Palaeocast

On the origin of my first species: Dinodontosaurus isiyavamanda

Hady George — Fri, 18 Sep 2026 15:42:27 +0000

I can’t imagine how ecstatic 9-year-old me would have felt if I told him he would get to unveil to the world a new fossil species, let alone one that informs on the oldest dinosaurs. 25-year-old me is pretty happy with the work too, but to be honest, it is not thrill or joy that overwhelm me. Instead, my awareness of the limitations of the research and the new questions it begs a fossil nerd to ask leave me with a deepened sense of humility and curiosity. A list of follow-up research projects in my phone’s notes app has been gradually lengthening the more I think about the new knowledge my collaborators and I have uncovered. Now that our knowledge can finally be shared with the wider scientific community in the form of an open-access research paper and publicly accessible data, I hope more eager scientists find the interest to pursue this avenue of research. I also find myself reflecting on the path it took to get here from when I was first invited onto this project, and I think that too may be worth sharing. Here, I hope to walk you through the “behind-the-scenes” of this research in case you may be interested in knowing what can go into identifying a new fossil species or palaeontological research more generically. If you’re an experienced, wise scientist you might find it amusing to read about the silly mistakes the new generation makes while attempting to navigate academia. And if you’re not a scientist at all, please be reassured I’ve left out a lot of the boring technical terms.

The main portion of the skull and the lower jaw of the holotype of Dinodontosaurus isiyavamanda. Photos originally taken by Jonathan Jackson before layered on top of one another by Hady George.

My involvement in the research project started not long after Tetzoocon 2022, which I attended a few months after completing my master’s in palaeontology at the University of Edinburgh. My friend Cassius Morisson, who was a speaker at the convention introduced me to Dean Lomax who was also a speaker. Dean and I had a quick chat over coffee where I mentioned to him my interest in researching dicynodonts, a peculiar group of ancient relatives of mammals that sort of look like pigs but with a beak and tusks. Dean then mentioned he had a friend who had been trying to publish some work he did as a master’s student on a dicynodont, and that he would put me in touch with him. I cannot stress enough how important it is too put yourself out there in your academic circles, no matter how much imposter syndrome may be pulling you down. I still find it surreal I get to meet and chat with friendly and brilliant scientists from all over the world who are regularly reshaping our perspectives on the deep past. Sometimes, you find people who introduce you to your future coauthors. In this instance, Dean introduced me to Nigel Larkin, a world-leading conservator of natural history specimens and preparator of fossils.

One of Nigel’s first ventures into the world of fossils was through his master’s, and like mine, it was on dicynodont fossils. In 1994, Nigel completed his master’s research project (degree awarded by the University of London) on fossils collected from the Triassic of Tanzania in a 1963 expedition. A huge swathe of bones was collected, altogether they represent a 240ish million-year-old graveyard of dicynodonts. The leaders of the expedition, Barry Cox and John Attridge (both have sadly passed away since) brought the collection to the Natural History Museum in London (NHM). Nigel must have spent long hours figuring out exactly what sort of dicynodonts were discovered. Eventually, he came to the conclusion that the best-preserved material represented a new genus and species: Ruhuhuungualasaurus croucheri. The first part of the binomial translates to ‘clawed reptile of Ruhuhu’, the Ruhuhu Basin being the geological unit in Tanzania where the fossils originally came from. The second part honours Ronald Croucher who was once the head of the Palaeontology Laboratory at the NHM. In his thesis, somewhat presciently, Nigel argues Ruhuhuungualasaurus may be related to the South American Dinodontosaurus.

Nigel was delighted to learn I had done my master’s thesis on a dicynodont, albeit one from older, Permian rocks. I had studied the natural mould of a skull of a 252ish million-year-old critter from the North of Scotland belonging to the species Gordonia traquairi. Generously, Nigel offered me the opportunity to take the next step in my research by helping transform his 1994 thesis into a peer-reviewed paper of the 2020s. As soon as I could turn my attention to this project, I was faced with what might forever be one of the most intimidating obstacles of my career: Dinodontosaurus. At first, when I compared the Tanzanian material to that of the South American dicynodont, I could see some clear differences in the anatomy. I came to the decision that the Tanzanian fossils did not belong to Dinodontosaurus, but rather a new genus. Unfortunately, due to the official rules surrounding the naming of species, we could not go with Ruhuhuungualasaurus. Maybe this is for the best though, as seven Us is so excessive that this name has been misspelt every time it’s appeared in the scientific literature. I won’t say what was the name we were going for at the time as a replacement, I’m planning on keeping it available for a future species. I was confident enough with my assessment I decided to present my work at a scientific conference. Our conclusion was presented in a poster I put together for the Symposium of Vertebrate Palaeontology and Comparative Anatomy (SVPCA) in 2023, and this is when Dinodontosaurus first crept up on me.

