Neural spine bifurcation in sauropods, Part 1: what we knew a month ago

April 5, 2012

The discussion over the new paper by Woodruff and Fowler (2012)–see this post and the unusually energetic comment thread that follows–made me want to go back to the literature and see what was known or could be inferred about neural spine bifurcation in the Morrison sauropods before the recent paper was published.

In this post I’m mostly going to stick with the “classic” specimens–those that are reasonably complete and have been monographed.  So this is the stuff that anyone could easily have figured out for themselves–I am just gathering it here so that it’s conveniently in one place.

Cervical vertebrae

Apatosaurus louisae, CM 3018, and A. parvus, CM 563/UWGM 15556

Apatosaurus parvus CM 563/UWGM 15556 cervicals 7, 5, 4 and 3 in anterior and right lateral views, from Gilmore (1936:pl. 31)

CM 3018 is the mounted Apatosaurus at the Carnegie Museum in Pittsburgh, which was exhaustively described by Gilmore in his 1936 monograph. In that paper Gilmore also described CM 563, which is now mounted at the University of Wyoming Geological Museum and cataloged as UWGM 15556. Gilmore referred CM 563 to A. excelsus, but in a specimen-level phylogenetic analysis Upchurch et al. (2005) found CM 563/UWGM 15556 to belong to A. parvus–hat tip to alert reader LeeB. for catching the revised specific referral. On the subject of neural spine bifurcation, Gilmore wrote (1936:195):

Unfortunately the type of A. louisae lacks most of the spine tops, only  those of cervicals eight, ten and twelve being complete; thus the point of change from single to bifid spines cannot be determined in this specimen. In specimen No. 563, C.M., identified by Hatcher as pertaining to Brontosaurus, an identification with which I concur [Apatosaurus louisae and Brontosaurus excelsus were then considered by some to be separate genera], there are nine cervical vertebrae preserved, three of which I regard as cervicals, three, four, and five. These show the spines to be single as far posteriorly as the fifth vertebra. Since C. 7 shows a well defined notch between the metapophyses, it seems to be a fair conclusion that C. 6 in Apatosaurus is the first vertebra to show a notch on the summit of the spine.

In Camarasaurus supremus Osborn and Mook make the observation that, “In C. 5, the spine has a very slight median notch.” In C. lentus [Gilmore 1925:369], however, the first notched spine is that of C. 7. From C. 6 to C. 9 inclusive, the spinal notch increases steadily in depth. From C. 9 to C. 15 inclusive the spine is completely divided into two metapophyses.  [An odd statement, since Gilmore (1925:351) says that there are twelve cervicals in C. lentus. By “C. 15” he must mean the 15th vertebra, i.e. D3.]

In CM 3018 the spine of C7 appears to be split, based on Gilmore’s Plate 24 (note that for a deeply notched spine, it is not necessary to have the spine tips preserved; the base of the trough will do), and the spine of C6 might be. Gilmore doesn’t mention it in the text, but the plate seems to show C6 with broken metapophyses (at least in posterior view, there are break lines across the metapophyses at mid-height) but with a preserved trough in the middle, which if accurate is enough to confirm that the spine of C6 was also at least partially bifid.

In CM 563/UWGM 15556, there is little doubt: C3, C4, and C5 all have single spines, and the spine of C7 is deeply split. I am happy with Gilmore’s estimates of serial position, by the way. I don’t think the putative C3 could be further back, given how small and short it is, and it’s the wrong shape for an axis. C4 and C5 fall into line in both size and morphology. There is clearly at least one missing vertebra between C5 and C7. The C7 might just possibly be a C8 but I really don’t think it could be any farther back than that, and I agree with Gilmore that it makes more sense as a C7.

The mounted apatosaurs at the Yale Peabody Museum, the AMNH, and the Field Museum are not much help.* The cervical series of all three are heavily reconstructed, and at least for the YPM and AMNH specimens it is quite difficult to tell where the bone ends and the plaster begins.

