Shaking the tree ...

Evolutionary biologists are constantly striving to improve our knowledge of the tree of life. Although this may seem like an esoteric exercise, of interest to a select few, understanding the pattern of evolution is fundamental to everything else that we might want to know regarding evolutionary history. Schemes of relationships offer a framework for understanding rates of evolution, the geographical history of a group, the ways in which different anatomical, physiological and behavioural features develop and change, and provide insights into the dynamics of adaptive radiations and extinctions. Given the foundational nature of this work, it should be unsurprising that scientists are forever revisiting the relationships of even the most familar of animal groups, in order to check and double-check that they've got things right. An impressive recent example of this comes from work on the interrelationships of the major mammalian groups, where new analyses of primarily molecular data were able to solve many features of the mammalian tree that were previously mysterious and that created a number of new, previously unsuspected groupings, such as Afrotheria, that paved the way for a different understanding of mammal evolution.

In 1887, Harry Govier Seeley, then a Professor at Kings College London, proposed a new classification for a group of extinct reptiles whose remains were being unearthed in Europe and North America. These animals, dinosaurs, were known to share a number of features in common, but there had been little consensus over their relationships and how they should be classified. He noticed several features that differed consistently between the various animals that had been assigned to this group and proposed that they could be divided to into two great tribes - Saurischia (the 'lizard-hipped' dinosaurs) and Ornithischia (the 'bird-hipped' dinosaurs). Unsurprisingly, the most obvious feature he used was hip structure, but Seeley also noted other features that supported these groupings and that distinguished them from each other. Seeley considered that his two groups were probably not particularly closely related and that they arose from different ancestors. Indeed, his arguments were so persuasive that his scheme held sway unchallenged for almost 100 years. The first challenge came in the early 1970s when Robert Bakker and Peter Galton (1974) argued that Saurischia and Ornithischia were each descended from the same common ancestor, making Dinosauria a natural, or monophyletic, group, rather than distant relatives as Seeley and others had advocated. This challenge was debated by the scientific community and upheld by analyses of expanded datasets (and new computational methods for assessing relationships) during the 1980s. Since the late 1980s, the palaeontological community has accepted dinosaur monophyly as dogma (and expanded this dogma to include birds firmly within the dinosaurian radiation), in comparison with the previous 100+ years where dinosaur polyphyly was the accepted model. However, in spite of this radical change, Seeley's dichotomous division of dinosaurs into Saurischia, composed of Theropoda and Sauropodomorpha, and Ornithischia survived (e.g. Gauthier 1986; Sereno 1999).

Since the 1980s, various dinosaur experts have looked at these fundamental splits in the dinosaur tree and their work upheld Seeley's model, with most of the debate concentraing on the exact positions within the tree of a few early, controversial species, such as Eoraptor and Herrerasaurus (e.g. Langer & Benton 2006). However, although these analyses often examined numerous detailed features they relied on only a handful of animals to inform the shape of the dinosaur tree. The past 20 years have witnessed a rush of discoveries of early dinosaurs, from all major lineages and from many parts of the world, and have also seen the recognition of a whole new group of close dinosaur relatives, the silesaurs (e.g. Nesbitt et al. 2010). This new information was incorporated into a variety of analyses, but all of these were focused primarily on the particular relationships of these new animals rather than the overall structure of the dinosaur evolutionary tree - probably because of the broad consensus that surrounded Seeley's neat and logical scheme.

