To behold the 200 million years of the history of a species in one bone. A St Petersburg University palaeontologist speaks about the study of the evolution of crocodiles
Ivan Kuzmin, an early career scientist from St Petersburg University, has been studying modern and fossil crocodiles for ten years already. He is one of the leading authors of scientific articles on this topic in Russia. In his Candidate’s dissertation, he examined the development of the braincase of crocodiles and their earlier ancestors. He found out what evolutionary changes this part of the skull had undergone in different representatives of this group.
Your research team has pioneered the study of the crocodylomorph braincase. Why have these modern creatures been unexplored until now?
This is a good question. The 2021 study, which my dissertation is largely based on, was mainly devoted to modern crocodiles. At present, there are from 25 to 29 crocodile species on the planet. This discrepancy is due to the fact that some researchers define species by the structure of their bones, while others do that by soft tissues and genetic data. In this later case, up to 29 species can be distinguished. Despite the fact that crocodiles live side by side with us and are represented in any zoological museum in the world, different parts of their skulls have been studied inhomogeneously. It is much easier to study the bones of the face or the skull roof, since they are externally exposed. However, it is more difficult to get to the internal sections and structures. Previously, in the course of such studies, skulls had to be destroyed. They were cut in half or into several sections with a circular saw. Naturally, with this approach, the details of the study were scarce, since the three-dimensional structure turned into a set of two-dimensional ones. The emergence of modern methods such as CT scanning and 3D modelling has allowed zoologists and palaeontologists to study structures without destroying them. In this case, the same principle is used: the tomographic scanner virtually (i.e. harmlessly for the collection sample) cuts it into a large number of sections. After that we use special software to look for the boundaries of the bones and, as in a children’s colouring book, trace the structures of our interest. Then, the computer assembles a three-dimensional object from them. You can virtually "take out" the desired bone from the skull, even if it is the only fossil specimen in the world.
In total, working on my dissertation, I studied about 75 specimens of modern crocodiles and 72 specimens of extinct species. There are not many skulls available for study in Russia. Although there are some in the collections of major museums, I had to use open data from non-Russian repositories and other resources to compare them with the available materials. About 19 modern species were studied like that. Some of them were represented by 10 to 20 samples. Some cases were studied using scientific publications. To do that, I had to read even 19th century sources. Indeed, a lot of data has been accumulated by science. Yet, they are all scattered. You can find one fact in an article written back in 1850, another − in the one published in 1920. As a result, when I put everything together, we were able to learn a lot about how the braincase of crocodiles is constructed and create a more visually representation of data compared to previous works. Despite the fact that this part of the skull has been studied for almost two and a half centuries, many questions remained, since it was difficult to get to the internal structures by any methods applied in those days. I also think that perhaps no one was interested in it as much as I did. I have been working on this subject since 2014 and in total devoted nine years to it.
Why are crocodiles so attractive for you?
It is quite a funny story. As a child, I liked dinosaurs for a long time. I remembered it only by my third year at St Petersburg University, when I had to choose my major. Then I asked Pavel Skutschas, my future research supervisor, who in the Department of Vertebrate Zoology was dealing with that subject. I did not even know back then that he was a palaeontologist himself. I wanted to study dromaeosaurs. If you remember the Jurassic Park movie, there were velociraptors that ran on their hind legs in packs and communicated with each other. These two species belong to the same family. Yet Pavel Skutschas told me that it was not all that simple in palaeontology. To study dinosaurs, you need bones, and all Dromaeosauridae samples in Russia were already being studied by other researchers. Fossil rhinos and crocodiles remained the only options available. I chose crocodiles, of course. They are practically close relatives of dinosaurs; they both belong to one large taxon of archosaurs. So I have been dealing with this topic since the period of my bachelor’s thesis, and it is worth mentioning that I still managed to get to the dinosaurs. When I got older and mastered some scientific methods, my colleagues and I published an article in 2020. It was devoted to the dinosaur brain, and we are still working on this subject.
Why the braincase? To be able to work on my bachelor’s thesis, I was provided with materials. It was a rather large volume of isolated bones, which was difficult to work with. Imagine that 100 specimens of four species were taken apart, mixed and laid out in front of you. It was quite a puzzle! To divide the bones into categories, it was necessary to know in detail the anatomy and features characteristic of different groups. I started studying modern crocodiles in order to understand how each bone is arranged, what features it has, and how they differ in different groups of crocodiles.
Then I began to pay attention to the structure of the braincase. The neurocranium of modern alligators seemed different from that of crocodiles, yet the bones that are 90 million years old were similar to it. When I started looking for a confirmation of my observations in the sources, I realised that previous researchers had never noticed that. So I realised that I had found a scientific problem and I would work in that area. In my thesis, I collected into a single analysis all my observations on the differences in the structure of the braincase of crocodiles and crocodylomorphs. I managed to: encode all the characters; conduct a phylogenetic analysis; and get quite a detailed tree of crocodile relationships. I managed to create a representative tree, yet some fragments differ from those accepted in other sources. This suggests that we still do not fully understand the evolution of crocodylomorphs.
If one fragment of the skull can be used to build the entire family tree for the period of 200 million years, it is an important element that carries systematically significant features. Further work with it will enable us to better understand the phylogenetic relationships and introduce new features into the analysis of the group.
Since 2014, after defending my bachelor’s thesis, together with the study of fossil materials, I began to work with modern crocodiles in order to understand their anatomy and use this knowledge to study their ancestors. Comparing them with each other, I found many new features that had not been noted by researchers before. To study fossil specimens, I had to review many publications and visually find those features in the images that were not indicated in the text of the article. And I really managed to discover a lot of the unknown. An article about the braincase anatomy of modern crocodiles that summarised the seven years of my research experience was published in the Journal of Anatomy and hit the cover. This was the basis for my work with fossil crocodylomorphs.
In 2021, you and your colleagues described all the structures in the braincase of crocodiles and assigned a single name to each component. How far have you progressed in your research over the past year?
In fact, this work is completed now. We are currently working on studying the evolution of the brain of archosaurs. The inner surface of the braincase to some extent reflects the structure of soft tissues. During its reconstruction, it is therefore possible to obtain a cast of the brain referred to as an endocast. Previously, to do this, we had to saw the skull through, clean out the cavity and fill it with latex. Now, it can be done virtually using a CT scan and 3D modelling. However, the work is done by hand, and it takes about a week to create one endocast. We are planning to create a sample of over one hundred three-dimensional models, to see how the brain of modern crocodiles changes during their development, from the embryo to the adult. I would also like to detect the differences between the main modern groups. These are: alligators and caimans, crocodiles, and gharials. In the course of this work, we are planning not only to visually assess the endocasts of fossil and modern crocodiles, but also to apply statistical methods, including geometric morphometry. By placing marks on the key places of the endocasts, we will get point clouds that can be compared using the software. Artificial intelligence will perhaps be used to analyse the series of samples.
Caimans, alligators, crocodiles and gharials, as well as their common ancestor who lived in the Late Cretaceous about 100 million years ago, are part of the crown group that unites modern animals and their last common ancestor. All of them are called crocodiles (the Crocodylia group). The animals that preceded them were quite similar in appearance. That is why they are often called ancient fossil crocodiles, but in fact they are more closely related to other fully extinct groups. This group is titled crocodylomorphs (Crocodylomorpha) meaning "similar to crocodiles". They appeared 230 million years ago in the Triassic period. When we speak of crocodylomorphs, we mean a larger group, which includes both modern crocodiles and their extinct "great-grandfathers".
In addition, my students and I created 3D models for previously scanned skulls of fossil crocodiles and crocodylomorphs. We are continuing to compile a collection of materials and hope to receive funding next year to complete this work. Among the unpublished materials, there is a model of the braincase of an alligator that lived in what is now Uzbekistan about 90 million years ago. It is one of the oldest crocodiles in the world. In collaboration with our colleagues from Portugal and Spain, we also published an article devoted to the Portugalosuchus’s skull. I was one of the co-authors. Thanks to my extensive experience with CT scanning, I can find those features in the crocodile braincase which others cannot see. So I was able to help colleagues with 3D reconstruction and comparison of neurocranium bones. By the way, Portugalosuchus is also one of the oldest representatives of crocodiles.
How was the scanning and subsequent modelling of the samples carried out? What were the challenges you faced in the process?
If the sample we are interested in is small and suitable for microtomography, we take it to the Centre for X-Ray Diffraction Studies, where experts carry out the necessary work. We give large specimens to medical workers who scan them using their equipment for human tomography. Palaeontologists do not participate directly in the process of obtaining scans, but we can communicate our wishes to the operators. The microCT has a very high resolution allowing you to get a large number of sections, up to 4,000, with the distance between them being just several microns. Of course, we do not work with such large sets of images. The largest braincase scanning I have worked with was the one from the fossil alligator from Uzbekistan. It consisted of about 2,000 sections. It took me three months to process it. On average, it takes about a week or two to work with one set of tomograms.
