Thursday, April 26, 2012

The "King Lizard" and the origin of cranial asymmetry in whales

 Whales and dolphins are among the most bizarre and derived of all modern mammals, to the point where it took a while for a consensus to develop that they were even mammals. To skip all of the obvious soft tissue features that the average joe can point out, and get to the part most paleontologically minded folks are interested in - the skull of cetaceans is extremely weird in a number of ways. From cranial telescoping and the posterior location of the blowhole, the strangely indistinct orbit, homodont and polyodont dentition (in odontocetes), a series of complex basicranial sinuses, and the straplike jugal bone - the skulls of cetaceans bear little similarity to other mammals. Toothed whales - the suborder Odontoceti - are further perplexing in the asymmetry of their skulls. Baleen whales - mysticetes - are at least more normal in having symmetrical skulls. The asymmetry of toothed whale skulls occurs in a number of ways. Firstly - asymmetry is confined to the facial bones of the skull, and primarily the premaxilla and maxilla (the bones that form the middle and sides of the snout or rostrum, respectively). Secondly, the foramina or holes in the facial region (for nerves and arteries) are differently positioned on each side. Thirdly, bones of the right side of the face are wider than their counterparts of the left side, and the very top of the skull (called the vertex) is accordingly displaced to the left side. Fourthly - in some modern toothed whales, the whole posterior facial region has undergone a clockwise rotation. The skulls of sperm whales and ziphiids in particular are among the most asymmetrical of odontocetes, while porpoises, some delphinids, and the Franciscana (Pontoporia) have what appear to be symmetrical skulls (but still exhibit slight asymmetry).
A new reconstruction of Basilosaurus isis by the folks at University of Michigan, with small and edible child for scale. From

Asymmetry of the odontocete skull has been conclusively tied to their unique mode of sound production. The left nasal passage is unmodified and retained for breathing, while the right nasal passage is hypertrophied and has a host of soft tissue structures and muscles which produce the sounds (i.e. the well-known whistles and clicks of bottlenose dolphins) used during echolocation.  These facial muscles attach to the skull, and because they are larger on the right side - these bones are enlarged on the right side, resulting in asymmetry. And thus, odontocetes exhibit soft-tissue facial asymmetry as well. Cranial asymmetry shows up in different manners in different groups of odontocetes - the basal odontocete Simocetus shows slight asymmetry in the shape and proportions of some of the skull bones, as do certain other Oligocene odontocetes. It has been argued before that cranial asymmetry has evolved multiple times within odontocetes, while facial asymmetry is probably a shared derived feature of all odontocetes. The obvious lack of asymmetry in mysticetes and apparent symmetry of archaeocete skulls suggested for the longest time that cranial and facial asymmetry was unique to odontocetes.
Much to my (and everyone else's) surprise, I saw an abstract at last year's conference on Secondary Adaptations of Tetrapods to Life in the Water (SATLW) by friend and colleague Julia Fahkle on the discovery of cranial asymmetry in archaeocetes. I was skeptical at first - suggestions of looking for asymmetrical archaeocetes and mysticetes were murmured in the mid 1990's regarding the provocative and absurd hypothesis that sperm whales were more closely related to baleen whales (rendering Odontoceti paraphyletic), and I've read elsewhere about people speculating that ichthyosaurs and plesiosaurs may have echolocated and cranial asymmetry should be looked for in these groups (taking the ecological analogy a bit far, I think). When I saw Julia's talk, though, I was surprised, impressed, and dumbfounded- and no longer skeptical. I was happy to see the paper (Fahlke et al. 2011) come out only a couple of months later (although it's taken me about 6 or 7 months to get around to posting about it).
The skull of Basilosaurus isis, from

After taking CT scans of a skull and lower jaws of Basilosaurus isis from the Eocene of Egypt, Julia thought that the twisting of the snout was due to post-burial deformation, and attempted to correct for deformation and modeled the skull to be symmetrical. However, the digital model of the jaws would not close properly - she discovered when the non-modified skull scan was used, the jaws would close properly - suggesting that it was natural. In Basilosaurus, she noticed that the snout was curved a little to the left (insert inappropriate joke here), and that the midline of the skull was deviated to the right side behind the orbits; Fahlke et al. (2011) characterized cranial asymmetry in Basilosaurus as 'curvature and axial torsion of the cranium'. Furthermore - there are thin parts of the lateral wall of the lower jaw called the 'pan bone', which in modern odontocetes are symmetrical in their thickness; In Basilosaurus, the thinnest parts of the pan bone are placed differently - in the left jaw, the thinnest part is placed further forward than on the right. 

Fig. 4 from Fahlke et al. (2011), showing the different placement of the thinnest part of the pan bone in the lower jaws of Basilosaurus isis.

