Friday, July 19, 2013

Field notes I: 2013 - summer of Tyrannosaurus rex

Right frontal in dorsal view of a juvenile Tyrannosaurus rex (DDM-344.1) from the Hell Creek Formation of southeastern Montana that I found last August. Rostral is to the right, medial is toward the top of the image. Abbreviation: Dinosaur Discovery Museum, Kenosha WI.
This series of posts will describe the upcoming Carthage College Expedition to the Hell Creek Formation of southeastern Montana. We have been collecting there since 2006 on exposures that are on sections managed by the Bureau of Land Management, and we leave for there in the morning!
Little Clint Quarry
This year we (myself, my technician, 10 students, 2 field assistants, 2 volunteers) resume excavation of a multitaxic and monodominant bonebed, the Little Clint Quarry (named for Thescelosaurus expert Dr. Clint Boyd).
Little Clint is the nickname of a partial skeleton of a juvenile T. rex that led to the bonebed. So far, we have collected a frontal, rib, tibia, and pedal phalanx of the specimen. We have also found a partial maxilla, a pair of tibiae, and a metatarsal III (see image below) belonging to a large T. rex, and a tooth and pedal phalanx of a medium-sized T. rex that we have nicknamed ‘Big Clint’.
In addition to T. rex, we have recovered a complete hadrosaurid fibula, a partial skeleton of Triceratops that includes cranial and postcranial bones, and two Thescelosaurus femora. The bonebed also includes nondinosaurian amniotes, including isolated bones of crocodilians and turtles. In some parts of the quarry the dinosaur bones are so densely concentrated that they are stacked upon each other. This year we will resume excavation of this lag deposit, and I am optimistic that we will recover at least a few additional bones of the large T. rex.
Partial left metatarsal III (DDM-35.131) in anterior view of a large Tyrannosaurus rex from the Little Clint Bonebed.

Little Hyslop Locality
Last summer I went prospecting with one of my long-standing volunteers, Mr. Andy Prell (Kenosha, WI) a few hours after the students left for home. About 15 minutes from camp I walked up a ravine, and to my astonishment, I saw a small tyrannosaurid frontal bone lying upside down on the slope (see image above). Andy joined me a few minutes later and searched upslope, where he found a partial pedal ungual (D I-1) of a small T. rex. Moments later I found a small T. rex tooth a short distance north of the frontal (see image below).
The bones and tooth correspond to the same size of animal, between 3 and 6 meters long, and so there is a good chance that at least a partial skeleton lies below the surface. Andy and I searched intently on the hillside for more evidence of the skeleton, without success. Several days later I returned to the locality with our team of volunteers, but again without finding anything else. I hope that this year’s spring rains have brought more of the specimen to the surface, unless Andy and I found the last of it. The Little Hyslop locality is named for Mr. Dan Hyslop (UW-Madison graduate).
The tooth (DDM-344.3) of a juvenile Tyrannosaurus rex that I found at the Little Hyslop Locality. It is consistent in size with the frontal bone, which suggests that there is more of the specimen to find.

I will make the best effort to keep you posted on our progress in the field, as far as T. rex discoveries are concerned!

Sunday, July 14, 2013

Q&A I: Teratophoneus curriei ontogeny

The skull of the type specimen of Teratophoneus curriei in left lateral view, modified from Carr et al. (2011).
If I have sufficient time, I’ll answer questions put to me directly that illuminate aspects of tyrannosauroid paleobiology. Recently I was asked about Teratophoneus curriei, a new genus and species of tyrannosaurine that I named recently (2011) with Thomas Williamson, Brooks Britt, and Ken Stadtman. These questions come from a reader in South Korea.
Q: If the base of the interfenestral strut (see diagram) in T. curriei is concave, and that is considered to be a Stage 3 (adult) adult feature of Albertosaurus libratus (Carr, 1999), then why in T. curriei is it considered to be a subadult feature?
A: In the article our goal was to establish the relative maturity of the holotype specimen based on my earlier work (Carr, 1999) on other tyrannosaurids, particularly A. libratus. Since we were working on the only specimen available to us at the time, we could not establish its maturity with precision. In addition to a high number of immature features, it turned out that the specimen had several that clearly indicated that it was neither a young juvenile nor a full adult. So what about that nagging strut?
The table in Carr (1999) shows that the strut is flat in small stage 1 (i.e., juveniles), whereas it is concave in large stage 1 specimens. Ergo, the presence of a concave strut in T. curriei is consistent with what is seen in relatively immature A. libratus. However, in A. libratus, this is not a perfectly clean pattern, because individuals of greater maturity (Stage 2; subadults) also have the flat condition, whereas the concave condition is seen in the most mature specimens (Stage 3; adults). So where does that leave us?
Keep in mind that in Carr (1999) I established growth stages – large intervals of ontogeny - that would assist in distinguishing ontogenetic variation from phylogenetically informative variation. However, many specimens can be grouped into the categories of small juvenile, large juvenile, subadult and adult, indicating that growth stages have limited resolution because the relative maturity of the specimens within those categories is not specified. This indicates that the growth stage categories themselves are arbitrarily defined. Is there a way to solve these issues of resolution and arbitrariness?
Clearly this is an area where a more rigorous approach to ontogeny is required, which, in my view, is solved by cladistic analysis of ontogenetic characters (Carr and Williamson, 2004; Carr 2010). It is only through this approach, which solves both problems in one stroke, can we distinguish ontogenetically informative variation from individual variation. I have work in progress for all of Tyrannosauridae, which I have presented over the past several years at the annual meeting of the Society of Vertebrate Paleontology.
In the meantime, I suspect that the concave condition does indicate a relatively mature condition (it is not seen in the smallest, presumably least mature, juveniles), but that its timing is individually variable, somewhat like tooth eruption in people (for example, none of my ‘wisdom teeth’ have erupted and I am middle aged!). We’ll have to wait and see what the results of the analyses show. I’ll be able to answer your question with more clarity and depth once that work is published.
Q: Teratophoneus has a low tooth count, a condition that reflects its short snout. What do you expect the tooth count to be at the opposite extremes of its ontogeny?
A: One trend in tyrannosaurids is to increase tooth count, then reduce it (Carr, 1999); in others the tooth count is somewhat stable (Tsuihiji et al., 2009). I expect that T. curriei will show the latter pattern, where it will show little if any variation in tooth count. The short snout imposes a limit on the number of alveoli (tooth sockets) early in ontogeny, and I do not expect that would change as the animals increased in size. However, if T. curriei increased the size of its teeth in the manner of T. rex, then I would expect a reduction in tooth count with increasing maturity. Presently the answer awaits new specimens.
References cited
Carr, T. D. 1999.  Craniofacial Ontogeny in Tyrannosauridae (Dinosauria, Coelurosauria). Journal of Vertebrate Paleontology 19:497-520.