* Before some outraged curator, collections manager, docent, or museum-goer gets up in my grill about this, I’m not saying they’re bad mounts, or that the necks are entirely fictional! It’s just a sad fact that the fragile neural spine tips are easily broken off and therefore rarely preserved intact. The museums are not defrauding anyone by sculpting in replacements–but that reconstructive work does make it hard to use those specimens as data if neural spine bifurcation is the character of interest. That’s all I’m saying.

Apatosaurus ajax, NSMT-PV 20375

Apatosaurus ajax NSMT-PV 20375, cervical vertebrae 3, 6 and 7 in anterior and posterior views. Modified from Upchurch et al. (2005: plate 2)

This is the new(ish) A. ajax described by Upchurch et al. (2005). They wrote (pp. 27-28):

The neural spine is unbifurcated in C3, but strongly bifurcated from C6 onwards; this seems to be the typical location for the onset of spine bifurcation in Apatosaurus, since it also occurs in this manner in CM 3018 and UWGM 15556.

Note that the intervening vertebrae are missing from that specimen, so we can’t tell exactly where the bifurcation began, but it is perfectly consistent with CM 563/UWGM 15556.

Diplodocus carnegii, CM 84/94

Diplodocus carnegii cervicals 2-15 in posterior view, from Hatcher (1901:pl. 6). Note that the bifid spines in C3-C5 are sculptures; there is no evidence that these spines were bifurcated when the vertebrae were intact.

Hatcher (1901:20-21; emphasis added):

Cervicals Three, Four, and Five.–All of these vertebrae are more or less injured. The neural spines and transverse processes especially are not well preserved. […] Commencing with C. 3 the neural spines of these vertebrae have been restored as bifid both anteriorly and posteriorly, each spine consisting of a broad thin plate of bone formed by the union of the pre- and postzygapophyseal laminae of their respective sides. These are made to appear free anteriorly and posteriorly, but united, except at their apices, throughout the inner sides; conditions which prevail in the succeeding cervicals.

Cervicals Six, Seven, Eight, Nine and Ten.–These vertebrae differ so little in their more important characters that they may be very conveniently described together. They are all fairly well preserved and show certain characters which are gradually more emphasized in the succeeding vertebrae of the series. Commencing with C. 6 they  regularly increase in length posteriorly. The neural spines become more completely bifid, resulting in a pair of transversely placed perfectly free spines on the tenth cervical consisting of  triangular plates of bone diverging superiorly and terminating at the summit in a rather blunt, rounded process.

Unfortunately Hatcher’s plates do not show which areas have been reconstructed and which have not–a very common failing in these classic monographs, and one which Upchurch et al. (2005) happily did not duplicate. Hatcher says that certain characters are more emphasized in more posterior vertebrae, and that by C10 the metapophyses are “perfectly free”, which suggests that they might have been less than perfectly free before. The spines of C3-C5 have been reconstructed as bifid, but that was to make them consistent with the succeeding vertebrae. So CM 84/94 is like its neighbor in the Carnegie dinosaur hall, CM 3018, in that in both specimens we know that the spines are bifid by some point but we don’t know what was going on in C3-C5.

Barosaurus lentus, AMNH 6341

A. Barosaurus lentus AMNH 6431 cervical vertebra 8 in anterior, left side, and posterior views. B. Diplodocus carnegii CM 84 cervical 8 in anterior, left side and posterior views (from Hatcher 1901). (McIntosh 2005:fig. 2.2)

This is the mostly-complete skeleton famously mounted in a rearing pose in the AMNH rotunda. McIntosh (2005:47-48):

The neural spine of cervical 8 is flat across the top, and that of 9 shows the first trace of a divided spine (Fig. 2.2A). This division increases gradually in sequential vertebrae, being moderately developed in cervicals 12 and 13, and as a deep V-shape in cervicals 15 and 16. This development is in sharp contrast to Diplodocus, where cervical 3 already shows the first trace of division, and where the division is already quite deep in cervical 7 (Fig. 2.2B). By cervical 11 it is as well developed as cervical 16 of Barosaurus (AMNH 6341; Fig. 2.3A). In the spines, a further difference is that those of the last two cervicals (14 and 15) of Diplodocus project anterodorsally, whereas those of Barosaurus are all directed dorsally (Fig. 2.3B).

