Two Tales


Birds can have resplendent tails. Wonderful arrangements and bizarre shapes. We may all be familiar with the lyrebird, whose male’s lateral tail feathers (retrices) have been modified from their typical planar vaned structure into a pair of curly feathers bracing an array of plumes whose vanes are sparsely ornamented but colored. The peafowl male’s “tail” isn’t; there are a long array of beautiful feathers, but these are actually modified from contour feathers of the back and base of the tail, whereas the normal retrices are a stiffened array. The “tail” drapes over the real tail, and when the latter is lifted, the “tail” is also, and thus spread out. Birds preen their retrices, plucking the vanes to created the lyrebird-look, or have developed unusual shapes, such as Wilson’s bird-of-paradise in which the retrices have become dense and curled. Quetzals have enormously long retrices in shimmering pearlescent hues. The racket-tailed hummingbirds have two long retrices to accompany their normal irridescent — but normally-shaped — tail feathers but are nothing but long raches terminating in a fan-like array: the eponymous “racket.”

But a recent paper has suggested that this diversity in tail shape is far older than even modern birds.

The first ancient tail is almost certainly that of Archaeopteryx lithographica. Upon examination in 1861 of the specimen that would eventually become the London specimen (NHM 37001), Hermann von Meyer concluded that his fossilized feather and this specimen were evidence of a new bird, and so named Archaeopteryx lithographica. The London specimen beautifully preserves the tail, extended downward along the slab, with nearly symmetrical arrays of feathers to both sides, suggesting that the bony bit supported a line of retrices, unlike any modern bird. Living birds, rather, have a shorter bony tail surrounded by a mass of specialized muscles, and it is into the fat and muscle that the quills of the tail are embedded. Yet for Archaeopteryx, the feathers appear to link into the bone. This is made even more likely when one considers that the tail is seemingly inflexible, nothing much more than a body rod, and the supports for the leg-pulling and tail-flexing muscles were located almost strictly at the base. So all those feathers would also likely have been immobile relative to the tail.

(In a curious twist, Christian Foth and Oliver Rauhut reported at SVP in LA at the beginning of this month that it seems likely that Hermann von Meyer’s initial feather may have been a retrex — a tail feather, rather than a wing feather [or remex, pl. remiges], as has long been assumed.)

Archaeopteryx seems to have been building on a bauplan for the tail, one started long before it in the early Jurassic, when it seems its theropod ancestors would have had a seemingly normal, “reptilian” tail which may very well have been covered in a dense or sparse covering of filamentous “fluff.” The tail, muscular down to nearly its tip, would have been highly mobile. But more advanced, birdier theropods have stiffer tails, and so less muscle would work, and as such less muscular attachments are found in the distal and even middle tail. And some of these, such as the oviraptorosaurians and dromaeosaurs, show evidence that the tail tip bore a large fan-like array. The first of these was described as Protarchaeopteryx robusta, and it suggested to its describers this more “primitive” bird would have preceded Archaeopteryx‘s very avian tail with just a distal “fan.” Archaeopteryx, then, would develop the fan towards the base of the tail, thus paving the way for a tail shortening.

A long bony tail with an array of feathers down the side at the beginning of avian evolution spells an interesting origin for further bird tails. Not long after Archaeopteryx, Confuciusornis sanctus in the ancient lagoons and forests of eastern Asia would provoke its own twist to the development of a “fluffy” hind section. Rather than a long array of feathers, or a broad fan as in many modern birds, a short array of retrices are found, but two long thin feathers are also present. The tail, then, is more like modern birds, although some doubt is present in whether the two thin “streamers” are actual retrices. One angle, as it were, is to examine the angle of deflection, and it does seem these two strap-like feathers arise from the tail, and thus Confuciusornis isn’t sporting a peafowl-like “pseudotail.”

But a new discovery attempts to further twist what may have happened to early avian tail evolution between Archaeopteryx and Confuciusornis. One might presume, as I did above, that the feathery array preceded tail shortening rather than assume, as some might, that Archaeopteryx‘s tail is an exaptation unique to it. Normally, fossils of the primitive bird Shenzhouraptor sinensis (Ji et al., 2002) are found plastered on their sides, their tail extended behind them, and their wings spread. All the integument that is recovered is a faint halo around the main body and head, and like many fossils from Liaoning, the extended arms. Long remiges are common in winged fossils from the Early Cretaceous lagerstätten of Liaoning and Hebei provinces, and across Liaoning’s border with the autonomous region of Inner Mongolia.