After my master’s supervisor Christian Kammerer caught a glimpse of the poster, he emailed me about errors I made in interpretation of the anatomy of the skull I was focusing on. Christian is as knowledgeable about the anatomy of dicynodonts and other protomammals as one possibly can be. Beyond simply pointing out some basic errors (e.g. I thought the crack at the tip of the jaw was a real biological feature, and not just the fossil being broken; very silly I know), he also mentioned that a lot of what I thought were unique anatomical traits were actually present in Dinodontosaurus. He sent me almost 800 high-quality photos of Dinodontosaurus fossils found across Argentina and Brazil, and I could see what he meant. My taxonomic assessment, that the Tanzanian material was a new genus, was completely wrong. This whole time I had been looking at Dinodontosaurus, and I hadn’t even realised it. Some of the worst imposter syndrome I’ve ever felt was at this moment. If I can’t even get this assessment right, what hope did I have at getting the tricker details right? Surely, I wasn’t capable of writing an academic paper on these fossils. Keep in mind that as much help as Nigel and my other coauthors were throughout this project, none of us were senior professors with a longstanding expertise in the ancient relatives of mammals (Mike Day wasn’t a coauthor at this point). After spending a day or two collecting myself, I returned with renewed enthusiasm. I refused to give up, and I’m glad I made this choice. We are discovering the secrets of past alien-like animals after all.

The poster where I got the taxonomy all wrong. This was presented at SVPCA 2023 and should NOT be used as a reference.

I again went through all the anatomical intricacies that I could find in the skull I was studying. I took more care with making comparisons to other species and carefully consulted the hundreds of images Christian had sent me. This was during the first year of my PhD at the University of Bristol (as of writing this, I am in my final year). In the Life Sciences Building, first year students are typically kept in a shared basement office. I spent long hours in there ruminating over how sloped some parts of some skulls seemed to be and other tedious bits of anatomy. It certainly was not fun and came at the cost of missing out on social events and exercise, but I did manage to come to new conclusions. The Tanzanian material seemed to be Dinodontosaurus. It had all the basic features of the genus including tusks, a swollen boss above the beak, and the back of the head was not notably exaggerated in any way. In many ways, Dinodontosaurus is generic (pun not intended). Its gross skull anatomy is like the default version of Triassic dicynodont skulls. This is where Dinodontosaurus gave me yet another headache. ‘Being basic’ is not a substantiative argument that survives peer-review. This is not to say Dinodontosaurus doesn’t have unique, derived anatomical features. It absolutely does, but they are characteristics of the sutures (rigid joints between skull bones) that I cannot identify in the Tanzanian skull because the surface quality of the fossil has been damaged by erosion. You can see why this was so troublesome.

The roof of the mouth (technically called the palate) of the Tanzanian skull I was studying had a clearly unique feature. What is normally a super thin cutting ridge was instead a bulbous structure. This was originally the main feature I had based my previous assessment on. I could not identify anything like this in the South American Dinodontosaurus material, but one or two Brazilian specimens Christian sent me photos of seemed to have something like it. I came to the conclusion that we must instead identify the Tanzanian fossil as Dinodontosaurus cf. tener. For the uninitiated (lucky you), this basically means: ‘it looks like Dinodontosaurus tener (the main Brazilian species), but it might not be Dinodontosaurus tener, and I don’t really have a clue. It’s an honest, but somewhat disappointing conclusion to come to. Here, I am beyond glad my coauthor Juan Escobar came to save us.

The unique palate anatomy of Dinodontosaurus isiyavamanda. Note how the ‘pr’ (posterior ridge) is bulbous rather than ridge like. Photo on left taken by Jonathan Jackson.

