When dealing with research from a particular few scientists – namely, the BANDits – none of them intrigue me more than the work of Theagarten Lingham-Soliar (hereafter, TLS). It isn’t just that the subject matter is intriguing (structure of skin, body shape and contour, reconstruction of body in mosasaurs and ichthyosaurs from sharks and cetaceans) but that amongst BANDits, TLS is the most practical, the more methods-driven. TLS’s primary work has been to uncover the structure and arrangement of fibres that comprise skin, and associated structures, tissues like collagen, elastin, and the surface of this skin. When I first began reading his work on dinosaur skin, focusing on Sinosauropterys prima, I was extremely interested: I wanted not merely to know how he got his conclusions, but how the work he sought to refute got theirs. I was fully willing to entertain that TLS, a seeming maverick, was a nuts and bolts scientist, applying a methodology towards discriminating epidermal, dermal, and extra-dermal structures. TLS had cut his teeth on shark and whale skin, applied it to other animals, and was confirmed in his conclusions (or at least not doubted). It seemed TLS had caught a fish, and it was a big one.
In 1996, Chinese farmers uncovered a fossil trove near the town of Sihetun in China’s northwest Liaoning Province. In a mere year, the site would effectively revolutionize paleontological discourse. One of the first specimens shown to outsiders made its way to the IVPP’s Ji Qiang; the specimen, GMV 2123 (NGMC 2123), preserved on two halves of a sandstone slab and split down the middle, is of a small animal, but Ji noticed a peculiar halo of fine filaments around the body and, despite its anatomy, described it as a bird. “Sinosauropteryx prima” means, literally, first Chinese feathered bird. Soon, another specimen followed (GMV 2124), notable in which the tail wraps underneath the skull, but its inclusion in Sinosauropteryx prima has lately been questioned, and it may belong to a new taxon – either a new species of Sinosauropteryx, or a new equivalent taxon altogether. Scientists soon came to China, either the Canadians joining the Chinese following the Silk Road, or the Americans, passing through China on their way to revisit the Late cretaceous Gobi Desert in Mongolia and surroundings, and together they noticed that the specimens weren’t quite birds, but nonavian theropod dinosaurs: they were feathered – so it seemed – but the pelvis was propubic and “simple,” unfused, while the arms were ridiculously short with a giant thumb but three clawed, functional digits. Might the filaments be primitive feathers?
A century ago, artists and scientists paid little heed to the origin of birds: if there was a kink to evolution, it was the question of the origin of birds – or of whales, which intrigued Darwin. We know this story is not so much a kink as a matter of incomplete knowledge, as indeed all “missing links” are; but early in the last century, people like Gerhard Heilmann and William Beebe argued that the ancestor of the bird was a small, lizard-like animal, and that feathers looked like large, segmented scales. This view had advanced far into the last century, even gaining traction into the 1980s and 1990s, where the “quintessential” bird-ancestor was pegged as being Longisquama insignis. The opposing viewpoint was, since the 1970s, largely anchored onto Deinonychus antirrhopus and John Ostrom’s and Peter Wellnhofer’s revision of Archaeopteryx lithographica – the latter unquestionably considered a bird by both sides of the debate. But despite comparisons, the one side argued that Archaeopteryx could not be compared to dinosaurs, or any theropod, for a variety of reasons, include pelvic structure, dentition form, hand morphology and digit count (a debate that still goes on) and even the presence of feathers. It is the last that concerns Sinosauropteryx prima, as it clearly lacked most of the other “avian” traits; but the presence of feathers in an animal so clearly non-avian would concretely place it in the middle of the debate over bird origins.
It was the joint attention on Sinosauropteryx prima‘s integument, the bizarre structures on Longisquama insignis, and the incredibly well preserved but ambiguous innards of Sciopnyx samniticus, and their role in the dinosaur-bird debate, that would cement the two camps. On the one hand, you had BAD (Birds Are Dinosaurs, whose proponents are “BADists”) and there aren’t any sub-camps; on the other, you have BAND (Birds Are Not Dinosaurs, whose proponents are BANDits), which now contains a group called MANIAC (Maniraptorans Are Not In Actuality Coelurosaurs, whose proponents are MANIACs – you can guess who coined the terms by their obvious political slant). The latter also contains a semi-movement, not so much a group, which might be called ABSRD – Anything But a Small, Running Dinosaur, a nod to the last sub-camp’s argument that, no matter what, birds cannot arise from within Dinosauria, even if they have to take theropods with them. In my last post on this, I referred to this as “moving the goal posts,” but it will stop when that moving wall hits Sinosauropteryx prima. There has developed a literature calling into question the integument of Sinosauropteryx prima, and TLS is in its forefront, previous installments which I’ve covered here, and here. It should be further noted that, as in those posts, several factors of the development of feathers from a conical structure which develops into a single filament, splits into a radiating morphology, or conjoins at the base to become a fan-like array, and eventually forms a pinnate, “true” feather, has been known for some time. The old case of the scale splitting into a feather is no longer plausible, though they do both develop in the same way …initially. Scales unfold into a flat, placard morphology, whereas feathers retain their conical structure and merely branch.