The main part of the researcher’s work is segmentation. That is the selection of component parts in the sample. These include individual bones and cavities in which the brain lay, and blood vessels and nerve channels passed through. The Amira/Avizo software is used for segmentation. 100 to 2,000 black-and-white tomograms are uploaded there. The white areas are those with the greatest density, while the black ones are those filled with air. Between them, there is a huge gradation of shades of grey that you need to learn to distinguish. Moreover, in the skull of a modern animal, the brain cavity is cleaned and all the details are easily recognisable, whereas in a fossil specimen, it can be clogged with rock that has almost the same density as bones. The search for boundaries between bones is also not always easy. In a modern specimen, there are clearly distinguishable black boundaries between them. These are sutures filled with air. It can be difficult to notice them in ancient specimens. Skulls that have been lying in the ground for millions of years can be squeezed. In this case, one still need to understand to which direction a bone was displaced and why it lies in a different projection. My experience with modern animals helps a lot here, so I recommend that novice palaeontologists should first practice with modern species.
It turns out that every modern crocodile, to some extent, goes through the 200 million years of the crocodile evolution during the first days of its development, while it is still in the egg.
Working with tomography is rather monotonous yet interesting. At the end of the day, you suddenly notice that you have found a new structure, and a whole series of questions arise related to it. We are often asked why artificial intelligence is not used for this work. In my opinion, working with modern skulls can be easily automated, since the contrast between the colour of bones and cavities is quite noticeable. Recently, neural networks have begun to be used to study the brain of bees. In 2022, Chinese palaeontologists tried to use this technology to identify individual bones in fossil specimens. Yet humans still win here in terms of accuracy.
Have you discovered something new to yourself during the research or come to unexpected conclusions?
Yes, definitely. It was surprising for me to understand that in the process of embryogenesis of a modern crocodile, the bones of its braincase change in a similar way to how they had been transformed during the evolution of the entire group of crocodylomorphs. As far back as the 19th century, it was noted in some publications that something similar happens to other bones, such as the palate, so I was surprised to find out that the same also works for the neurocranium. Now, my students and I are planning to map the changes taking place in the skulls of crocodylomorphs in the course of their evolution and modern crocodiles during their embryogenesis, and compare them with each other. Now, it seems to me that different parts should develop asynchronously, but some stages will resemble the evolutionary transformations that were taking place in remote ancestors.
Please tell us a little about your dissertation defence. Is it difficult to defend a dissertation in accordance with St Petersburg University’s own standards?
In fact, I really liked it, because these standards make life easier. The only difficulty for me was the bilingual system. I wrote my dissertation in English, like my other scientific articles, so I had to translate it into Russian. Usually, it is probably the other way around. The application process and the procedure itself were quite simple. It was new to me and I learned the details from my colleagues who had defended their dissertations before. The dissertation department answered my questions in detail and very promptly, so there were no organisational problems. Everything proceeded rather smoothly, so my experience was a positive one.
The results of your research were used to create the first publicly available online course in palaeontology, and last year you made it to the finals of the "Fidelity to Science" competition. Are you planning to continue educational activities?
Of course. The ability to talk about your research in a simpler language is important for any scientist. If you cannot explain what you do, it means that you yourself do not understand your subject very well. Working with a wide audience to popularise palaeontology and zoology is very interesting and important. For example, when children learn that the animals living around us have been evolving into their modern during millions of years, they realise that global extinction caused by human beings can be a great tragedy. All that inspires an interest in science and a love for nature. Yet the number of palaeontology students is increasing every year, so I am gradually moving to a different format of work related to the education of young personnel who will further promote Russian palaeontology. Now, we are engaging young people in scientific activities. For example, one of the co-authors of an article about the ankylosaur brain published in 2020, was a tenth-grade school student. We started working together as early as his eighth grade and within two years he took a full part in the study. At present, he is a second-year student at St Petersburg University. He continues to study this subject.
There is a high interest in dinosaurs in popular culture, but do many of such fans become true scientists in future? How popular are palaeontologists today?
Indeed, part of the audience is later cut off, since science poses more complex problems than just finding out how many dinosaurs were discovered in the new year and what they are called. Nevertheless, the number of students wishing to study palaeontology is gradually increasing, in no small part due to the dynamic popular science activities of Pavel Skutschas and other members of our team. Our students understand that the work of a palaeontologist is not just looking at bones covered with dust through a magnifying glass. For example, I encourage students to learn 3D modelling. These skills will be useful not only in their scientific activities. They will also enable them to switch to game development or create custom models if they wish. Our team is thinking about creating new online courses in palaeontology, and in this area, modelling is required to create a visual picture for students.
In fact, palaeontology still remains a rather narrow niche. I probably know almost all vertebrate palaeontologists working in Russia. In my opinion, higher competition would better stimulate the development of this science. In addition, this year St Petersburg University reduced the number of postgraduate places in Biology, so master’s programme students graduating in 2023 will have to look for other opportunities for further development in this area. I would like the University to pay attention to this problem. But in general, popular science content is in increasing demand among a wide audience, so the number of palaeontologists in Russia is increasing. I can therefore say that at least our team has experienced evident progress in recent years, and this is encouraging.