Carr, T. D. 2010. A taxonomic assessment of the type series of Albertosaurus sarcophagus and the identity of Tyrannosauridae (Dinosauria, Coelurosauria) in the Albertosaurus bonebed from the Horseshoe Canyon Formation (Campanian–Maastrichtian, Late Cretaceous). Canadian Journal of Earth Sciences 47:1213-1226.
Carr, T. D. and T. E. Williamson. 2004. Diversity of Late Maastrichtian Tyrannosauridae from western North America. Zoological Journal of the Linnean Society 142:479-523.

Carr, T. D., T. E. Williamson, B. B. Britt, and K. Stadtman. 2011. Evidence for high taxonomic and morphologic tyrannosauroid diversity in the Late Cretaceous (Late Campanian) of the American Southwest and a new short-skulled tyrannosaurid from the Kaiparowits formation of Utah. Naturwissenschaften 98:241-246.

Tsuihiji T, M. Watabe, K. Togtbaatar, T. Tsubamoto, R. Barsbold, S. Suzuki, A. H. Lee, R. C. Ridgely, Y. Kawahara, and L. M. Witmer. 2011. Cranial osteology of a juvenile specimen of Tarbosaurus bataar (Theropoda, Tyrannosauridae) from the Nemegt Formation (Upper Cretaceous) of Bugin Tsav, Mongolia. Journal of Vertebrate Paleontology 31: 497–517.