(McIntosh’s reference in the text to cervical 7 of Diplodocus carnegii is partly in error; the neural spine cleft is fairly deep in C7 but the vertebra shown in the figure is C8, as correctly noted in the figure caption.)

Suuwassea emilieae, ANS 21122

Suuwassea emilieae C6 in anterior, left lateral, and posterior views, from Harris (2006:Text-fig. 7A-C)

Harris (2006) on C2 (p. 1096, Text-fig. 4):

The spine gradually widens mediolaterally toward the distal end, which is rendered heart-shaped by a 12-mm-deep, sagittal, parabolic notch.

C3 (Text-fig. 5) is missing the end of its spine, enough remains to show that if it was bifid the cleft could not have been very deep.

C4 is missing.

C5 (p. 1099, Text-fig. 6):

The spinous process expands mediolaterally toward its apex, attaining maximal width just proximal to its terminus. A long, narrow crack at the distal end gives the appearance of bifurcation, but the collinear dorsal margin indicates that no true split was present.

C6 (p. 1101, Text-fig. 7, above):

The distal end of the spine is cleft by a parabolic, 11.8-mm-deep intraspinous sulcus, marking the initial stage of bifurcation.

Camarasaurus lewisi, BYU 9047, C. supremus AMNH 5761, C. lentus CM 11338 and YPM 1910

Camarasaurus supremus AMNH 5761, C2-C7 in anterior, left lateral, and poster views, from Osborn and Mook (1921:pl. 67)

McIntosh, Miller, et al. (1996:76) on Camarasaurus (= “Cathetosaurus“) lewisi:

The spines are placed well back on the arches and rise higher than in C. grandis, C. lentus, or C. supremus. The cleft in the bifid spine is much deeper and narrower than in any of those species. The cleft in cervical 3 of C. grandis (YPM 1905) is barely perceptible, very modest in numbers 4 and 5, and distinct in 6.

The metapophyses are transversely broad. If the arrangement portrayed by Osborn and Mook (1921) for the large adult C. supremus (AMNH 5761) is correct, the notch between them first appears in cervical 7, the same position reported by Gilmore (1925) for the juvenile C. lentus (CM 11338). However, a small depression is present in cervical 5 of the holotype (YPM 1910) of the latter species. In C. lewisi (BYU 9047) a narrow, deep, sharp cleft is already present in cervical 3 (Jensen, 1988). The depth continues to increase greatly to cervical 8, the last cervical in which reliable measurements of this feature can be taken, where it appears as a very steep, V-shaped notch. This feature appears to be unique to C. lewisi.

McIntosh’s comments here perfectly match the descriptions provided by Osborn and Mook (1921) and Gilmore (1925) so I have not bothered pasting in the relevant sections of those papers (though I did reread them to make sure everything matched).

Camarasaurus grandis, GMNH-PV 101 (WPL 1995) and YPM 1905

McIntosh, Miles, et al. (1996:11-12):

As in C. grandis YPM 1905, the bifurcation of the spine begins as an incipient notch on cervical 3, but it is not until cervical 5 that one observes the fully developed U-shaped trough, and even then the cleft is not nearly as prominent as in the cervicals of the diplodocids. In this respect GMNH-PV 101 Camarasaurus, and most of the other specimens of the genus, differ from the deep, narrow bifurcation seen in C. lewisi BYU 9047 (see Table C).

However, this description does not match the illustrations. While fig. 25 does show a very subtle notch in C3, fig. 26 shows no bifurcation whatsoever in C4 which has a distinctly convex neurapophysis. Subtle bifurcation returns in C5, and deepens slightly in C6, C7 and C8.

Inferences on bifurcation in cervicals

Remember that here I am not trying to either support or challenge the work of Woodruff and Fowler (2012), I’m just looking at what had been published previously and seeing what inferences could be drawn from that evidence only.