Juan has recently completed his PhD on the Argentinian representative of Dinodontosaurus (D. brevirostris), and few can rival his knowledge of the accursed beast. Juan suggested we shouldn’t be drawing comparisons to every single Dinodontosaurus specimen I had seen in Christian’s photos. Many were unpublished and had yet to be formally assigned to a specific species. Juan gave me a list of which ones to exclude from our comparisons, and I adjusted the taxonomic assessment accordingly. None of the official specimens of Dinodontosaurus tener had the unique palate of the Tanzanian skull. We also now know that Dinodontosaurus brevirostris isn’t a perfect match for it thanks to Juan’s 2023 paper on its lower jaw anatomy. Those are the only two established species of Dinodontosaurus and both are from South America. The former is from Brazil, and the latter is mainly from Argentina but is now also known from Brazil. This means our Tanzanian skull was not only the first record of Dinodontosaurus outside of South America, but also a new species. We can look back on this and see this as Juan correcting my mistake of drawing comparisons to unpublished specimens, but there are so many Dinodontosaurus specimens it is difficult to keep track of them. Palaeontologists are used to having one or two specimens to compare to for each species, or at least one good reference specimen (e.g. holotype, paratype, lectotype), but the convoluted history of study of Dinodontosaurus has left us with a bit of a mess. I’m not the first and won’t be the last to get headaches from this animal 240ish million years after its extinction.

That was the hardest part done, and now that we had Dinodontosaurus in Tanzania, it got me thinking about how old the rocks it’s from must be. The fossils we studied are from the same geological unit that yields the silesaur Asilisaurus and potentially also Nyasasaurus. Both of these reptiles have been argued to be the oldest dinosaurs as the unit has historically thought to be Anisian in age (the middle of the Middle Triassic). But the South American Dinodontosaurus fossils are not that old, and we know that from radiometrically dated rocks in Argentina and Brazil. The age of the Tanzanian unit was based on outdated links to units in South Africa. After compiling a very long list of fossil species in all these mentioned countries, it was evident that the Tanzanian unit can’t be that old. If we’re deciding its age based on which other unit it has the most in common with, since we can’t radiometrically date the rocks because we don’t have the right minerals, then the Tanzanian unit should be considered the same age as the South American rocks. Although on different continents (which were connected back in the Triassic) the faunas are incredibly similar. Previous researchers had suggested this based solely on the cynodonts (another group of protomammals) and in discussion of other units like that in Zambia (and also very recently based on poposauroid material, but I totally missed this paper when it was published; very silly I know), but now with Dinodontosaurus in play, the evidence is overwhelmingly in our favour. The Tanzanian unit, and its alleged oldest dinosaurs, are probable no older than end Ladinian – Carnian or late Middle Triassic to beginning Late Triassic in simpler terms. We don’t know exactly how many years we’ve shortened its age, but it’s something like 6-8, maybe 10 million years younger now.

Coming to that realisation, and convincing coauthors and later reviewers of it, has been extremely satisfying. ‘Recalibrating the origin of dinosaurs’ if you want to put it so dramatically, never got any real pushback. After I presented this at the Society of Vertebrate Palaeontology meeting of 2023, I asked curator of mammals at the Field Museum and stem-mammal expert Kenneth Angielczyk what he thought of the whole thing. His response was simply: “it’s convincing”. Ken is one of the greats we proverbially stand on the shoulders of, so that immediately became a career highlight. I may have screwed up the taxonomy repeatedly, embarrassing myself time after time, but this “bigger picture” conclusion I got right from the beginning. When I need to, I look back at this to remind myself my failures don’t define me, my successes do too, and so do the journeys I took to arrive at both.

The new and traditional equivalences between Gondwanan geological units in Argentina, Brazil, Tanzania, Namibia, and Zambia. It is clear Dinodontosaurus is only one of many connections, but it is arguably one of the most important ones considering its abundance in South America. Schematic made by Science Graphic Design.

In December of 2024, a few weeks before going on Winter vacation, I submitted the work we’d done as a manuscript to Journal of Vertebrate Palaeontology. At this stage we also bolstered our taxonomic assessment with the findings me, Juan, and my friend Edmund Moody gathered from phylogenetic analyses. Juan and I handled one mathematical approach to recovering evolutionary relationships and Ed handled another two. Ed is renowned for his work on producing genealogies from genetic data, and is well known for his recent work on the reconstruction of the genome of the ancestor of all life on Earth. Dicynodont evolution is small fry to him. Together with Nigel, Mike Day who is a curator at the NHM and has played a huge role in kickstarting my career by hosting me time and again to study fossils, and Charles Saanane who is our colleague in Tanzania that helped with deriving the name of the new species, we wrote the paper.