Osteology IV: Craniofacial skeleton in dorsal view

The bones of the craniofacial skeleton of Albertosaurus libratus seen in dorsal view, © Dino Pulerà. The bones of the palate are not shown (see text for details). Carbon dust plate by Mr. Dino Pulerà.
This image shows the bones of the braincase and facial skeleton in dorsal view; it does not show the quadrate or the palate. The missing bones reflect the limited amount of material that I had to work with at the time I drafted the line drawing on which this rendered imaged is based. However, in future posts of this series each bone will be featured in multiple views, which hopefully will compensate for the deficits in this otherwise informative image.
Basioccipital: In dorsal view, the basioccipital is represented by a sliver of bone that extends along the dorsal midline of the occipital condyle (the ball joint that articulates with the first vertebra of the neck) between the overlying otoccipitals. The basioccipital is a single unit and midline bone.
Frontal: The paired frontal bone contacts the nasal rostrally, the prefrontal rostrolaterally, the lacrimal laterally, the postorbital caudolaterally, the parietal caudally, and its complement medially. It contributes a slip to the margin of the orbital fenestra and forms a part of the rostromedial margin of the dorsotemporal fenestra. The frontal forms most of the interorbital region of the dorsal skull roof and the rostral part of the temporal region.
The dorsotemporal fossa covers approximately the caudal half of the bone, whereas the rostral half is covered by smooth subcutaneous surface. The dorsotemporal fossa is a depression with an irregular surface that functioned as the origin of attachment for the adductor (jaw closing) muscles in life. The triangular rostral end of the bone that lies outside of the fossa is informally termed here the forehead.
Jugal: In dorsal view, the jugal is a mediolaterally narrow strap that bounds the orbital fenestra ventrolaterally. It extends from caudolaterally to rostromedially along its course from the wide temporal chamber caudally to the base of the snout rostrally. The temporal part of the jugal extends nearly directly caudally, whereas the orbital region extends rostromedially. The jugal gives the orbital fenestra the rostral vector of its orientation. Caudally the jugal is overlapped laterally by the quadratojugal, whereas rostrally it overlaps the caudal end of the maxilla as a narrow wedge. The postorbital and lacrimal conceal the jugal’s contact with those bones.
Lacrimal: In dorsal view, the lacrimal is a large block-like structure with a rapidly tapering rostral process. Caudally the lacrimal inserts into the frontal, medially it extends along the prefrontal, and rostrally it inserts into the dorsolateral surface of the nasal; however, its ventral contact with the jugal is blocked from view by the lacrimal and its contact with the maxilla is out of the plane of view.
The lacrimal is a dominant structure above the transition between the orbital region and the base of the snout. It is widest in the region of two confluent ornamental structures, the caudolateral shelf and the cornual process. The shelf extends laterally over the orbital fenestra, whereas the cornual process extends rostrolaterally between the orbital and antorbital fenestrae.
Laterosphenoid: The laterosphenoid is paired bone that forms the rostrodorsolateral corner of the braincase. In dorsal view, it forms the rostromedial margin of the dorsotemporal fenestra. It contacts the frontal rostrodorsally, the parietal caudodorsally, and the prootic caudally. The part that can be seen forms a laterally extending ridge that is situated above the joint surface for the epipterygoid, a bone of the palate; as such, the laterosphenoid shielded that delicate bone from the adductor musculature.
Maxilla: In dorsal view, the maxilla forms nearly the entire length of the lateral surface of the snout; it is excluded from the dorsum of the snout by the premaxilla and nasal. Caudally the maxilla is overlapped by the jugal, dorsally by the nasal, and rostrodorsally and rostromedially by the premaxilla. Although the maxilla does contact the lacrimal caudodorsally, it disappears below the nasal above the antorbital fenestra. Like the jugal, the maxilla slopes slightly ventrolaterally, bringing it more broadly into view.
Below the jugal, the caudalmost part of the maxilla extends caudolaterally toward the orbital region. In this region the maxilla extends further laterally than it does medially. When seen from above, the maxilla forms the ventrolateral and rostroventral margins of the antorbital fenestra.
Nasal: The nasal is the longest bone of the dorsal skull roof, where it extends from the rostral part of the orbital region to nearly the tip of the snout. The nasal contacts the frontal caudally, the prefrontal caudolaterally, the lacrimal laterally and dorsally, the maxilla ventrally and rostrolaterally, and the premaxilla rostrally. Between the lacrimal caudally and the maxilla rostrally, the nasal spans the width of the snout above the antorbital fenestra and part of the antorbital fossa.
In dorsal view, the nasal is seen to be a long and narrow structure that forms the dorsomedial margin of the antorbital fenestra and the caudolateral and caudomedial margins of the bony naris. The nasal bears the coarsest region of the rugose ornamental surface of the facial skeleton, a condition that is also seen from the side.
Otoccipital: The otoccipital is an extensive bone mediolaterally and dorsoventrally, such that it requires two labels here. The otoccipital contacts the basioccipital ventrally, the supraoccipital dorsally, the parietal dorsolaterally, and the squamosal rostrally.
The otoccipital forms nearly all of the occipital condyle in dorsal view, each side separated from each other by a narrow midline strip of the basioccipital. The otoccipital extends at a low caudolateral angle behind the squamosal, distally forming the caudolateral corner of the skull.
Parietal: In dorsal view, the parietal is an extensive bone that contacts the frontal rostrally, the laterosphenoid rostroventrally, the prootic caudoventrally, the squamosal caudolaterally, the otoccipital caudoventrally, and the supraoccipital caudally. The parietal is widely exposed between the dorsotemporal fenestrae; in contrast, it is narrowly exposed along and above the occiput, where it extends mainly dorsally in the vertical plane as the tall nuchal crest. The parietal is the widest of the bones of the dorsal skull roof, which reaches its greatest breadth across the caudolateral processes that extend between the squamosals and otoccipitals.
The rostral part of the parietal lies entirely within the dorsotemporal fossa. The sagittal crest separated the complementary fossae along the dorsal midline. In this region, the parietal is dorsoventrally deep, curving lateroventrally to the laterosphenoid, prootic, otoccipital, and squamosal. Depite its depth, it only forms a short rostromedial extent of the margin of the dorsotemporal fenestra.
The caudal surface of the nuchal crest received vertebrocranial musculature, whereas its rostral surface anchored adductor musculature. In life, muscular contractions on the nuchal crest from behind lifted the head, which pivoted on the occipital condyle.
Postorbital: In dorsal view, the postorbital is a slim bone that extends caudolaterally from the frontal to the squamosal. The postorbital contacts the frontal rostromedially and the squamosal caudomedially.
The postorbital delimits the dorsotemporal fossa rostrolaterally by a ridge, which does not extend onto the frontal. Although it is tilted out of the plane of view, the postorbital overlaps the lateral surface of the squamosal. In the image it appears that the postorbital overlaps the squamosal medially; but this is not the case – a V-shaped notch splits the rostral end of the squamosal in this region, exposing the postorbital to view ahead of the laterally overlapping contact.
As seen from above, the cornual process of the bone can only marginally be seen. In contrast, the dorsotemporal fossa is the most widely exposed part of the bone in this view.  Behind this, the postorbital forms the mediolaterally narrow rostral end of the upper temporal bar.
Prefrontal: The prefrontal is divided into two processes, and the dorsal of these is seen from above, where it is situated between the lacrimal laterally, the frontal caudally, and the nasal medially. The prefrontal is the smallest of the dorsal skull roof bones and it is located at the level of the rostral end of the interorbital region. This bone is shaped somewhat like a teardrop, where the wide end is positioned caudally and the tapering end points rostrally, pinched between the lacrimal and nasal. The prefrontal is smooth and is not coarsened by the rugose ornamental surface that is seen in the lacrimal and nasal.
Premaxilla: In dorsal view, the premaxilla forms the front of the snout, where it articulates with the maxilla laterally, and the nasal caudolaterally and caudodorsally. It forms the rostroventral and rostral boundaries of the bony naris. Most of the dorsal surface of the bone is flattened and smoothed by the narial fossa, except its rostral surface is covered by the coarse subcutaneous texture. In Albertosaurus libratus, as pictured here, the tips of the separate premaxillary diverge from each other such that a wedge from the nasal separates them; in other taxa these processes are apposed to their tips.
Prootic: Only a narrow slip of the prootic can be seen in dorsal view, where it forms the caudomedial margin of the dorsotemporal fenestra. This part of the bone forms a ledge that would have deflected adductor musculature away from the space and contents of the middle ear.
Quadratojugal: In dorsal view, the quadratojugal caps the caudolateral corner of the craniofacial skeleton. The vertical stalk of the bone is seen, as well as its long rostral process that extends onto the lateral surface of the jugal. As such, the quadratojugal forms the lateral half of the lower temporal bar.
Squamosal: In dorsal view, the squamosal completes the mediolaterally narrow upper temporal bar; it contacts the postorbital rostrolaterally, the otoccipital caudomedially, and the parietal rostromediodorsally. The medial process of the bone is wedged in a groove between the parietal dorsally and the prootic ventrally. The bone widens caudally, where it forms the flat surface of the dorsotemporal fossa, which is bounded laterally by a ridge.
Supraoccipital: The supraoccipital is a single unit midline bone; it contacts the otoccipital ventrally and the parietal rostrally. In dorsal view, it forms the midregion of a transverse (mediolateral) bar that extends caudally above the foramen magnum and lower half of the occiput (caudal surface of the braincase), and it forms a rostrodorsally-inclined rectangular block that extends a short distance up the midline of the nuchal crest.

