Uncinate processes in birds, crocodylians, and their fossil relatives
Almost all birds alive today, with the minor exceptions of the Australasian incubator birds and the South American screamers, have hook-like bony prongs protruding backwards from several successive pairs of ribs in the shoulder region. Scholars and their students have been pondering the possible function of these structures, called uncinate processes, for centuries, with structural reinforcement of the ribcage and enhancement of ventilation as the most common hypotheses. Nearly two decades ago, the ventilatory hypothesis received important experimental support from a study carried out in the Canada goose by J. R. Codd and some colleagues. Perhaps, their work suggested, the uncinate processes of birds act like levers, contributing to the ability of the trunk muscles to pump air into and out of the body. Crocodylians, the closest living relatives of birds, have uncinate processes too, but crocodilian uncinate processes take the form of soft, cartilaginous tabs rather than rigid bony hooks. How crocodylian uncinate processes might function, and whether these structures have a common evolutionary origin with the uncinate processes of birds, are open questions.
Bony uncinate processes broadly resembling those of modern birds occur in many extinct members of Pennaraptora, the group of dinosaurs that includes birds and their close relatives, and in the somewhat more distantly related ornithomimosaurian Pelecanimimus. These dinosaurs, including birds, are all classified among the predominantly bipedal and carnivorous Theropoda. Elsewhere in Dinosauria, several members of the herbivorous group Ornithischia possess tabs on their ribs. These structures are conventionally called “intercostal plates”, but look suspiciously similar to the uncinate processes of crocodylians, and I consider them true uncinate processes. They may contain some bone, but are probably made predominantly of cartilage that was hardened during life by a process called calcification. Araripesuchus, a Cretaceous relative of crocodylians, has spike-like uncinate processes that are also likely composed of calcified cartilage.
Taken at face value, the evidence indicates that bony or calcified uncinate processes evolved separately in Araripesuchus, some ornithischians, and a subset of theropods including birds, with soft cartilaginous uncinate processes occurring only in crocodylians. However, the logic of evolution and fossil preservation suggests another possibility. The softness of normal, uncalcified cartilage ensures that this substance almost never survives the process of decay to enter the fossil record. Soft cartilaginous uncinate processes like those of modern crocodylians could have been widespread in dinosaurs, without ever turning up in an actual fossil quarry. Perhaps they were even ubiquitous in Archosauria, the group that contains birds, other dinosaurs, crocodylians, and their assorted close relatives. Crocodylians, Araripesuchus, ornithischians, and theropods could conceivably all have inherited uncinate processes from the common ancestor of archosaurs, with evolution either calcifying the original soft cartilage or transforming it into bone in some but not all cases. Such a scenario would change our picture of the anatomy of extinct archosaurs, by adding uncinate processes to their ribcages, and would imply a new understanding of uncinate process evolution. But how could it be tested?
Finding uncinate processes that aren’t there
In 2016, my coauthor Corwin Sullivan and I paid a research visit to the Evolutionary Studies Institute at the University of the Witwatersrand in Johannesburg. At that point we were just getting interested in uncinate processes, and in the possibility that they might have once been common in extinct archosaurs. We pulled some ribs of a modern Nile crocodile out of a drawer in the Institute’s fine anatomical collections, and saw exactly what we’d been hoping for – distinct facets, which we soon took to calling “uncinate scars”, on the edges of the ribs where cartilaginous uncinate processes had once been attached. If we could find similar scars on the ribs of extinct archosaurs, we could deduce that they must have had uncinate processes too, even if the processes themselves weren’t preserved.
That promising discovery had an unexpected sequel on our last night in South Africa, when we visited a game restaurant with several colleagues. Alongside impala and kudu meat, among other delicacies, crocodile ribs were on the menu, and Corwin took one and managed to scrape off enough of the succulent flesh that we were able to get our first good look at an intact crocodylian uncinate process. More formal crocodile dissections followed upon our return to the Institute of Vertebrate Paleontology and Paleoanthropology in Beijing, where we were both based at the time. Furthermore, I learned that uncinate scars occur not just in crocodylians, but also in birds that are young enough for the bony uncinate process to be bound to the rib by soft tissues. In adult birds, the uncinate processes are normally fused to the ribs and cannot be detached without breaking them.
With the help of funds from the Dinosaur Research Institute, I went to look for uncinate scars in major palaeontological collections in Canada, China, and the United States. Staff at the museums I visited were helpful and hospitable, and at Yale Peabody Museum and the New Mexico Museum of Natural History were even gracious enough to remove troublesome rock from specimens in their collections that I wanted to scrutinise. At the American Museum of Natural History in New York, I was lucky enough to find an incomplete fossil crocodylian rib bearing an uncinate scar that looked nearly identical to the modern crocodile uncinate scars we’d seen in South Africa. This gave our spirits a nice boost, as the observation provided evidence to validate our use of uncinate scars to infer the presence of uncinate processes.
At the Canadian Museum of Nature in Ottawa, a skeleton of the tyrannosaurid Daspletosaurus torosus was on display in the centre of the dinosaur gallery. As I passed through a crowd of visitors and looked at D. torosus from the side, uncinate scars on five consecutive ribs gradually revealed themselves to me. The previously undescribed scars had been hidden in plain sight, as that D. torosus mount has been on display since the 1970s!
Making sense of it all
When the collection visits were done and dusted, I had found 66 individual uncinate scars distributed among at least 19 fossil archosaurian species, though many of the scars were on specimens that were too incomplete to be identifiable to the species level. I had also seen 19 preserved uncinate processes, all undoubtedly made either of bone or calcified cartilage. In total, I had evidence of the presence of uncinate processes in at least some members of several major groups of dinosaurs, of course including birds, and in three extinct groups of crocodylian relatives, as well as in actual crocodylians. These groups were scattered across the archosaur family tree, indicating that uncinate processes might have been a widespread archosaur feature. But just how far back in archosaur evolution did they first arise?
Corwin and I worked with Leon Claessens, an expert on the evolution of archosaur respiratory evolution who has known Corwin since they were PhD students together, to make sense of our data. To test whether uncinate processes are likely to have been an ancestral feature for dinosaurs, or even for archosaurs as a whole, we used R, a statistical computing language and environment, to run what is called an ancestral state reconstruction. This type of analysis, which can be carried out with a variety of algorithms and starting assumptions, uses the known distribution of a feature on a phylogenetic tree to infer that feature’s potential evolutionary history. When we applied this method to analyse our data on the distribution of uncinate processes, trying different algorithms and assumptions to test their effects, the analysis strongly indicated that the presence of cartilaginous uncinate processes was the ancestral condition for both Dinosauria and Archosauria. Such results are always probabilistic, and subject to change as new data come in, but presently available evidence suggests that soft cartilaginous uncinate processes were inherited from ancestral archosaurs and broadly distributed across Archosauria, evolving into bony or calcified uncinate processes in comparatively rare cases where additional strength and stiffness might be beneficial.
This finding raises the question of why ancestral archosaurs needed uncinate processes in the first place. Could early archosaurian uncinate processes have played a role in ventilation, like their counterparts in modern birds? A 2019 study carried out in the American alligator, again led by J. R. Codd, provided interesting empirical evidence that this might have been the case. Crocodylian uncinate processes are largely embedded in a back muscle, the iliocostalis, and Codd and his coauthors showed that this muscle is activated during exhalation. When an alligator is breathing hard, they wrote, contraction of the iliocostalis appears to force the ribs to swing “inwards and backwards”, helping to drive air out of the lungs. Assuming that the uncinate processes help enable the iliocostalis to rotate the ribs, as seems likely, this discovery demonstrates that a ventilatory function for uncinate processes is not confined to birds and may even have been present in the common ancestor of all archosaurs, which is also the most recent common ancestor of birds and crocodylians.
If uncinate processes did indeed assist with ventilation in ancestral archosaurs, then they represent a fundamental archosaurian adaptation for enhanced ventilatory performance that was ultimately inherited by modern crocodylians and birds. Among birds, in particular, uncinate processes are part of a sophisticated breathing apparatus that now supports the demands of powered flight, but that role is remote from the context in which they first evolved.