The review process was brutal and super lengthy (submitted Dec 2024, published Sept 2026). I won’t get into all the details, but I will say it would have been much easier if we had a professor or some sort of senior academic to supervise. My coauthors were incredibly helpful, especially Mike as he’s an experienced describer of therapsids, but writing scientific papers well is an extremely difficult task to master. The reviewers and editors had a lot to say, much more about the writing than the science. There were still important comments on the science though. Christian and Ken reviewed the paper after all, and they can be incisive. I mean this in the best way. The paper was greatly improved thanks to their constructive criticism. At first Ken gave major revisions (he said ‘moderate’, but they felt ‘major’), and in the second round of review Christian gave major revisions. It was not easy, but I am grateful. I’m especially thankful for Ken pointing out that the referred specimen he collected does indeed have one of the diagnostic sutures of Dinodontosaurus, alleviating me of some of the mental pain I was suffering from. I could now say more than just a lot of technical anatomical words that basically meant ‘it looks basic’. Also, I did all this while juggling my PhD on jaw mechanics in fish and amphibians. Time management was everything, and I have sacrificed loads of my free time to get here. It was definitely worth it, but I can’t see myself undertaking such a side-project again (not to anyone’s surprise, writing a 100-page manuscript is not as fun as going out with your friends).

Once all the science was done, we had to decide on a name for this new Dinodontosaurus species. Originally, I floated the idea of naming it after Christian, but he wasn’t the biggest fan of the idea when I mentioned it to him. I totally sympathise; it is Dinodontosaurus at the end of the day. Afterwards, we thought naming it after Barry Cox would be nice. He did co-lead the expedition that found the fossil, and he had studied the Brazilian Dinodontosaurus fossils long ago. It felt appropriate to honour him for his contributions to ‘dicynodontology’ as I hoped to put it. However, at the Society of Vertebrate Palaeontology conference in 2025, Emma Dunne gave a presentation on naming fossil species in the languages of the native peoples of regions fossils are from, and we chatted about this afterward. I thought it would make more sense to instead commemorate the locals who helped with the collection of the Tanzanian skull in the 60s. Their names were never recorded so we cannot honour them individually, but we can honour them as a people. ‘Isi ya vamanda’ are the Manda words for ‘country of the Wamanda’, who are the local people of Southwest Tanzania where the fossils are from. To wrap up, that’s how we arrived at the new name. Actually registering an official name for a species of animal is super easy. It was by far the easiest part in this whole four-year ordeal. You just fill in some boxes in a website. This side-quest has both raised and humbled me, and I stand before you as someone who is half-expecting a future scientist to refute my findings and synonymise/invalidate the novel species once new data is collected, as is often the case in science, especially when this science concerns dicynodonts. Whether this happens next year or next century is a mystery, so for now, please welcome Dinodontosaurus isiyavamanda.

Hady George is a palaeontology PhD student at the University of Bristol researching jaw function across the fish-tetrapod transition among other things like dicynodonts, and seemingly always has a pop science book somewhere.

Cover art by Gabriel Ugueto

Full scientific paper here: https://tandfonline.com/doi/full/10.1080/02724634.2026.2692542

Episode 177: Cambrian Fecal Revolution

David Marshall — Mon, 17 Aug 2026 16:42:00 +0000

The Cambrian Explosion is one of the most significant events in evolutionary history. Classically, it describes the sudden appearance of complex fossils (including all major animal groups) at the beginning of the Cambrian Period. Before then it was thought that only relatively simple single-celled or multicellular life existed. We now understand this event to be a lot more nuanced, recognising many of the geological biases at play and having discovered complex fossils in the earlier Ediacaran Period. Many researchers therefore consider the term ‘Cambrian Radiation’ to be a better representation of the event.

Whether an explosion or a radiation, determining the events of the Ediacaran and Early Cambrian and their causes is a constant area of palaeontological research. Numerous drivers behind this event have been hypothesised ranging from oxygen availability and nutrient cycling to an arms race between predators and prey. It is likely that no single factor will be identified as an underlying cause, but the relative extent to which each contributed to the event will be debated for many years to come.

In this interview, we are joined by Dr Julien Kimmig, Staatliches Museum für Naturkunde Karlsruhe, Germany, who recently co-authored a study on the coprolites of the Early Cambrian, the evolution of guts, and the influence the ‘Cambrain Fecal Revolution’ may have had on the wider Cambrian Radiation.

Within the modern marine carbon cycle, the processes driven by biological activity are often referred to as the ‘biological pump’. These biological processes are incredibly important for how carbon is cycled and is the foundation of the whole marine food web. Within this cycle, particulate organic carbon (POC) represents all small organic remains such as detritus and faecal material. Whilst this POC represents a modest 2.3 gigatonnes of material at any time, around 50 gigatonnes can be cycled each year, meaning POC has the highest turnover rate of any organic carbon pool on the planet. In some environments, faecal matter can account for 100% of all POC, though the average amount is typically below 40%. Regardless, faecal matter still represents a significant proportion of the biological pump and understanding how and when animals first started to produce it is a key part in understanding the evolution of global carbon cycles through time.