Wednesday, July 10, 2013

Research as it happens II: MOR round-up

Clockwise from left to right: the first draft of the monograph with marginalia, laptop with manuscript in progress, premaxillary teeth and the first maxillary tooth from the new species, loupe, and digital calipers.
I’ve put in full days over two weeks in the collections of the Museum of the Rockies (MOR; Bozeman, MT). In this entry I’ll provide a summary of what I’ve accomplished, what loose threads remain, and a description of the challenges in writing a monograph on limited time. As in the first entry, this follows a question-and-answer format.

How does the MOR collection compare to others?
For this specific project – the description of a new taxon – the MOR collections are among the best of the best for several reasons:

(1) There are multiple specimens of the new species.
Multiple specimens means that individual variation can be distinguished from ontogenetically- and phylogenetically-informative variation. A taxon based on a single specimen inevitably leads to false positives regarding assessments of the significance of features. For example, relatively primitive features turn out to be typical of juveniles, or unique features turn out to be individual variation after the sample size increases.

(2) A growth series is represented.
Ontogeny is at the core of my research program, and a solid understanding of ontogenetic variation is central to hypotheses of diversity and phylogenetic characters. In the monograph, each paragraph includes a description of ontogenetic variation, where possible; this approach lends a description the proper depth with which to identify features that are truly diagnostic of a new species. It sometimes turns out that specific growth changes can also diagnose a species.

In circumstances where I work without a growth series, I feel utterly blind regarding the data I collect – I simply do not know what it means, and it is a disquieting experience. Thankfully at the MOR I do not have to worry that I am working in the dark.

(3) The skulls are in some cases articulated, whereas others are disarticulated.
Although a completely articulated skull is a marvelous spectacle in a museum gallery, to a paleontologist’s eye it is a lot of missing data. Why? All of the surfaces of contact between apposed bones cannot be seen, and such surfaces contain information regarding ontogeny, phylogeny, function, and size.

The growth series that I have been working on includes a nearly completely articulated skull that thankfully is separated into several large sections. In this condition the specimen is much more maneuverable than a single unit of bone, which lends itself to a more complete and accurate description. The growth series also includes a completely disarticulated skull and skeleton, which gives me access to every surface of each bone that are otherwise blocked from view in the articulated skull. Finally, the series includes isolated bones that also provide a complete view of every bone.

(4) The skulls have associated postcrania.
Although much of my published work has focused on the craniofacial skeleton, I also work extensively on the postcranium. Fortunately, several of the skulls I am working on have associated postcranial skeletons in varying states of completeness. This provides the description with an appropriate osteological scope, and permits the identification of the various types of variation. This provides a more complete data set for quantitative phylogenetic- and ontogenetic analyses.

(5) There are multiple specimens of multiple taxa.
The MOR has more than one species of tyrannosaurid in its collection against which I can check to see if the features that I am describing in the new species are in fact unique. On several occasions I have found that a feature that I thought was diagnostic is seen in other species, whereas I have been able to identify features that are novel to the new species. Without a collection of this comparative scope, I would be utterly ignorant of the significance of the features that I document regardless of having the growth series.

With each of these qualities considered, the MOR is toe-to-toe with the collections at the American Museum of Natural History (New York, NY), Canadian Museum of Nature (Aylmer, QC), and the Royal Tyrrell Museum of Palaeontology (Drumheller, AB), at least when tyrannosaurids are considered.

Is there a cost to such a fantastic state of affairs?
Unfortunately, there is a cost to everything; several excellent specimens of skulls and postcrania that constitute a growth series means a tremendous amount of time must go into the research project to recover the data. For instance, on this project I have been writing up at least three specimens for each bone, which triples the amount of work and time that is spent, in contrast to cases where only one specimen is available.

This research trip will provide a baseline that I can use to estimate the time of completion for similar descriptive projects. So far, two whole months have gone into this project and I predict that it will take between one and two weeks to finish the collections-based part of it, which is longer than I expected that component to take.

What sort of flexibility do you allow yourself?
Although I do arrive at a collection with a plan to follow, I have to stay flexible in case time does become available to expand sections or add new bones. For instance, although I had planned for the monograph to stay focused on the skull, my progress on the latest trip was brisk enough that I did expand it to include the postcranium of what will become the type (name-bearing and reference) specimen. I then limited the comparisons with other specimens to that set of bones. I will also sometimes move ahead to writing up other bones if the inertia of having worked on the same bone for two days gets the better of me.

What did you get done, really?
Upon arrival, the monograph was 426 pages long (double spaced), and on the last day I had increased it to 611 pages – 185 pages written over two weeks. In addition to that, I took approximately 1157 measurements over one-and-a-half days; this is an overestimate since not all specimens are complete, but I did check each set of landmarks on each side of every specimen where necessary. I added 45 pages of tables, and over the space of one day I took 824 photographs.

What challenges do you face?
I think that the most significant challenge on any research trip is intrinsic, not extrinsic. It may come as a surprise that such work could be grueling, but the initial excitement of writing up a new species based on excellent fossils does become displaced by the tasks required to produce a work that meets the standards of the field. A description is a serious task that requires close attention and an ability to focus on several objectives at once. These include writing the basic description of what I see, taking note of ontogenetic variation, accommodating what has been said previously in the literature, quantifying features, etc.

Although writing an osteological description is primarily a task of the mind, it is tiring. My best time of day is in the morning, even though I am not a morning person. I can usually work for two days with high energy before I hit the ‘brick wall’ in the early afternoon of the third day and the others that follow. I experience the brick wall as an overwhelming sensation of inertia regarding the task before me and I can no longer sit still and write. As a remedy, I pace around to rid myself of the nervous, distracting energy. The best way that I combat the brick wall is to take advice given to undergraduate students when they study for an exam – don’t take on the task for several hours straight, instead take a break for 10-15 minutes every hour.

What advice do you have for students who are starting their first descriptive work?
Approach the fossil in a hierarchy from general to specific for the entire skull and then for each bone. After writing the description for each part, take measurements, followed by photographs. The idea is to complete one section before moving to the next. I’ll give an example of this approach in the bone-by-bone Osteology posts. The key is to follow an organized sequence; however, depending on the project, and the amount of material and time available to me, I’ll delay taking measurements and photographs until the last two or three days.

How is this work funded?
Small-scale projects like this do not qualify for large grants, such as those offered by the National Science Foundation (NSF). However, my travel costs for my first two trips to the MOR have been defrayed by small travel grants from my institution, Carthage College. On the occasions where those funds are not made available, I have carried the cost out of my own pocket.

What’s left for this project?
In terms of the collections work, several tasks are ahead of me: (1) expansion of the description of the pterygoid, (2) denticle counts for the mesialmost maxillary teeth of the type specimen need to be obtained, (3) the comparative description of the postcranium requires completion, (4) photographs of the postcranial bones of referred specimens need to be measured and photographed, (5) the subcutaneous surface of the ornamental cranial bones needs to be described, (6) the distribution of lesions are presently undocumented, and (7) several observations need to be included, based on the marginal notes in the first draft of the manuscript. Nearly there!

Monday, July 8, 2013

Osteology III: Craniofacial bones in lateral view

The bones seen when a tyrannosaurid skull is viewed from the side, © Dino Pulerà.The mesethmoid is not included (see text for explanation). Carbon dust plate by Mr. Dino Pulerà.