The 'pan bone' is an extremely thin wall of relatively dense bone in the lower jaw, which is placed alongside the very enlarged 'mandibular foramen' (a hole in the jaw that transmits arteries and nerves), which in modern toothed whales is filled with a large lens-shaped body of fat, termed the mandibular fat pad. The enlargement of the bony opening has resulted in the loss of bone on the inside of the jaw - so that looking at the medial surface, there is a boneless window exposing the mandibular fat pad. This structure has been implicated as a key innovation in cetacean evolution as an adaptation for directional hearing in water. Land mammals - including humans - have earbones that are firmly sutured to the skull, and airborne sounds travel slowly and bounce off of soft tissue and are funneled into the ear canal. The 'bouncing' of sound waves is called acoustic impedance, and only occurs when there is a strong contrast in the density of a given material. In water, sounds travel much faster, and because soft tissue is nearly the same density as water - waterborne sounds travel extremely quickly and because there is little to no acoustic impedance between water, flesh, and bone - sounds travel through (rather than bouncing off) soft tissue and bone, arriving at each ear too quickly for the direction of the sound to be determined by the brain. This is called bone-conducted hearing, and most (if not all) non-cetacean mammals hear through this manner when underwater. Next time you're in a pool, try an experiment in functional morphology and close your eyes and have a friend make noises (talking or yelling works), and try to correctly tell the direction of the sound - I challenge you!  The expansion of the mandibular foramen and the mandibular fat pad forms an acoustic pathway to each ear. In addition to this, modern cetaceans exhibit a series of extremely complex air-filled sinuses which surround each set of earbones, and the earbones have lost their bony connections to the skull, which are ways that the ears of cetaceans are acoustically isolated (these sinuses and loss of bony connections appear in baleen whales as well). The pan bone, air sinuses, and earbones separated from the skull are all features that have been previously identified as adaptations for directional hearing underwater.
In her talk last June, Julia showed a slide that further solidified the identification of asymmetry as natural rather than just deformation in archaeocetes - they identified, in a number of protocetid and basilosaurid skulls (roughly a dozen well-preserved specimens from different ages, localities, formations, and countries) - that the direction of torsion and curvature of the snout was the same in each specimen. Fahlke et al. (2011) argued that the appearance of cranial asymmetry in archaeocetes followed closely after the ability to hear directionally underwater, and that it must somehow be related. Pan bones and enlarged mandibular foramina (a bony correlate of a mandibular fat pad) are found in nearly all cetaceans with the exception of Pakicetids; if I recall correctly, the earliest appearance of a pan bone is in the "crocodile-otter" remingtonocetids, and pan bones are known in protocetid and basilosaurid archaeocetes, in addition to archaic mysticetes (read more about that here). Fahlke et al. (2011) argue that the asymmetry in archaeocetes is possibly analogous to that in owls - which enhances the hearing ability of owls in the dark.
Figure 1 from Fahlke et al. (2011).

While Fahlke et al. (2011) convincingly establish archaeocetes as having cranial asymmetry - an extraordinary and provocative discovery - they didn't spend much time in the article contrasting cranial asymmetry of archaeocetes and odontocetes. They do explain that cranial asymmetry in archaeocetes is NOT related to echolocation, as archaeocetes clearly lack many of the facial features of odontocetes. A major point worth stating is that asymmetry in archaeocetes is related to sound reception, while in odontocetes, these types of asymmetry are lacking, and the facial asymmetry is instead related to sound production. While this point was not elaborated on much by Fahlke et al. (2011) - which is certainly no problem anyone should complain about, given the page limits of the journal PNAS in which it was published - it certainly complicates the picture, and if anything - certainly makes early whale evolution much more interesting. It implies that cranial asymmetry might even be decoupled from echolocation altogether, but also that there are now two known modes of cranial asymmetry in cetaceans - longitudinal curvature and torsion, versus facial asymmetry. It suggests that cranial asymmetry predated echolocation and facial asymmetry in odontocetes; perhaps cranial asymmetry as documented by Fahlke et al. (2011) laid the structural blueprints for the derived form of asymmetry in odontocetes. Secondly – Fahlke's discovery implies that mysticetes are secondarily symmetrical, because they evolved from ancestors (basilosaurids) with asymmetrical crania. So then, what is going on in toothed mysticetes? Do they have symmetrical skulls, or not? If not, then why did they lose the asymmetry seen in archaeocetes? Were they already hearing at low frequencies like modern cetaceans, rendering directional symmetry less useful? If not in toothed mysticetes – then when was asymmetry lost in baleen whales? Julia Fahlke's exciting discovery really throws a giant wrench in what previously appeared to be a simpler view of cetacean hearing evolution, and leaves us asking more questions - and most excitingly, a series of weird questions we did not expect.

Edit: There is another hypothesis for cranial asymmetry that has been proposed, which this discovery (among other fossils) demolishes. But, that's a topic for another time.
Fahlke, J.M., P.D. Gingerich, R.C. Welsh, and A.R. Wood. 2011. Cranial asymmetry in Eocene archaeocete whales and the  evolution of directional hearing in water. Proceedings of the National Academy of Science 108:35:13545-13548.

No comments: