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. 12

Synopsis of Chapter 12: Structural Adaptations of Early Archaeocete Long Bones

Sandra I. Madar, Department of Biology, Hiram College, Hiram OH 44234


1.1 Introduction

This decade has provided more of the “appendicular” skeletons of early cetaceans, so: “When coupled with archaeocete craniofacial, dental, and axial remains, a much clearer picture is emerging of the morphological transitions (!!) that occurred during cetacean evolution.” As we know, they are linked (morphologically, at least) to the mesonychians. So Madar expects to see “… series of structural modifications that occurred in a move from complete terrestrial competence, through an amphibious or semiaquatic stage, to the type of highly specialized aquatic locomotion that characterizes modern cetaceans.” So we’ll find out if she finds anything.

She’s examining long bones of both fossil archaeocetes and mesonychians and extant aquatic, semiaquatic, and terrestrial mammals to see how the archaeocete taxa have “adapted (!!) their limbs to an aquatic niche.” Notes that the physical environment affects the mechanical/ anatomical requirements.

Next paragraph: “The move into a novel environment should also result in modifications of the structural properties of the skeletal elements, as the forces transmitted through the bones change.” Why? Animal density is quite important here. “In transitional (!!) taxa, the need to balance buoyancy concerns and load bearing should be reflected in skeletal structural properties.”

Builds on work of Currey and Alexander (1985). Design of tubular bones depends on their role in structural support and locomotion. Structure can be optimized. She’s going to use radiography to look at the distribution of compact bone and features that affect buoyancy control.

1.2 Locomotor Morphology and Functional Constraints of Aquatic Locomotion: Modern Analogues

Shallow-diving mammals and highly aquatic birds have dense diaphyses (= LONG BONE SHAFTS) relative to terrestrial, allowing neutral buoyancy at depth. Allows energy efficiency and minimal physical effort. Gee, I wonder why that would be advantageous to aquatic animals? Analagous to gas bladders in fish.

“Recent” invaders of the aquatic habitat, like penguins and otariids, use several types of hydrostatic controls. Modified lung volume, gastroliths (stomach stones), different levels of bone density. (We’re going to see the terms pachyostosis, osterosclerosis, and pachyosteosclerosis frequently.) Discusses sirenians, which have adaptations that allow approximate horizontal orientation.

1.3 Predicted (!!) Archaeocete Bone Morphology Based on Behavioral Reconstructions

She wants to determine by comparison to extant mammals if the architecture of the long bones is consistent with the locomotor reconstructions.

Ambulocetus: brackish environment, eye orbits facilitating above-water vision (hadn’t heard that before), and hind-limb swimming. Suggested that it might feed like crocodiles: lurking under the water. Should have bone density to aid flotation prevention, and weight-bearing capability.

Remingtoncetids: Amphibious. Diaphyseal morphology “similar to Ambulocetus”? (remember these are predictions). Heavy skeleton more likely than in the contemporary protocetids.

Rodhocetus kasrani: shallow marine environment. Gingerich has already concluded that it was a “pursuit predator” moving via spinal undulation, but likely still had terrestrial competency. Similar to otariids? Either increased bone density, or light skeletons (similar to phocids and modern cetaceans) that utilize hydrodynamic buoyancy control. Based on Georgiacetus, the former appears to be the direction taken. (Preview of what’s coming: Rodhocetus is a mix.)

Basilosaurus isis: a bit problematic, since anguilliform swimming implies that it was not a pursuit predator, though dentition indicates it fed via prey capture. Increased bone density appears to be systemic, as speed and deep diving weren’t part of its lifestyle.

Zygorhiza kochii, Dorudon atrox, Ancalecetus simonsi: We know from the last chapter that these related archaeocetes have vertebral structures very similar to modern cetaceans. SOOO… should have been deep divers and pursuit predators. Ribs are light, so the long bones should be too.


Wow. Table 1 lists the specimens she examined. 42 by my count. Table 11 lists the measurements made. 6 on the femur, 7 on the radius, various on the metapodials and diaphyseal elements. Details count!

Tells how she did the radiographs (X-rays for us normal people). Determined % of given bone cross section devoted to cortical vs. spongy bone using the method of Currey and Alexander. Radiographs used instead of sections to determine cortical thickness.

Some specimens have medullary cavities filled with trabecular bone. *SCRABBLE ALERT* Trabecular means “a small bar, rod, bundle of fibers, or septal membrane in the framework of a body organ or part.” Helps bones to withstand concentrated tensile or compressive forces. Figure 1 shows where this occurs in the barking deer femur and the phocid Mirounga angustirostris (elephant seal). Wish there had been dimensions here: makes me radically change the way I think a femur is supposed to be shaped.)