When the skull is considered as a single functional unit, it is clear that aside from their flexible interfaces with each other, bones as parsed segments are an illusion. However, dinosaurs and all other osteichthyans are descendants of a common ancestor whose head and pharynx were enclosed externally and supported internally by separate bones. Although separate bones are the result of historical constraint, their separateness provides us the opportunity to access the library of evolutionary information they hold.
The goal of this section is to provide a summary of the bones of the craniofacial skeleton that can be seen from the side. The bones are organized by alphabetical order, and their contacts, general structure, and functional significance are described.
Ectopterygoid: The ectopterygoid is a paired palatal bone that extends caudoventrally below the facial skeleton as a stout hook. Laterally it is braced against the medial surface of the jugal; ventrally, medially, and dorsally it joins the pterygoid. In terms of the skull frame, it separates the subtemporal fenestra from the suborbital fenestra. This bone connected the facial skeleton with the pterygoid close to the junction of that bone with the braincase and quadrate. The surface of the ectopterygoid that can be seen below the facial frame is coarsened by an extensive muscle attachment surface, presumably for a slip of the pterygoideus that inserted onto the mandibular ramus.
Frontal: The frontal is a paired bone of the dorsal skull roof, that caudally joins the parietal, caudoventrally the laterosphenoid, laterally the postorbital, rostrolaterally the lacrimal, rostromedially the prefrontal, ventrally the orbitosphenoid, rostroventrally the mesethmoid, and rostrally the nasal. In lateral view it is seen to roof the orbital space that contained the eyeball and its associated neurovasculature and musculature, and that it formed the rostral extent of a large basin (dorsotemporal fossa) from which originated parts of the adductor musculature. Also from the side, it is seen that the frontal anchored the temporal and antorbital regions at a common point above the orbital fenestra; the frontal forms a narrow slip of the fenestra between the postorbital and lacrimal in species where those bones are separate from each other.
Jugal: In lateral view, the jugal is a paired bone that overlaps the maxilla rostroventrally and the lacrimal rostrodorsally; it is overlapped in turn by the lacrimal ahead of the orbital fenestra; caudodorsally it is overlapped by the postorbital, which extends a short distance medial to it behind the orbital fenestra; caudally the jugal is overlapped by a long process from the quadratojugal; medially the jugal abuts the ectopterygoid below the orbital fenestra.
From the side, the jugal is seen to participate in many openings, including much of the ventral margin of the orbitotemporal region, and it extends forward into the caudoventral corner of the antorbital region. Rostrally it forms the caudoventral corner of the antorbital fenestra, dorsally it forms the ventral margin of the orbital fenestra, and caudally most of the rostral and ventral margins of the laterotemporal fenestra. Its position relative to the ectopterygoid shows that it formed the rostrolateral wall of the adductor chamber and the lateral wall of the suborbital space. The jugal extends below the ventral margin of the facial skeleton as a low ornamental bump.
Lacrimal: The lacrimal is a paired bone that separates the orbital fenestra behind from the antorbital fenestra ahead. Dorsally it inserts into the frontal caudomedially, joins the prefrontal caudoventrally, the nasal rostrodorsally, and the maxilla rostroventrally. The ventral end of the bone overlaps the lateral surface of the jugal caudally and is overlapped by it rostrally; its rostromedial surface joins the palatine. The lacrimal forms the rostrodorsal and rostral margins of the orbital fenestra, and the caudodorsal, caudal, and part of the caudoventral margin of the antorbital fenestra.
The lacrimal extends above the dorsal surface of the skull as a prominent bump; its coarse surface is continuous with that of the rugose dorsal surface of the nasal ahead, and with the rough lateral surface of the postorbital behind. Therefore, in additional to its structural necessity in the facial frame, its upper end passively transmitted information to conspecifics regarding relative maturity and species identity.
Maxilla: The maxilla is a paired bone; it is the largest bone of the facial skeleton, when it is seen from the side. It contacts the premaxilla rostrally, the nasal dorsally, the jugal caudoventrally, the lacrimal caudodorsally, the palatine caudomedially, the vomer rostroventromedially, and its complement rostromedially. The maxilla forms the deeply concave rostral margin of the antorbital fenestra, and the rostrolateral margin of the bony choana.
It contains all except four of the teeth contained in the upper jaw, and forms part of the ventral margin of the orbital region. The maxilla is an important part of the facial frame above and below the antorbital fenestra. Above, it contributes to the dorsal skull roof with the lacrimal and nasal, and ventrally the lower bar of the snout with the lacrimal, jugal, and (out of the plane of view) the palatine. The lateral surface of the bone is excavated by the extensive antorbital fossa, which is the large depression that also extends onto the lacrimal and jugal.
Nasal: In lateral view, the nasal forms over half of the antorbital region. In tyrannosaurids it is a single unit bone, but open sutures rostrally and caudally indicate that it develops from separate ossifications that later fuse along the midline. Caudolaterally the nasal contacts the lacrimal, ventrally the maxilla, and rostrally the premaxilla. The nasal contacts the premaxilla at two points, above and below the bony naris, of which it forms its caudomediodorsal and caudolateroventral margins. The complementary nasals are joined along the dorsal midline to receive the caudodorsally extending process from the premaxilla, whereas each nasal separately extends a long process along the rostrodorsal surface of the maxilla to join the corresponding process from the premaxilla to enclose the bony naris. The nasals enclose the airway from above and support a coarse surface externally that was a part of the extensive signaling structure that is largely concentrated along the top of the orbital and antorbital regions of the skull.
Orbitosphenoid: The orbitosphenoid is a small oosification located ahead of the laterosphenoid, below the frontal, and behind the mesethmoid (not pictured because it was not preserved in any of the specimens used to produce the image). Although often described as paired bones, the orbitosphenoid in tyrannosaurids is seen as a single unit structure. The dorsal surface of the orbitosphenoid contributes to the enclosure of a canal that in life extended branches of the olfactory nerve (CN I). The orbitosphenoid was also located above the opening for the branches of the optic nerve (CN II). The orbitosphenoid, like the mesethmoid, can be thought of as ossified proximal portions of the continuous interorbital and internasal septa.
Otoccipital: In lateral view, this large paired bone is largely concealed by the surrounding palate and frame of the facial skeleton. Although its rostrodorsal part can be seen through the laterotemporal fenestra, its distalmost end forms the caudalmost point of the skull, which gives an indication of its true extent and structural importance.
The otoccipital extends medial to the prootic rostrally, the basisphenoid rostroventrally, the squamosal caudodorsolaterally, the quadrate caudodorsally, and the basioccipital ventromedially. Much of the lateral surface of the otoccipital forms part of the medial wall of the middle ear and it is penetrated by the associated pneumatic recesses, receives the stapes, and its caudoventral corner serves as the origin for the mandibular depressor, the muscle that assists gravity in opening the jaw.
Palatine: In lateral view, the palatine is the dominant bone of the palate when seen through the antorbital fenestra. This paired bone articulates caudomedially with the pterygoid, medially with its complement, rostromedially with the vomer, ventrolaterally with the maxilla, and caudolaterally with the lacrimal. Between the maxilla and lacrimal, there may be a contact with the jugal.
The palatine participates in the boundaries of several openings, including the caudal, caudodorsal, and caudoventral margins of the bony choana (internal bony naris), the rostral margin of the suborbital fenestra, and the rostrodorsal margin of the palatine fenestra. The palatine fenestra can be seen in lateral view, between the pterygoid caudodorsally and the palatine rostrally.
The palatine is an important structural component of the craniofacial skeleton, where it forms part of a massive strut that braces the antorbital region transversely (from side-to-side), and also participates in a strut that extends for nearly the entire length of the skull, starting caudally with contralateral flanges that extend rostromedially from each quadrate that continue through the pterygoids before converging at the palatine, extending rostrally through the vomer and finally below the palatal shelf of the maxilla and possibly the premaxilla. The palatine connects the facial skeleton to the palate into a mediolaterally stable frame.
Parietal: The parietal is a single unit bone, and it is the caudalmost bone of the dorsal skull roof. In lateral view, the parietal is saddle shaped, but it is mostly blocked from view by the dorsal temporal bar, except for the dorsally extending rostral extreme of the sagittal crest and the tall nuchal crest that extends above the entire dorsal skull roof. The parietal contacts the frontal rostrally, the postorbital rostrolaterally, the laterosphenoid rostroventrally, the prootic ventrally, the otoccipital caudoventrally, the squamosal ventrolaterally, and the supraoccipital caudally. It also forms the caudal half of the ceiling of the endocranial cavity (the space contains the brain and its surrounding venous and dural structures).
In lateral view, two primary landmarks of the parietal can be seen, namely the sagittal crest and the nuchal crest. The sagittal crest extends from the rostral end of the parietal along the midline to the rostral surface of the nuchal crest, where it fades. The sagittal crest forms the sharp midline boundary between the paired dorotemporal fossae that served as the surface of origin for the adductor (jaw closing) muscles.
The nuchal crest is the stable bony interface that separates the adductor musculature rostrally from the cervicocranial musculature caudally. The latter muscles extend from the vertebrae caudally onto the occipital surface of the skull rostrally. In lateral view, the surface of the nuchal crest that can be seen is a part of the dorsotemporal fossa; neck muscles would have inserted on the concave caudal surface that cannot be seen from the side, except for a marginal insertion surface along its caudolateral edge.
Parabasisphenoid: The parabasisphenoid is a single unit bone that is one of the largest bones of the braincase; however, very little of it can be seen in lateral view aside from a slip of its caudoventral corner between the quadrate and jugal within the laterotemporal fenestra, and nearly the entirety of the sphenoid rostrum within the orbital fenestra. The parabasisphenoid contacts the laterosphenoid rostrodorsally, the prootic dorsally, and the otoccipital caudally. This bone forms the ventral and rostral margins of the basicranium (the region below the endocranial cavity).
The slip that can be seen through the laterotemporal fenestra pertains to the caudoventral margin of the basicranium between the basal tuber caudodorsally and the basipterygoid process rostroventrally. The structures are blocked from view by the quadrate and jugal, respectively.
The sphenoid rostrum is a midline structure that is excavated by a concavity caudodorsally, which is called the Turkish saddle (sella turcica) that lodged part of the pituitary gland from behind and below. The rostrum extends rostroventrally into the orbital space. Ahead of the saddle, a long furrow excavates the entire dorsal surface of the rostrum to its tip, whereas its ventral surface is convex and smooth. The cartilaginous interorbital septum almost certainly fit ventrally into the rostrum and was stabilized from above by the orbitosphenoid, mesethmoid, frontal, and nasal. If so, then the septum was much wider dorsally than it was ventrally.
Postorbital: The postorbital is a paired bone that is triradiate in shape, which sends a process rostrally that articulates with the frontal rostromedially, the parietal caudally, and the laterophenoid caudoventrally; it sends a process caudally that fits into a deep and narrow slot in the lateral surface of the squamosal to complete the dorsal temporal bar; and it sends a process ventrally that fits onto the rostrolateral surface of the jugal to complete the postorbital bar. In some tyrannosaurines the rostral process of postorbital reaches far enough forward to contact the lacrimal.
In lateral view, the postorbital separates the laterotemporal and orbital fenestrae, where it forms the rostrodorsal corner, and the caudodorsal and caudal margins of the openings, respectively. The lateral surface that extends between the ventral and rostral processes carries an ornamental structure, the cornual process of the postorbital. This is seen in all derived tyrannosauroids, except in the least mature specimens. In Albertosaurus libratus, as pictured here, the process it situated along the orbital fenestra and along the ventrolateral edge of the rostral process. In other tyrannosauroids, this ornament is much larger and it is situated caudal to the margin of the fenestra. The lateral surface of the ventral process has a coarse texture, which is continuous with texture that also covers the entire antorbital region.
The postorbital forms the dorsal margin of the skull between the squamosal caudally and the frontal rostrally (lacrimal in some tyrannosaurines). The rostral half of the bone flanks the orbital space, whereas it caudal half flanks the adductor chamber. Also, the postorbital is limited ventrally by the jugal to approximately the midheight of the skull frame, where it stops above the level of the tooth-bearing regions of the maxilla and premaxilla.
The ventral edge of the rostral process and the concave margin between the rostral and ventral processes form the caudodorsal margin of the orbit proper (the part that actually surrounded the eyeball in life). The irregular and vertically oriented rostroventral margin of the ventral process flanked the space situated below the eyeball. Pterygoideus musculature that originated from the palatine and extended caudally and ventrally over the ectopterygoid to insert onto the lower jaw, occupied this space; this group of muscles would have almost certainly provided speed to the bite, as is seen in modern crocodilians.
Prefrontal: In lateral view, the prefrontal is a small triangular wedge of bone situated at the rostrodorsal corner of the orbital fenestra; it is a paried bone. Caudally it inserts into the frontal and rostroventrally it fits into a groove in the caudomediodorsal surface of the lacrimal. The caudoventral edge of the bone forms part of a strut that extends rostrolaterally from the cranial crest (crista cranii) of the frontal to the orbitonasal ridge of the lacrimal. Regionally, this part of the prefrontal separates the orbital space caudolaterally from the nasal airway rostromedially. The prefrontal is situated below the rostrolateral surface of the frontal, and so it does not reach the midline of the skull. The prefrontal is a deep structure that does not contribute to the lateral surface of the face or to the margin of the orbital fenestra.
Premaxilla: The premaxilla is a paired midline bone that surrounds the bony naris rostrodorsally, rostrally, and rostroventrally; it contacts the nasal caudodorsally, caudally, and the maxilla caudoventrally. The premaxilla also contains the mesialmost part of the tooth row. The contact between the premaxilla and the maxilla is separated into dorsal and ventral parts by the subnarial foramen.
In lateral view, the caudodorsal surface of the premaxilla is flattened, smoothed, and excavated by the narial fossa. The coarse subcutaneous surface covers the rostral surface of the bone, which extends caudoventrally below the subnarial foramen onto the maxilla, and dorsally along the bony naris onto the nasal. Large foramina penetrate the subcutaneous surface and narial fossa, a rostral continuation of the row of foramina that extends along the ventral margin of the maxilla.
Prootic: The prootic is an extensive paired bone that is situated on the lateral surface of the braincase, and much of it can be seen through the laterotemporal fenestra. The prootic contacts the laterosphenoid rostrally, the parabasisphenoid rostroventrally, the otoccipital caudoventrally and caudally, the squamosal caudodorsally, and the parietal rostrodorsally. The prootic overlaps the lateral surface of the otoccipital and a gap separates them at the rostral end of their contact.
The prootic is a structurally important bone, where in lateral view it surrounds the rostrodorsal part of the middle ear, forms part of the bony boundary between the dorsotemporal fossa and the middle ear, and between the middle ear and the orbital space. The rostral end of the auditory (hearing) ossicle, the stapes, is lodged between the otoccipital medially and the prootic laterally.
Pterygoid: In lateral view it can be seen that the paired pterygoid is one of the longest bones of the skull, and the longest of the palate. The pterygoid is represented by dashed lines in the carbon dust plate of A. libratus, because I did not have a specimen of a pterygoid available to me at the time I drafted the line drawing, some time between 1996 and 1998.
The pterygoid contacts the quadrate caudodorsally, the parabasisphenoid caudomedially, the ectoptergyoid laterally, its complement rostromedially, the palatine rostrolaterally, and the palatine and vomer rostrodorsally. In lateral view, the pterygoid is seen to form the caudodorsal boundary of the palatine fenestra.
Quadrate: In lateral view, the ventral and rostral parts of the paired quadrate can be seen. The quadrate contacts the squamosal dorsolaterally, the otoccipital dorsomedially, the quadratojugal caudodorsally and caudoventrally, the pterygoid rostrolaterally, and ventrally the mandibular ramus (surangular laterally and articular medially).
When seen from the side, the long and deep orbital process can be seen extending through the adductor chamber. The orbital process would have been positioned deep (medial) to the adductor musculature. At the same time, the process bounded the middle ear space laterally, in life partly shielding its contents from the bulging contractions of the jaw closing muscles. Also, the orbital ramus has an extensive articulation with pterygoid (not shown here, for the reason given above). As such, this part of the quadrate is a functional and structural part of the palate.
The ventrally extending part of the quadrate is the mandibular process; this structure is separated by a groove into a lateral and medial condyle. A ridge formed by the articular fits into the groove between the condyles. The groove extends rostromedially, which constrained the excursion of the mandibular ramus upon opening and closure. As a result, the high mediolateral forces that occurred at this joint during the capture and rendering of prey were prevented from dislocating the jaw.
Quadratojugal: The quadratojugal is the paired L-shaped bone that forms the caudoventral corner of the facial skeleton. This bone forms the caudoventral margin of the skull, a shallow concavity that in life was situated ahead of the ear canal (external auditory meatus). The quadratojugal also forms the ventral margin of the skull for a short distance ahead of the jaw joint; the bone also completes the caudoventral corner of the laterotemporal fenestra. The quadratojugal overlaps all of its surrounding bones: the squamosal dorsally, the quadrate caudodorsally and caudoventrally, and the jugal rostroventrally. The squamosal and jugal are overlapped in the vertical and rostrocaudal planes, whereas the quadrate is overlapped in the mediolateral and vertical planes. Also, a blade-like process from the quadrate extends laterally through a slot at the caudodorsal corner of the bone.
The quadratojugal is composed of two primary processes: a dorsal process that expands rostrally toward the squamosal and a long rostral process that covers the lateral surface of the jugal. The dorsal process and squamosal together form a wide flange that nearly cuts the laterotemporal fenestra in two. The quadratojugal and jugal form the lower temporal bar. Based on its position and contacts, the quadratojugal helps to stabilize the jaw joint between facial skeleton laterally and dorsally, and the braincase dorsomedially.
Squamosal: In lateral view, the squamosal is a complex paired bone that can be divided into three processes, including a rostral process that forms most of the upper temporal bar and receives the postorbital on its lateral surface, a caudal process that overlaps the otoccipital, and a rostroventral process that extends along the top of the quadratojugal into the laterotemporal fenestra.  The squamosal contacts the postorbital rostrally, the parietal, otoccipital, and prootic medially, the otocciptial caudally, the quadrate ventrally, and the quadratojugal rostroventrally.
The squamosal is extensive, where it forms the entire caudodorsal corner of the craniofacial skeleton, including its lateral, caudal, and dorsal surfaces. Although it is adjacent to the adductor chamber, the internal surface of the bone almost certainly was not occupied by musculature; instead, it appears that it was a part of the middle ear cavity. The squamosal forms the tallest part of the temporal region and extends caudoventrally along the upper temporal bar to the otoccipital. Together these bones form a process that extends caudally above the region of the external ear opening.
The squamosal is a structurally important bone in that it stabilized the top of the quadrate against itself and the lateral surface of the braincase. It also completed the connection between the lower and upper temporal bars, and enclosed the adductor and middle ear spaces caudodorsally.
Vomer: In lateral view, the single unit and midline vomer is seen through the antorbital fenestra, where it extends rostrally from the medial surface of the palatine to the ventral surface of the palatal shelf of the maxilla; caudally the voner contacts the pteryoid, a contact that is concealed by the palatine. The vomer forms most of the dorsomedial margin of the bony choana and it forms the ventral bony part of the internasal septum with the palatine.
The vomer completes the axial strut that begins caudally with the widely separate quadrates and rostrally-converging pterygoids. The vomer would have helped to resist dorsoflexion (extension) of the antorbital region.