1. There is no evidence in any of the North American diplodocoids of a bifid spine farther forward than C6. The bifid spines in the mounted skeleton of D. carnegii are sculptures; Hatcher was doing his best with imperfect fossils and limited information. The appearance of a split spine in C5 of Suuwassea is caused by a vertical crack and a small amount of missing bone. In the very large AMNH 6341 Barosaurus, the first partially split spine is on C9.

2. Adult sauropods can show unbifurcated spines, partially bifurcated spines, and fully bifurcated spines serially in the same individual. This is true even in very large individuals (e.g., Apatosaurus parvus CM 563/UWGM 15556, Barosaurus lentus AMNH 6341, Camarasaurus supremus AMNH 5761), so it is unlikely to be an artifact of ontogeny. Therefore single spines do not always indicate juveniles, bifid spines do not always indicate adults, and incompletely bifid spines did not always become fully bifid–in all of the specimens listed above, the most anterior bifid spines are only shallowly divided. We should probably describe vertebrae with shallow splits as ‘incompletely’ bifid rather than ‘incipiently’ bifid; the latter term implies that the bifurcation was going to deepen with time, which did not always happen depending on serial position.

3. The evidence from Camarasaurus is consistent with an ontogenetic increase in bifurcation. The juvenile C. lentus described by Gilmore (1925) has the first incipiently bifurcated spine at C7, whereas the larger, presumably adult individual of the same species represented by YPM 1910 has the first split at C5, as do the individuals that make up C. supremus AMNH 5761. In C. lewisi BYU 9047 and C. grandis YPM 1905, and arguably in C. grandis GMNH-PV 101 the first spine to be partially split is C3. It is tempting to interpret the difference between adult C. lentus and C. supremus on one hand (first split at C5) and C. lewisi and C. grandis on the other (first split at C3) as interspecific variation, but it might be individual variation considering that we are dealing with usually just one individual from each species (for neural spine bifurcation in adults; I am aware that there are other individuals not mentioned here).

Dorsal vertebrae

I’m going to report the results here in a more compact form than I did for the cervicals. My convention of convenience will be: spines that are split over more than half the distance from the tips to either the postzygapophyses or transverse processes (whichever are higher) are described as deeply bifid, and those split over less than half that distance, including very shallow dorsal indentations, are described as shallowly bifid. The expected dorsal counts are 10 in Apatosaurus and Diplodocus, 9 in Barosaurus, and 12 in Camarasaurus.

Apatosaurus louisae CM 3018 (Gilmore 1936): Deeply bifid  in D1-D3, shallowly bifid in D4-D6, unsplit in D7-D9, D10 spine missing.

Apatosaurus parvus CM 563/UWGM 15556 (Gilmore 1936): Deeply bifid in D1-D3, shallowly bifid in D4, D5-D10 spines missing.

Apatosaurus ajax NMST-PV 20375 (Upchurch et al. 2005): Deeply bifid in D1-D4, shallowly bifid in D5-D6, unsplit in D7-D10.

Diplodocus carnegii CM 84/94 (Hatcher 1901): Deeply bifid in D1-D5, shallowly bifid in D6-D9, unsplit in D10.

Diplodocus longus USNM 10865 (Gilmore 1932): Deeply bifid in D1-D5, shallowly bifid in D6-D8, unsplit in D9-D10.

Barosaurus lentus YPM 429 (Lull 1919): Deeply bifid in D1, D4, and D5, unsplit in D6-D9 (NB: Lull interpreted the latter as D7-D10 on the expectation of 10 dorsals, based on Diplodocus).

Barosaurus lentus AMNH 6341 (McIntosh 2005): Deeply bifid in D1-D3, shallowly bifid in D4-D8, unsplit in D9.

Camarasaurus lentus CM 11338 (Gilmore 1925): Deeply bifid in D1-D3, shallowly bifid in D4-D6, unsplit in D7-D12.