3.1 Modern Mammals

Modern terrestrial mammals have a tight range of values for various relatable quantities (an example is R/t, the ratio of the mid-shaft radius to cortical thickness). Hippopotami, though having similar values, have higher bone densities due to increased trabecular bone. Tapirs are similar. Polar bears are slightly different than grizzlies or browns. There’s a whole bunch of detail on otters that I’m going to skip.

Otariids and phocids are a bit more interesting. All of the pinnipeds have trabecular diaphyses. Otariids are different from other phocids as they have systemically thicker cortical bone.

3.2 Mesonychians

Most fossils were densely mineralized, preventing clear radiography. But they did appear to have a typical terrestrial structural morphology (good, since they were terrestrial). Nothing remarkable here.

3.3 Archaeocetes

Figure 4 shows diagrams of each of the femur radiograph results, with both an adult and juvenile Basilosaurus represented.

3.3.1 Ambulocetus

Also somewhat difficult to radiograph, but show well-developed cortex and dense distribution of cancellous bone near the epiphyses, “consistent with weight bearing”. Femoral cortical indices are at the “high end” of extant terrestrial mammals and mesonychians = thinner cortical bone at midshaft. Denser diaphyses likely. Manual and pedal metapodials show reduced cortical thickness too, but overall density is high with compact trabeculae along each of the shafts at some point.

3.3.2 Remingtoncetids

Extremely thin cortex near midshaft. Dense trabeculae in the femoral head and neck, oriented obliquely along the medial margin of the neck. Indicates vertical loading of the femoral head. Dalanistes femora had uniform, tightly-packed spongy bone.

3.3.3 Rodhocetus

Some differences here. Femoral diaphysis filled with compact, but thin spongiosa, distinguishable from the endocortical margin. Cortical bone relatively thin midshaft, remains thin proximally and distally (unlike remingtoncetids). Unclear on the femur, but trabeculae do not form solid columns as in otariids, Ambulocetus, and terrestrial mammals.

3.3.4 Basilosaurus

Femora similar to other archaeocetes in lack of open medullary cavities; unique in having extremely thick diaphyseal cortices. Cancellous bone extending into the neck is randomly distributed, lacking patterned orientation of trabecular columns.

3.3.5 Dorudontids

Thin layers of cortical bone lining the diaphyseal proximal margins. Extremely diffuse trabeculae in the rest of the radial diaphysis. No well-defined trabecular columns in the proximal third of either radius.

4. DISCUSSION (already?)

Reminder that modern terrestrial mammals have tight range of relevant ratios. Guess what? Archaeocetes and semiaquatic extant taxa are highly variable! This is interesting:

“Terrestrial species have a diaphyseal structure that optimizes fatigue strength for withstanding repeated loading (values), rather than ultimate strength against impact loading. Values of K (medullary cavity size) for the archaeocetes parallel the range seen in modern semiaquatic lineages of recent origin, such as lutrines (otters).”

This next paragraph rocks: “Aquatic organisms need to balance the demands of maintaining neutral buoyancy at a depth necessary for finding food (tending to require increase skeletal mass), and the need for locomotor efficiency (tending to require decreased skeletal mass as body size increases to limit drag). The limb elements for early cetaceans all appear to have osteological modifications commensurate with highly aquatic adaptations (!!), and in most cases support the predictions (!!) that were made for each taxon based on gross morphology and ecological reconstructions.”

4.1 Mesonychians

Terrestrial. Oh yeah, and “Cross-sectional morphology, combined with diaphyseal proportional indices (Table), indicates that the mesonychians are most similar to ungulates.” Doesn’t appear that they spent much time in the water.

4.2 Ambulocetus

Long discussion:

Sizable trabecular column in the femoral neck; uniformly thick cortex in long bones = terrestrial locomotion was important. Trabecular columns similar to terrestrial species and otariids.

[I realize that I’ve performed a disservice to anyone reading this who is unfamiliar with the pinnipeds, because I had to look this up. Otariids are the “eared” pinnipeds, like the sea lion. They can “walk” on their flippers. Phocids are the earless seals, which have to hump their way along the beach (spinal undulation). Since the difference between these two groups is being repeatedly cited as analogous to the archaeocete evolutionary stages, it’s a good idea to have them in mind for the sake of visualization. Sorry I haven’t said that before.]

Ambulocetus is similar to other semiaquatic taxa in the lack of a medullary cavity. Diaphyseal structure is similar to lutrines, suggesting a hydrostatic response allowing it to overcome neutral buoyancy at the surface. Systemic osteological response of increased bone density in other bones. Madar calls it ballast, aiding submergence.

Lengthened metatarsals and phalanges are longer than mesonychians, consistent with aquatic use. Overall, most similar to lutrines; shows “shifts in appendicular proportions consistent linked to an aquatic mode of locomotion”.

4.3 Remingtoncetids

Femoral diaphyseal architecture is not like phocids, despite ratio similarities elsewhere. More structurally similar to otariids. Again, hydrostatic buoyancy control is likely.

Reduced femora length, moderately so, similar to Ambulocetus and lutrines. Femoral length/vertebral centrum index is intermediate between terrestrial carnivores/ungulates and lutrines. Same capability of hind-limb terrestriality as Ambulocetus and most otariids.

4.4 Rodhocetus

Ratios similar to some pinnipeds, but different cortical bone distribution. High skeletal density is indicated, again for hydrostatic buoyancy control. Terrestrial limb use “intermediate” between phocids and otariids, less terrestrial than Remingtoncetus. Consistent with Gingerich’s predictions of locomotor adaptations. We’ve heard it before, but it only had “limited terrestriality”, perhaps with hind limbs augmented by the axial skeleton.

4.5 Basilosaurus

All features indicate no terrestrial function. Thick cortex in the femoral midshafts is somewhat similar to sirenians and aquatic reptiles. Increased cortical development “surpasses” pinnipeds, lutrines, and semiaquatic ungulates. Consistent with need to maintain neutral buoyancy at shallow depths and maintain horizontal trim.

Increased density (in contrast to conclusions based on rib pachyostosis) indicates that it also (unlike sirenians) had systemically increased skeletal weight. Similar to the usual suspects, though the tendency is “advanced in this archaeocete”. ??? So why with a long skeleton did it have high bone density in the “vestigial hind limb”? Indicates that it didn’t do sustained swimming or deep diving.

4.6 Dorudontids

Diaphyseal structure in forelimbs is similar to modern cetaceans, consistent with hypothesis that doroduntines liked deep water, basilosaurids shallower waters. Possible development of hydrodynamic buoyancy control based on Zygorhiza ribs compared to Basilosaurus.


Figure 5 summarizes the functional implications of the structural characters. Nice diagram; should be on a Web site somewhere!

Ambulocetus probably walked like lutrines (otters, remember).

Remingtoncetids intermediate between Ambulocetus and Rodhocetus; architecture of femora “strikingly similar” to otariids. Indicates “amphibious” lifestyle, not as terrestrial as ambulocetids.

Rodhocetus: as we’ve heard, less terrestrial than remingtoncetids. Skeletal features reminiscent of phocids.

Variable architecture of basilosaurid and dorudontid diaphyses matches morphological differences. The dense nature of the Basilosaurus hind limb is unexpected, given its “vestigial nature” (hmm, 2nd time she said that), and would affect body trim. (*** Would it really? These limbs are not big compared to the whole animal! But I guess if I was swimming and attached a 2 pound ankle weight to my feet, I’d notice the difference. OK.) The rest of the skeleton was systemically dense, too, like sirenians and crocodiles.

*** Insight light! This is why so many Basilosaurus skeletons have been found. They inhabited shallow seas and likely beached or got washed up on shore frequently.

Dorudontids have wholly modern structural characters. This is consistent with the “highly derived locomotor adaptations”. (Is it just me or should Dorudon atrox be getting a lot more publicity?) Diaphyseal morphology implies hydrodynamic buoyancy control (like modern whales).

Results of analysis here support the hypothesis [might even be considered a prediction] of Felts and Spurrell in 1965 that early whales should exhibit heavy bones before significant modifications of articular morphology. Pachyostosis is not of the sirenian form, however.

Discussion of limb ontogeny: skip. Further down: “Thus, the diaphyseal architecture of archaeocetes and aquatic mammals suggests that variables such as diving depth, swimming speed, and metabolic rate may be influencing diaphyseal architecture to a greater degree than transportation costs.”

Structural modifications affecting density occur “early” in archaeocete history. Individually variable; they outline “a progressively increasing dependence on aquatic habitats in these taxa”. Because limb elements seem particularly sensitive to changes in mechanical loading, this type of analysis can be used to “detect architectural differences in fossil material for which locomotor behavior is otherwise ambiguous.”

[Final note: A lot of people wonder how paleontologists can learn so much from the limited material they find. Isn’t this a great demonstration of how they do it?]

Next week, we’ll undertake the sexy and exciting Chapter 13: [parents are strongly cautioned that this chapter contains mature content]

“Evolution of Thermoregulatory Function in Cetacean Reproductive Systems”

(which is also particularly exciting to me because it’s only 16 pages long with lots of half-page diagrams)

[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