But TLS has turned his attention rather than to one of these earlier specimens of this species but to a newer, relatively undescribed specimen, IVPP V12415, which would make the third, if not fourth, known specimen of this rather “world shaking” taxon. TLS originally described this specimen with Alan Feduccia and Wang Xiao-lin in 2007, in which they attempted to firmly assert the integument was comprised merely of degraded, macerated, and irregularly bundled collagen fibers. In that paper, TLS and colleagues argued that the structure, often with bulbous shape, of the filaments that had been called “feathers” by other authors, were in fact similar to degraded collagen bundles. Collagen forming the structure of especially thick skin or “frills” were also used as the comparison to the line of parallel-fibered “halo” around the body, such that it would seem to form a thin, mid-line frill. TLS would later return to this in 2012 and asserted that this structure was of a mid-line frill, and even provided a reconstruction. This reconstruction focuses on the filaments recovered dorsal to the body, as they are mostly found at the back of the skull, atop the neck and back and hips, and down the tail and along its underside; it makes them into segments of a mid-line frill, much like you’d find in the Jesus Christ, or basilisk, lizard, Basiliscus spp.
TLS has chops, integument ones, and he supports this by working on detailed analysis of what seems to be dermal structures. In looking at detailed sections of Sinosauropteryx specimens, some which may not necessarily pertain to prima, TLS has argued that the structures termed “filaments” are continuous with other structures which seem to be dermal. In looking between the chevrons of the tail, around the arms, and behind the neck, it seems that different integumental structures – dermal and extra-dermal, as in fur or feathers – may be present; Currie and Chen found no significant difference between filament-like features in the tail that appears between neural spines or chevrons, and those above/below them between bones. Problematically, the material in these regions are preserved in different ways: there are clear, organic stains of various reddish or brown hues across the body, both those of the internal organs are stained darkest, such as within the eye socket, whilst those around the body contours are a brown-orange; but they are not preserved as details of the slab, merely colored stains. Instead, fibrous preservation, including of filaments implied to be “fuzzage” (see [n1], here) is almost always as impressions into the underlying sandstone/mudstone/tuff into which the specimen itself is an inclusion. These are finely detailed, though no higher in resolution than the grains that make up the rock, and they cross from the lighter colors between bones and around the body outline, and the dark colors beyond that, there the seeming fuzzage is implied. TLS had argued that their orientation and structure implied degraded and macerated collagen fiber bundles, showed several experiments where he and his colleagues had deliberately mangled such structures in fish to produce the right shape. In this, they ignored the artifice of the subject, unable to support this claim at the time simply using ichthyosaur or mosasaur skin impressions, which always show a clean, smooth, and decidedly non-fuzzy outline.
It is also a peculiarity that TLS did not turn his investigative technique to confirm that these structures could not be extra-dermal, that they had to be dermal features, by considering the preservation in the same Sihetun sediments for mammals, birds, amphibians, fish; whose outlines were as varied as one might wish. Rather, TLS focused on the argument of degraded collagen, and has not wavered from this assertion for almost a decade. It would be a simple thing, really, to see if the same organization and pattern of stains to impressions of integument occur in Sinosauropteryx prima and mammals, using the self-same techniques employed so far, and used to argue such interesting ideas. But this has never happened. I’ve mentioned before the work of Foth to produce hypotheses regarding how feathers of real, honest-to-goodness birds preserve, in the form of flattened bodies whose fluids expel from the body and cause the feathers to become irregular, splayed out from the body in an unrealistic halo, which completely obscures the natural outline the animal would have in life – or death. I have mentioned that, under impression experiments, it is difficult, even impossible to discern barbs in all but the largest and heaviest of feathers, almost certainly only the flight feathers of the wings, legs, tail.
This immediately recalls that Currie and Chen had hypothesized that some of the structures they discerned in the tail of Sinosauropteryx prima represented pinnate structures, which resemble portions of stage II feathers (as defined by Prum, 1999, and Prum & Brush, 2000). The regularity of the branching, in precisely one other direction and evenly separated from each other, at a high angle to the main preserved filaments, as if they were raches of a single feather, suggested to Currie and Chen that these were, in fact, true feathery structures, antecedent to the morphology in modern birds; to Prum and Brush, they might be stage III structures, which occur as part of the development of modern feathers. It is this specific allegation that TLS has bent much of his work attempting to disprove, and almost always, he has relied on the implication of collagen fibers as the better analogue, including the use of a cross-hatched appearance,
TLS, concerned over new preparation on the specimen, has rushed to publish pre-prep photographs and examinations from them of IVPP V12415, as several regions were of concern to him. Below, in A, you will see the whole specimen, which is almost complete. The proximal chevrons have become somewhat disarticulated from their caudal vertebrae, and the middle caudals appear to break off. TLS calls attention to this latter point, as there appear to be two distinct caudal sequences, and as shown below, identified these as being the ends of a precise break in the caudal vertebrae. Between these, running perpendicular to the vertebral sequences, are features (either relief or incised features of the matrix) which have a elongated, longitudinally-running arrangement and which TLS argues are scales:
To be precise, TLS argues that these structures show regular width (they do) and are segmented (they seem to be) and are approximately parallel to one another between the vertebral sequences. These, TLS argues, are indications that the tail of Sinsoauropteryx prima was scaled, and that this disproved the assumption that the filaments (visible before and after the break along the tail) are structural fibers, collagen, and nothing more. Indeed, TLS points to a series of marks cranial to the second caudal section, and calls these “collagen,” pointing out their pattern as evidence for cross-hatching in a clear collagenous array.
TLS published last year a reconstruction of a decomposing, opisthotonic Sinosauropteryx prima, and in this has been somewhat concerned with the connection of opisthotonus when it comes to dessication in animals. In this discussion, he offered the idea that opsithotonus will cause elements of the vertebrae to become arranged in a manner quite distinct from their normal orientation, pointing to the base of a juvenile ostrich neck to suggest that the tail itself can, in fact, form a zig-zag arrangement in opisthotonus. Problematic with this hypothesis is that opisthotonic posture tends to follow natural curvature of the vertebrae, and in TLS’s bird example, the neck forms a strong C- (or L-) shaped arch, which this posture is natural in life … but in the tail? As contributing evidence for this, TLS offers the presence of scales in a perpendicular line to the tail series to show that there was in fact a Z-shaped tail section. For various reasons, I think this hypothesis is erroneous.
1. The preserved tail sections are not contiguous. Distal to the first section of the tail, there is a series of impressions of the tail, and this continues well past the cranial end of the distal section of vertebrae. This is visible in the original fossil, and in closeups, and is illustrated below.
2. The “scales” are probably not scales. As TLS himself refers to in his latest paper, the paper was necessary because the fossil was to undergo new preparation. In his mind, TLS is documenting features he feels were natural to the animal, but which the preparators feel were not, or required to be removed for better presentation. The reasoning is not given, and perhaps this is to support TLS’s position that he is correct to document these structures so as to preserve his views on things. As such, the observations and figures in the paper are produced almost solely from photographs. This is, by itself, not a [major] problem, but it means that we have to consider that the reasoning for preparation is to remove matrix from the slab, probably from the overlying material that failed to separate onto the counterslab when the block containing the fossil was split. My impression here is that lighting is the key: structures in relief are highlighted from the side nearer the light, in shadow the other side; structures incised, or negative, are highlighted from the side further from the light, and in shadow the other side. In this way, when the lighting source is from the top left, this is the side of a structure in relief (such as a bone) will appear lit, whilst a structure in negative, or impressed, will appear lit from the other side. So determining where the lighting comes from, and then where shadows falls, indicates (approximately) whether a structure is an impression or an inclusion.
In IVPP V12415, the caudal vertebrae are preserved in both forms, and the impressions of distal caudals of the distal segment show shadows on the side nearer the lighting source (top left), as do the array of marks between vertebral sequences. These, demonstrated by long rounded marks of relatively equal width, suggest to me rotary grinder tool marks. They may have been made by preparators examining the region between vertebral sequences. However, some features of the slab continue along the line formed by these “scales” towards an end of the slab, and these are decidedly not scale-like, and elicit no comment from TLS. Furthermore, they start out similar in width to those more proximal, but on one side become more skewed, and the marks appear to radiate outward. These structures appear as possible trace fossils, but they are certainly not scales, unless they represent scaly structures that somehow rotate completely from the direction of the tail and extend off the slab.
3. Those “collagen” fibers aren’t.
The collagen traces, traces which TLS argued represented the distended fibers as the two sections of tail pulled apart, are clearly irregular incisions into the matrix formed as large tooling marks around the tail. When matched side by side, sections of the tail at the break and further distally show the same orientation of criss-crossing marks, are regularly equidistance from one another and occur in only two orientations. Impressions of integument should show more irregular orientations, skewing, and even curvature of the elements when they are part of the rachis identified as a “feather.” None of these photos show traces of filaments, and they are not similar to the traces left around regions where filaments are identified, thus there is no conflict with identification of filaments with collagen. These instead look like sections where the preparators were examining the slab between vertebral sequences.
In either this case, or 2 above, TLS doesn’t attempt to discount the nature of either trace fossils, preparation artifacts, or irregular splitting on the slab. Perhaps TLS merely thinks preparators are removing “obvious” evidence of non-filamentous integument in these fossils, so as to support the orthodoxy (as Feduccia put it). But this seems irregular: If you want to make the case that there are non-filamentous integument, perhaps one should demonstrate what filamentous integument looks like, and how to discriminate it from non? Like, look at a mammal or two? Moreover, these structures end abruptly at scarps in the slab, scarps similar to the ends where bones terminate, such as the distal end of the proximal sequence of vertebrae.
But, what if?
Well, what does it matter if this speculation about integument doesn’t bear weight? What if those really are scales? Well, not much, actually. You can have animals with two types of integument. You can have scales and feathers on a single animal (birds), scales and fur (mammals; scale-tailed squirrels, pangolins, in which stiffened hair-based structures form scale-like shapes), or scales and filaments:
Scansoriopteryx heilmanni is one of the best described specimens of a small, bird-like theropod dinosaur group called Scansoriopterygidae, which I mentioned here. And a small patch of rounded, pebbly scales on the tail seems out of place when the rest of the body suggests it was covered in filaments. These are especially important when their form (thin lines in the slab) are in the same form as Sinosauropteryx prima, whilst their positions (around the head, extended outward from the arms) are in direct contradiction to the idea that these might represent collagen bundles. Similar filaments with similar distribution are found in the therizinosauroid Beipiaosaurus inexpectatus:
The presence of scales itself does nothing to disprove the presence of filaments. It is merely that TLS assumes the “scales” in IVPP V12514 are preserved where they became disarticulated through opisthotomus, but failed to account for preparators’ artifacts, trace fossils, or other natural processes. In many ways, TLS was ignoring potential outcomes which would reject his overall hypothesis. He took this,
and turned it into this,
If the filaments of Sinosauropteryx prima reflect the presence of a frill, then they should indicate so also in a larger variety of animals. That might make a bit of sense if one felt that region of preservation reflected place of likely habit. That is, that if you get preserved at the bottom of an lake, you dwelt in the lake. The fin-tailed Sinosauropteryx only makes sense when TLS makes it an aquatic animal, without any other feature of the animal’s anatomy indicating this. What does that say for the various birds, mammals, insects, etc.? I do not think TLS has thought this one through.
TLS is a fisherman, one may say, one searching for “the big one,” but like most fishy tall tales, he’s searching for an ephemeral creature, one which is more alive in his imagination than exists.
Currie, P. J. and Chen P.-j. 2001. Anatomy of Sinosauropteryx prima from Liaoning, northeastern China. Canadian Journal of Earth Sciences — Revue canadienne des sciences de la Terre 38(4):1705-1727.
Czerkas, S. A. & and Yuan C.-q. 2002. An arboreal maniraptoran from northeast China. in Czerkas, S. J. (ed.) Feathered Dinosaurs and the Origin of Flight. The Dinosaur Museum Journal 1:63-95.
Foth, C. 2012. On the identification of feather structures in stem-line representatives of birds: Evidence from fossils and actuopalaeontology. Paläontologische Zeitschrift 86:91-102.
Lingham-Soliar, T. 2013. The evolution of the feather: Scales on the tail of Sinosauropteryx and an interpretation of the dinosaurs opsithotonic posture. Journal of Ornithology 154:455-463.
Lingham-Soliar, T., Feduccia, A. & Wang X.-l. 2007. A new Chinese specimen indicates that ‘protofeathers’ the Early Cetaceous theropod dinosaur Sinosauropteryx are degraded collagen fibres. Proceedings of the Royal Society, B: Biological Sciences 274:1823-1829.
Prum, R. O. 1999. Development and evolutionary origin of feathers. Journal of Experimental Zoology 285:291-306.
Prum, R. O. & Brush, A. H. 2002. The evolutionary origin and diversification of feathers. The Quarterly Review of Biology 77(3):261-295.
Xu X., Tang Z.-l. & Wang X.-l. 1999. A therizinosauroid dinosaur with integumentary structures from China. Nature 399:350–354.