Camarasaurus supremus AMNH 5761 (Osborn and Mook 1921): Deeply bifid in D1-D6, shallowly bifid in D7-D8, unsplit in D9-D12 (NB: a bit of guesswork here, since Osborn and Mook were working with disarticulated material and interpreted Camarasaurus has having 10 or 11 dorsals).

Camarasaurus lewisi BYU 9047 (McIntosh, Miller, et al. 1996): Deeply bifid in D1-D8, shallowly bifid in D9-D12.

Inferences on bifurcations in dorsals

1. As with the cervicals, most adult sauropods have deeply bifid, shallowly bifid, and unsplit spines in serially adjacent vertebrae. In the diplodocids, the spines of D6-D10 (or D9 in Barosaurus) are always either unsplit or very shallowly indented at the tips.

2. The diplodocid genera show some interesting differences. In Apatosaurus the last four dorsals are always unsplit. In Diplodocus the spines are at least shallowly indented as far back as D8 or D9. Barosaurus goes both ways, with YPM 429 having unsplit spines in the four most posterior dorsals, and AMNH 6341 having an entirely unsplit spine only in the last dorsal.

3. In the diplodocids, deep splits are always confined to the first half of the dorsal series (D1-D5), and these are usually followed by a long run of vertebrae with very shallowly notched spine tips. The exception is Barosaurus YPM 429, which–if the vertebrae are truly consecutive (the series is missing at least two)–has a deep split in D5 and unsplit spines in D6-D9.

4. As with the cervicals, the evidence from Camarasaurus does not rule out an ontogenetic increase in bifurcation. In the juvenile C. lentus CM 11338, the spines are  only bifid as far back as D6; in the adult C. supremus AMNH 5761 to D7; and in the old C. lewisi BYU 9047 to D12. If these differences represent ontogenetic changes rather than interspecific differences (which also cannot be ruled out at this point), it is interesting that there is a bigger difference between the adult C. supremus and the old C. lewisi than between the juvenile C. lentus and the adult C. supremus: in other words, the greatest changes took place after adulthood was attained.

The rest of the series

Links to all of the posts in this series:

and the post that started it all:


  • Gilmore, C.W. 1925. A nearly complete articulated skeleton of Camarasaurus, a saurischian dinosaur from the Dinosaur National Monument. Memoirs of the Carnegie Museum 10:347-384.
  • Gilmore, C. W. 1932. On a newly mounted skeleton of Diplodocus in the United States National Museum. Proceedings of the United States National Museum 81:1-21.
  • Gilmore, C.W. 1936. Osteology of Apatosaurus with special reference to specimens in the Carnegie Museum. Memoirs of the Carnegie Museum 11:175-300.
  • Harris, J.D. 2006. The axial skeleton of the dinosaur Suuwassea emilieae (Sauropoda: Flagellicaudata) from the Upper Jurassic Morrison Formation of Montana, USA. Palaeontology 49:1091-1121.
  • Hatcher, J.B. 1901. Diplodocus (Marsh): its osteology, taxonomy, and probable habits, with a restoration of the skeleton. Memoirs of the Carnegie Museum 1:1-63.
  • Lull, R.S. 1919. The sauropod dinosaur Barosaurus Marsh. Memoirs of the Connecticut Academy of Arts and Sciences 6:1-42.
  • McIntosh, J.S. 2005. The genus Barosaurus Marsh (Sauropoda, Diplodocidae); pp. 38-77 in Virginia Tidwell and Ken Carpenter (eds.), Thunder Lizards: the Sauropodomorph Dinosaurs. Indiana University Press, Bloomington, Indiana, 495 pp.
  • McIntosh, J.S., Miles, C.A., Cloward, K.C., and Parker, J.R. 1996. A new nearly complete skeleton of Camarasaurus. Bulletin of the Gunma Museum of Natural History 1:1-87.
  • McIntosh, J.S., Miller, W.E., Stadtman, K.L., and Gillette, D.D. 1996. The osteology of Camarasaurus lewisi (Jensen, 1988). BYU Geology Studies 41:73-115.
  • Osborn, H.F. and Mook, C.C. 1921. Camarasaurus, Amphicoelias, and other sauropods of Cope. Memoirs of the American Museum of Natural History 3:247-287.
  • Upchurch, P., Tomida, Y., and Barrett, P.M. 2005. A new specimen of Apatosaurus ajax (Sauropoda: Diplodocidae) from the Morrison Formation (Upper Jurassic) of Wyoming, USA. National Science Museum Monographs No. 26. Tokyo. ISSN 1342-9574.
  • Woodruff, D.C, and Fowler, D.W. 2012. Ontogenetic influence on neural spine bifurcation in Diplodocoidea (Dinosauria: Sauropoda): a critical phylogenetic character. Journal of Morphology, online ahead of print.

16 Responses to “Neural spine bifurcation in sauropods, Part 1: what we knew a month ago”

  1. Jay Says:

    Damn fine post, Matt. Good effort compiling all this together in one convenient place

  2. Dino Hunter Says:

    Tell me it ain’t so, a post on sauropods! Really Sauropods, thats like dinosaurs, what happened? WOW, SVPow, talking about dinosaurs, I didn’t think it was possible… :) I keep waiting for someone to talk about the neck in Apatosaurs, Diplodocus so I can finnally correct my incorrect article I wrote in PT a decade or so ago…I thought Greg Paul or Scott Hartman was working on that…

  3. Mike Taylor Says:

    Tracy, I’m not sure where you’ve been, but this is the fourth SV-POW! article about sauropods in the last month (DIY project: remodel your basement…, Changes through growth in sauropods and ornithopods, Did sauropods support their weight by living in vast shallow lakes? and now this one), and in the same period there have been three more articles about other saurischian vertebrae (It came from my desk, Things to Make and Do, part 10: ostrich cervical diverticula, Things to Make and Do, part 10b: more fun with ostrich necks)

  4. LeeB Says:

    Following Upchurch et. al. 2005 isn’t UWGM 15556 now A. parvus rather than A. excelsus?


  5. Matt Wedel Says:

    Ack! You’re right, of course. I’ll revise the post accordingly.

  6. […] In the previous post in this series I looked at the some of the easily available raw data on neural spine bifurcation in Morrison sauropods. In this post I’ll explain how serial variation–that is, variation along the vertebral column in one individual–is relevant to the inferences made in the new paper by Woodruff and Fowler (2012). But first, a digression, the relevance of which will quickly become clear. […]

  7. […] in a series on neural spine bifurcation in sauropods, inspired by Woodruff and Fowler (2012). In the first post, I looked at neural spine bifurcation in Morrison sauropod genera based on the classic monographic […]

  8. […] there’s Matt Wedel’s long and detailed posts (parts 1 parts 2 and now part 3) about the birfurcation of sauropod vertebral neural spines. Part 2 is […]

  9. […] a known diplodocid, building on the information presented in the first three posts in this series (Part 1, Part 2, Part […]

  10. […] as we prepare the submission of the paper arising from our response to Woodruff and Fowler (2012) [part 1, part 2, part 3, part 4, part 5, part 6].  Here is an oddity. Sacra of Haplocanthosaurus. Top, H. […]

  11. […] (For a bit o’ fair-and-balanced, remember that this neck may not be complete, and that some of the neural spines are sculptures.) […]

  12. […] was cruising the monographs the other night, looking for new ideas, when the humerus of Opisthocoelicaudia stopped me dead in […]

  13. […] were giant storks“, which became Witton and Naish (2008). Then last year our string of posts (starting here) on neural spine bifurcation in Morrison sauropods became the guts–and most of the muscles […]

  14. […] way to generate a big paper quickly, even with the rewriting. It was just over two months from the first post in the destined-to-become-a-paper series on April 5 to submission on June 14 (not 24 as it says on […]

  15. […] other side. I believe they were the first to advance that idea in print (although I should probably take my own advice and scour the historical literature for any earlier instances), and needless to say, I think […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: