by the carotid rete (as in extant cetaceans) and possibly by the basilar artery; the latter is tentative

Post-pakicetid cetaceans
– spinal arterial retia appear to anastomose with the endocranial portions of the arteria diploetica magna (each side)
– post-pakicetid, pre-crown cetaceans have space for a vascular rete mirabile in the posterior cranial fossa
– development of the arterial caudal rete is followed by a decrease in the diameter of the transverse canal for the vertebral artery in the axis vertebra (notable in basilosaurids)
– the cervical arterial rete mirabile, lateral to the vertebral column in extant cetaceans, is probably homologous to the vertebral arteries

A discussion of how the retes interacted with the air sinuses as aquatic adaptations concludes this chapter. Note:

“Before the development of a separating bony ridge or lamina, the expansion and contraction of the air sinuses during diving could have been lethal, if there was no retial tissue between the petrosal and the cerebellum.” Hmmm. Diving adaptations developed from the existing air sinuses.

Actually, most of this has been seen before: Paraxonia, monophyletic Mesonychidae, Acreodi. See way above.

Vertebral arteries, found in terrestrial ungulates and likely in protocetids, have been lost by the basilosaurid stage.

The large arteria diploetica magna in mesonychids and pakicetids “formed a transition between cerebral blood supply via the internal carotid artery, the plesiomorphic condition, and via the spinal meningeal arteries or homologous arterial rete mirabile as in extant cetaceans”. (plesiomorphic = primitive character state)

The replacement of arteries as the primary blood supply by the arterial retia began to appear in protocetids and was fully established in basilosaurids (before the appearance of “modern” cetaceans)

Final note: I have frequently noted the gradual reduction of the hind limbs in the terrestrial mammal – whale sequence as a notable evolutionary trend. Yet it seems that many skeptics are seeking the development of “new and novel” features and functions. Well, the function of the rete mirabile as the primary blood supply route to the brain appears to fill that request, and this function is cited as an adaptation to an aquatic lifestyle that included diving behavior.

Thanks be to God! (especially now that I’m done with Chapter 6)

Chapter 7 is much more paleontological in perspective: “Middle Eocene Cetaceans from the Harudi and Subathu Formations of India” (and it’s only 20 pages long!)

However, Chapter 8: “Postcranial Osteology of the North American Middle Eocene Protocetid Georgiacetus” is 32 pages long and very detailed. But I am certainly not going to skimp on a chapter whose summary paragraph contains the phrase “… fills the morphologic gap…” You’ll just have to keep reading these to find out what fills what gap.

[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