The geographical paths to dinosaur dominance

Understanding how the dinosaurs spread to every corner of the Earth and how they and their geographical distribution evolved

Go to the profile of Ciara O'Donovan
Feb 05, 2018
Upvote 2 Comment

Our paper published in Nature Ecology and Evolution can be found here: 

From as early as the age of four or five, most children are fascinated by dinosaurs. Children can name various species, know what they looked like, where they were from and what they ate. But despite our familiarity with this charismatic group of animals, scientists in the field know surprisingly little about dinosaur biogeography and how the dinosaurs’ widespread distribution came to shape their morphological and ecological diversity. The earliest known dinosaur fossils come from Argentina, so it follows that South America is thought to be where they originated. There is however a rich fossil record of dinosaur specimens which have been and continue to be found all over the world. How then, did one ancestral species that existed more than 230 million years ago, produce hundreds of subsequent species that spread from South America to become the dominant, terrestrial organisms, across the entire globe?     

In the first week of my PhD I set out to answer this question. Armed with a wealth of data downloaded from The PaleoBiology Database’s Fossilworks portal ( I was able to plot the locations of every dinosaur fossil catalogued in the database at that time. We needed to use this data matched to a large phylogenetic tree (which describes how dinosaur species are related to each other), to reconstruct the locations of every dinosaurian ancestor in the tree. With more data than I knew how to cope with and the tasks of learning both R (a statistical programming language) and phylogenetic comparative methods (that account for how dinosaurs were related to each other), my supervisor and I decided to carry out a pilot study.

We used the iconic Tyrannosaurus rex and its 13 closest relatives (Tyrannosauroids) as the focal group for our pilot study. These species have a distribution that spans Asia and North America with only one species in Europe. So, did their ancestors cross mainland Europe or go over the Bering Straits land-bridge to get between these two locations? And how many times did these dispersals occur? It was whilst attempting to answer these questions that we realised the existing method we were hoping to use to estimate our reconstructions was not working as anticipated. This method, not built for biogeographical data, was being tricked by the longitude-latitude coordinate system because this system describes locations as if they occur on a flat Earth, or map. Whilst we know that -180o longitude and 180o longitude describe the same line on the globe or that lines of -179o longitude and 1o longitude are virtually next to each other, conventional phylogenetic comparative methods don’t. Where numerically these values are far apart, geographically they are not which resulted in our model reconstructing ancestors on the wrong side of the Earth. We set out over the subsequent months to alter our model of evolution to work in a spherical world, applicable to biogeographical data on a global scale. At this point, both the scope and scale of the project had increased beyond what we had planned and what we could have expected at the beginning.

After three years of getting to know the dinosaurs and the Earth they lived on, developing some semblance of competence in R, running thousands (if not more) of regression models on a supercomputer and comprehending two-way interactions, we had managed to interpret a wonderful and elegant story about spatial distributions, speciation and evolution.

Our final paper sees the dinosaurs move backwards and forwards between continents that were continually moving and changing and we characterise the dinosaurs’ dispersals over 170 million years in terms of direction and speed of movement through time. We find that the dinosaurs initially moved quickly across the Earth which had recently been decimated by one of the largest mass extinctions in its history (the End-Permian extinction event, 252 million years ago). This left an almost clean slate for the dinosaurs to spread into and populate. As the Earth filled up with dinosaurs (along with other organisms such as early mammals), the dinosaurs continued to spread but they moved smaller distances and more slowly. This slow-down-through-time pattern we detect, fits in with over one hundred years of theory about evolutionary radiations which have previously only been revealed by morphological evolution and speciation. We find that the ecological forces governing how much dinosaur species moved and where they moved to acted on all dinosaurs equally such that the slow-down-through-time pattern is universal across all dinosaurs. This allowed us to hypothesise the modes by which new dinosaur species arose and how these might have changed through time as the Earth and its niches filled up. We suggest that early in the dinosaurs’ history, speciation occurred more owing to geographical barriers such as oceans and mountains. This then perhaps shifted towards specialisation within habitats once moving large distances became less possible.

We managed to overcome what, on some difficult days, seemed like an endless number of hurdles, to uncover the geographical story behind the dinosaurs’ varied appearances and diverse ecologies. We have revealed the processes governing how the dinosaurs came to exist where they did, using a realistic model, a phylogenetic tree and the evidence of their deaths 66 million years ago.


Go to the profile of Ciara O'Donovan

Ciara O'Donovan

PhD student , University of Reading

No comments yet.