But what is less common are feather remnants, impressions or fossilized remains, closer to the body way, including around the legs and tail. Feather preservation is tricky (see here and references therein) and not all of a feather gets preserved. Most problematic, and an issue that has to some degree confounded reconstructions of Archaeopteryx‘ wings, is that feathers on the limbs in birds come in layers — there are not merely a single layer of slightly overlapping structures terminating on the trailing edge. Instead, feathers are differentiated into deep layers of down covered by contour feathers, and the large flight feathers extend from among these on the arm and tail; contour feathers on the arms and tail, on under and over-sides are termed “coverts” and these feathers are adapted to have similar shapes and properties to flight feathers with which they are associated, include certain levels of asymmetry.

Shenzhouraptor sinensis, which has been typically termed Jeholornis prima in the most recent literature surrounding is — and issue which, sadly, cannot be easily resolved even by resort to the ICZN despite claims from the describers of Jeholornis prima that it can, a problem that is fairly lengthy in explanation — has received new material in the form of several specimens described by Jingmai O’Connor and colleagues that, for the first time, reveal the tail integument. And one specimen suggests that there is something odd going on.

First, a distal tail array of feathers is known. Such a tail arrangement is known in some other paravian theropods, namely Microraptor zhaoianus (as Microraptor gui; see Turner et al., 2012) in which only the distal tail region is so enfeathered. But an additional arrangement of tail feathers is also known, and here is where things get sketchy. First, O’Connor relate us to several specimens preserving partial tail integument, including two specimens (holotype and paratype) referred to the not-jokingly-named Jeholornis palmapenis (no, really: it means “frond feathered,” so stop snickering), and these specimens demonstrate an arrangement including distal “fronds” around the tail tip, and some feathers further basal. Oddly, the middle section of the tail hasn’t demonstrated any preserved feathers, but feathers are preserved at the base, and here is where things get odd.

One specimen in particular is very well preserved, STM2-18, and was previously mentioned by O’Connor et al. (2012) in their description of Jeholornis palmapenis. O’Connor et al. note in this new paper that, rather than seemingly being displaced wing feathers, they were actually basal tail feathers, and represented retrices forming a basal fan.This, O’Connor et al. implied that the evolution of the tail didn’t progress with a broad fan attached to an ever-shortening tail, but rather a broad fan with a unique structure, a form of “racket,” way at the end of the bony tail; thus, that the shortening of the bony tail had little to do with the development of the avian-style tail fan.

Much work in the last two decades has been done to figure out how the avian wing-based locomotion module developed, and part of this was the argument brought forth in the 1990′s by Steve Gatesy and others that the tail is a part of the leg module of most terrestrial theropods, as indeed it is in many other sauropsids, including crocodilians. But towards the origin of birds, the tail eventually becomes decoupled from the hindleg and becomes a third locomotory module, which birds today use as part flight control, steering, and even partial lift-producer. Part of this work has been developed by the work of John Hutchinson on one hand and by Scott Persons and Heinrich Mallison on the other. All of these workers have looked into the coupling elements of leg and tail, and found that leading to birds, a distinct decoupling occurs where the tail functions as a locomotory module while it is still a long, bony thing, supporting tail feathers as in Archaeopteryx, where the feathers are preserved down the length of it. Such a tail functions well because it is long, stiff, and only the base is acted upon by muscles. But such a tail is costly, both to produce and to maintain. The hypothesis goes: To reduce cost of production and maintenance, the most expensive portions had to be reduced while preserving function, and this meant to reduce the bony tail but increase the feather component beyond previous sizes. Feathers are still costly, but may be controlled through passive muscles. If so, then the muscles that already existed could control the feathers instead of the body tail, and you could even shave off a few quills in order to bring them all into a more economic function. And of course, organisms can exapt these existing structures into a more economic form by selecting for better performing animals: Animals with more feathery and less bony tails would perform better because they were able to retain more energy, depending on their actions. They would tend to survive, being fit to survive.