Tuesday, July 2, 2013

Literature review I: Larson, 2013

The single most contentious dinosaur specimen in recent history: the Cleveland Museum of Natural History's juvenile Tyrannosaurus rex skull (CMNH 7541). Skull length is 752 millimeters. This is a digitally colorized version of the original carbon dust plate that was published in Carr (1999).

This entry is a challenge for me to pen, because I have two manuscripts in progress that have direct bearing on the ontogeny of Tyrannosaurus rex. I really do not want to deal with Larson's article in brevity, but my hands are tied because I will not put unpublished data here.

The bottom line is that Larson’s article “The case for Nanotyrannus” is a litany of straw man arguments and false analogies, misinterpretations of osteology, misunderstandings of variation and ontogeny, all rooted in a limited data set and a verificationist approach. None of it convinces me that the hypothesis I published in 1999 is incorrect: the weight of evidence shows that the Cleveland skull is referable to Tyrannosaurus rex, based on a comparison with the growth series of a well-known tyrannosaurid (namely, Albertosaurus libratus).

The Cleveland skull shows the juvenile features that are seen in all tyrannosaurid taxa, not just A. libratus, but also A. sarcophagus, Daspletosaurus torosus, and T. bataar (Carr, 1999; Currie, 2003). It also shows features that are shared with adult T. rex (yes, and adults lower in the tyrannosaurine phylogenetic hierarchy); ergo, the simplest hypothesis is that the skull is (1) a juvenile tyrannosaurid that is (2) referable to T. rex. From what I’ve read, Larson (2013) has set this line of reasoning aside in favor of something complicated.

On a lighter note, if that Late Cretaceous oviraptorid embryo (Larson, 2013: 44) – evidently in an egg – had the loving attention of so much ink, then we would have learned a lot more from that specimen about the rules of theropod ontogeny and evolution, instead of this taxonomic wallow!

References cited
Carr, T. D.  1999.  Craniofacial ontogeny in Tyrannosauridae (Dinosauria, Theropoda). Journal of Vertebrate Paleontology 19:497-520.
Currie, P.J. 2003. Cranial anatomy of tyrannosaurid dinosaurs from the late Cretaceous
Alberta, Canada. Acta Palaeontologica Polonica 48: 191226.
Larson, P. L. 2013.  The case for Nanotyrannus in J. M. Parrish, R. A. Molnar, P. J. Currie., and E. B. Koppelhus (eds.) Tyrannosaurid Paleobiology, University of Indiana Press, Bloomington and Indianapolis, pp. 15-53.