The vast amount of new data now available, in combination with new software packages that allow huge datasets to be analysed rigorously and rapidly, seemed to offer an opportunity to take another look at dinosaur relationships with fresh eyes. With this in mind, my PhD student Matt Baron together with my former PhD advisor (and Matt's other advisor) David Norman, and I decided to see if this new information might affect our knowledge of the dinosaur tree. Matt built a data matrix containing 74 species of early dinosaurs and their close relatives, using specimens from all over the world and concentrated in the Middle Triassic–Early Jurassic, the time at which these major splits took place. Over 450 separate anatomical features were checked for each of these species and the resulting data analysed. It's worth bearing in mind that all other recent analyses of this problem included no more than 12–15 examples of early dinosaurs, on which to base the entire early evolutionary history of the group. Analysis of these data resulted in the recovery of an unexpected and radical tree topology that offers a challenge to Seeley's 130 year old hypothesis. This tree found that theropods were more closely related to ornithischians than either group was to sauropods, thus removing much of the content from Seeley's Saurischia (the carnivorous herrerasaurs remained with the sauropodomorphs, however). This new grouping of Theropoda + Ornithischia has been dubbed Ornithoscelida in a paper published in Nature that presents our new hypothesis (Baron et al. 2017). Ornithoscelida is a name originally proposed by 'Darwin's bulldog' Thomas Henry Huxley in 1870 in a classification that preceded Seeley's, but that was largely ignored. The name means 'bird-limbed' and refers to the hollow, gracile and elongate leg and arm bones found in theropods and ornithishians. In our analysis this new grouping receives strong support, with 21 features that seem to unite these animals to the exclusion of the sauropodomorph dinosaurs.

Given previous definitions of various dinosaur groups, this new tree requires some new definitions for commonly used names if we are to keep all dinosaurs as dinosaurs. For example, the most common definition (the common ancestor of sparrows and Triceratops and all of its descendants), would now exclude Diplodocus and other sauropods from Dinosauria. This was a step too far for us and so we proposed a new scheme of definitions to help stabilise the use and definition of some of these names, not only to account for if the new tree is potentially right, but also that would allow the tree to change if we were wrong without altering the meaning of these names.

In addition to giving us a new dinosaur tree, if our results stand up to detailed scrutiny by other palaeontologists, then they might be used to provide many new insights into dinosaur evolution. This would include looking again at the timing and pace of dinosaur origins, the evolution of various key characters, such as feathers and carnivory, and into the areas where dinosaurs might have first appeared. As with anything in science, our tree is a hypothesis - it is there to be stretched and tested to see if it is stronger than those that have come before and if it has the power to explain more about dinosaur evolution than other competing schemes. If we're wrong and there are better alternative explanations for the patterns we see then we'll have to accept that evidence and move on - that's how science works. In the meantime we hope that people will look at this with an open mind rather than rejecting it in a knee-jerk fashion due to its challenge to a well established and embedded dogma. While Huxley might be cheering us from the shades, we're also aware that Seeley would be shaking his head. Time will tell which one of them was closer to the truth.

References

Bakker, R. T. & Galton, P. M. Dinosaur monophyly and a new class of vertebrates. Nature 248, 168–172 (1974). 

Baron, M. G., Norman, D. B. & Barrett, P. M. A new hypothesis of dinosaur relationships and early dinosaur evolution. Nature (2017).

Gauthier, J. Saurischian monophyly and the origin of birds. In The origin of Birds and the Evolution of Flight (ed. K.Padian). Memoir of the California Academy of Science 8 1–55 (1986).

Huxley, T. H. On the Classification of the Dinosauria with observations on the Dinosauria of the Trias. Quarterly Journal of the Geological Society 26, 32–51 (1870).

Langer, M. C. & Benton, M. J. Early dinosaurs: a phylogenetic study. Journal of Systematic Palaeontology 4, 309–358 (2006).

Nesbitt, S. J., Sidor, C. A., Irmis, R. B., Angielczyk, K. D., Smith, R. M. H & Tsuji, L. A. Ecologically distinct dinosaurian sister group shows early diversification of Ornithodira. Nature 464 (7285): 95–98 (2010).

Seeley, H. G. On the classification of the fossil animals commonly named Dinosauria. Proceedings of the Royal Society of London 43, 165–171 (1887).

Sereno, P. C. The evolution of dinosaurs. Science 284, 2137–2147 (1999).