But O’Connor’s model differs, and there may be some reasons why this difference makes the hypothesis less likely.

Jeholornis remigial-retricial feather distribution sm

A, composite model; B, STM2-18; C, STM3-3; D, STM@-11; E, STM3-30; F, STM3-4; G, SDM2009.109.1. 1, holotype of Jeholornis palmapenis, mainslab. Primaries; 2, Secondaries; 3, Tertiaries; 4, basal “fan” sensu O’Connor et al.; 5, slightly more distal basal feathers; 6, middle section; 7, section not preserved on any specimen; 8, distal “frond”; 9, reconstructed tail fan; 10?, possible open slot may be covered by feathers extended from the thigh. All left sides Show preserved portions, in comparison to the right side, which shows the full hypothesized preservation possible (as in A).

The first reason is that preservation of the tail structures suggests that the feathers extended along the bony tail, and were not limited to the tip, regardless of whether a basal fan was present. Preservation models in Liaoning are subject to degrees of veracity: Not everything is ever preserved, and contour and covert feathers tend to be least well preserved. The array of specimens referred by O’Connor et al. to Jeholornis (Shenzhouraptor), as shown above, demonstrate this for both the wings and tail.

The second reason is the quality of the locomotory model. The new model by O’Connor does not fit neatly into this, certainly not well enough to explain the gradual transformation of a retricial array along the bony tail into a retricial fan. Problematically, the muscle control of the basal fan wouldn’t even be as useful or exaptive due to their apparent nature as canards. Despite this, O’Connor et al. demonstrate some aerodynamic functionality of their hypothesis, noting higher performance than without the canards; but, that’s the point of canards.

Andrea Cau was the first to suggest that O’Connor et al. had identified possible tail coverts as remnants of a retricial fan, and this idea has merit (see E, below). But I am more inclined to think that the full-tail model, a slimmer version of the Archaeopteryx morphology, is more sound. It, at the least, fits the modular evolution of the disengagement of the tail from the hindlimb, which precedes tail retraction on the evolution of birds. Retention of a distal fan (not “frond”) implies a mechanical function is being selected for, and this functionality appears to flow more neatly into the “short” tail form seen in Confuciusornis sanctus.

Evolutionary shortening of the tail isn't so simple. A, basal tetanuran had filamentous integument; B, short-tailed maniraptoran oviraptorosaurs appear to have a broad distal fan, but short proximal tail feathers; C, dromaeosaurids had long stiffened tails with distal fans, but almost no proximal feathers; D, Archaeopteryx had relatively long feathers all along the tail; E, a reconstruction of the tail of Shenzhouraptor (Jeholornis) with the basal feathers reconstructed as coverts; F, Shenzhouraptor (Jeholornis) presented as argued by O'Connor et al.; G, a short-tailed bird, Confuciusornis sanctus had two much longer tail feathers; H, the magnificent tail of the male peafowl, Meleagris pavo.

Evolutionary shortening of the tail isn’t so simple. A, basal tetanuran had filamentous integument; B, short-tailed maniraptorans, here a basal oviraptorosaur, appear to have a broad distal fan, but short proximal tail feathers; C, dromaeosaurids had long stiffened tails with distal fans, but almost no proximal feathers; D, Archaeopteryx lithographica had relatively long feathers all along the tail; E, a reconstruction of the tail of Shenzhouraptor (Jeholornis) with the basal feathers reconstructed as coverts; F, Shenzhouraptor (Jeholornis) presented as argued by O’Connor et al.; G, a short-tailed bird, Confuciusornis sanctus had two much longer tail feathers; H, the magnificent tail of the male peafowl, Meleagris pavo. Arrows point to a deep V-split between left and right halves of the tail vanes. A V-split may be present in Archaeopteryx as well as Confuciusornis.

I am thus far more inclined to qualify this preservation as incidental, and not clarifying or confusing of tail evolution in birds. Instead, it merely reinforces the need to consider preservation biases and the vagaries of interpreting the lack of preservation as the preservation of a thing’s lack. It is a reasonable argument, I think, that O’Connor et al. attempt to disjunct the established model of avian tail evolution, but reliance on poor preservation in some specimens seems a poor way to test this model.

Gatesy, S. M. 1990. Caudofemoral musculature and the evolution of theropod locomotion. Paleobiology 16: 170-186.
Gatesy, S. M. 1990. The evolutionary history of the theropod caudal locomotor module. pp.333-346 in Gauthier & Gall (eds.) New Perspectives on the Origin and Evolution of Birds. Peabody Museum of Natural History, Connecticut.
Gatesy, S. M. & Dial, K. P. 1996. From frond to fan: Archaeopteryx and the evolution of short-tailed birds. Evolution 50: 2037-2048. http://dx.doi.org/10.2307/2410761
Hutchinson, J., Bates, K., Allen, V. 2011. Tyrannosaurus rex redux: Tyrannosaurus tail portrayals. The Anatomical Record 294 (5): 756-758.
Ji Q., Ji S.’a., You H.-l., Zhang J., Yuan C.-j., Ji X., Li J. & Li Y. 2002. [Discovery of an avialae bird - Shenzhouraptor sinensis gen. et sp. nov - from China]. Geological Bulletin of China 21 (7): 363-369. [in Chinese with English abstract].
Mallison, H. 2011. Defense capabilities of Kentrosaurus aethiopicus Hennig, 1915. Palaeontologia Electronica 14 (2) 10A: 1-25.
O’Connor, J. K., Sun C., Xu X., Wang X.-l. & Zhou Z.-g. 2012. A new species of Jeholornis with complete caudal integument. Historical Biology 24 (1): 29-41.
O’Connor, J. K., Wang X.-l., Sullivan, C., Zheng Z.-t., Tubaro, P., Zhang X.-m. & Zhou Z.-h. [2013.] Unique caudal plumage of Jeholornis and complex tail evolution in early birds. Proceedings of the National Academy of Sciences, Philadelphia [Published online ahead of print, 7 Oct, 2013; DOI:10.1073/pnas.1316979110]
Persons, W. S. IV & Currie, P. J. 2011. The tail of Tyrannosaurus: Reassessing the size and locomotive importance of the M. caudofemoralis in non-avian theropods. The Anatomical Record 294: 119-131.
Persons, W. S., IV, Currie, P. J. & Norell, M. A. [2013.] Oviraptorosaur tail forms and functions. Acta Palaeontological Polonica [Published online ahead of print: 4 Jan, 2013; DOI:10.4202/app.2012.0093]
Zhou Z.-h. & Zhang F.-c. 2002. A long-tailed, seed-eating bird from the Early Cretaceous of China. Nature 418: 405-409.

About these ads
This entry was posted in Paleontology, Science Reporting, Taphonomy and tagged , , , , , . Bookmark the permalink.

2 Responses to Two Tales

  1. Andrea Cau says:

    My “hip covert” hypothesis stems mostly from fig. S1 in O’Connor et al. (2013): the anteriormost “tail” feathers in that specimen appears as anchored above the ilium, thus are more plausibly interpretable as hip coverts instead of (tail) rectrices.

    • This is what made your idea really reasonable. But I am concerned nonetheless that distortion is present regardless. Measuring angles to find some structure doesn’t dismiss the feathers as being displaced, even as O’Connor et al. originally surmised. Taken at face value, and applying the model of tail coverts, I would assume that an Archie-like tail would preserve coverts running down the sides parallel to the retrices and not angled towards the midline as O’Connor et al. presume. So if the Archie-model is incorrect and the feathers are correctly angled, but the “unique tail model” of O’Connor makes little mechanical sense and the feathers are coverts, it suggests at least that the tail feathers so shown are not even coverts, merely long contours. Preservation is crappy nonetheless. It is easier to adapt the specimens into the Gatesy model than to presume some divergence from it, and I think the best interpretation of the fossils in the Gatesy model support the tail feathers as either retrices (my original surmise) but deflected, or coverts/contours that form some other function than relate to the aerial function. That said, the tail appears to be retricial along its entire length, and O’Connor et al.’s conclusion regarding it erroneous. We shall see!

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s