It is reasonable to assume that the origin of faecal matter would coincide with the first appearance of metazoans (animals) in the fossil record. We are now fairly confident that many of the organisms from the Ediacaran biota, such as Kimberella (pictured), are animals. In fact, we even have evidence that some specimens preserve traces of gut content. Despite this, no coprolites (fossil excrement) are known before the Cambrian Period. Given that the identity of many Ediacaran organisms remains problematic, it can be difficult to ascertain whether these (potential) animals had the requisite anatomy to produce faecal matter and in a way that can be fossilised and recognised at a coprolite. Image credit: Oleg Kuznetsov CC BY-SA 4.0.

The most convincing examples of Ediacaran animals with a gut anatomies potentially capable of producing faecal matter are the cloudinomorphs. These animals produce a funnel-in-funnel external shell and possess a through-gut stretching from end to end. If true, they would be closer related to us than to other basal groups such as jellyfish or sponges. The image above demonstrates the position of the gut (dark red) and the fossilisation process. The video below shows how the specimens were micro-CT scanned and how a 3D reconstruction was built up from interpretations of hundreds of 2D X-ray slices. Credits: Schiffbauer et al. 2020. Discovery of bilaterian-type through-guts in cloudinomorphs from the terminal Ediacaran Period. Nat Commun 11, 205. CC BY 4.0.

The image above shows how many coprolite-producing sites are known from each series/stage of the Cambrian Period and their locations. Whilst coprolites are found as soon as the Cambrian begins, the number of fossil sites remain relatively low. By examining the entirety of the Cambrian coprolite record in this way, Kimmig and Bicknell were able to characterise the evolution of guts and faecal matter through time.

In the first series of the Cambrian Period, the Terreneuvian, possesses very few coprolite-yielding sites and the relative size and diversity of coprolite forms remains low. This likely represents the lack of larger animals with more advanced digestive tracts at this time.

Small carbonaceous and phosphatic micro-coprolites, like those above, are the only forms encountered in the Terreneuvian. The carbonaceous forms are often flattened strings of subcircular faecal pellets that are below 2mm in length. They are preserved in a variety of shapes including straight and coiled forms. Since they don’t contain the remains of other animals, it is assumed they were likely produced by deposit-feeding organisms grazing the sea floor sediments.

Phosphatic microcoprolites include three-dimensional pellets which can occur in a variety of forms including individual pellets, trails and burrows. Similar to the carbonaceous microcoprolites, these were likely formed by deposit feeders.

By Series 2 of the Cambrian, many more coprolite-yielding sites are known and there is an increase in the size, complexity and abundance of coprolites. These sites include the first konservat-lagerstätten (sites of special preservation of fossils) that preserve the first large and relatively complex animals. Because of these sites, we are able to observe that modern trophic dynamics (feeding strategies) are more or less established by Series 2 and we are better able to identifying potential producers for each coprolite type. However, the effort put into collecting and studying material from these lagerstätten could introduce a human bias given the relatively limited attention offered to the earlier Terreneuvian sites.

There are a variety of different macroscopic coprolite groups appearing in Series 2. These can often contain the remains of other animals, thus giving us direct evidence of some of the earliest metazoan (animal) food webs. This example from the Emu Bay Shale contains the remains of trilobites.

A possible producer of this coprolite could be the relatively large trilobite Relichia rex. This trilobite is known to have strong, robust legs with large spines at their bases that would have been used to crush other animals it encountered on the sea floor. Whilst other large predators, such as Anomalocaris are known from Emu Bay, their nektonic (free swimming) lifestyle means that their coprolites would likely contain other nektonic prey such as bivalve arthropods. The coprolites of these nektonic predators are much rarer. Whilst it’s a now lot easier to reason who the producer of any given coprolite might be, there is still little direct evidence. Fortunately, by this time, we begin to find animals with their gut contents preserved in situ (i.e. still inside the gut), these are termed ‘cololites’ and together with coprolites and any other fossil produced from the digestive system of an animal are called ‘Bromalites’.

Phosphatic microcoprolites in Series 2 also include those formed of shelly material. These will often contain the remains of one type of organism such as worm sclerites (armour), bivalve arthropod exoskeletons and brachiopod shells. It is thought that these coprolites were probably produced by smaller predators or scavengers such as arthropods or worms.

By the Miaolingian, marine shelf communities were well established and were producing a lot of faecal matter.

We have some excellent evidence of trophic systems from this time, particularly from the priapulid worm Ottoia prolifica. This worm lived in the sediment, is abundant in many Cambrian lagerstatten (such as the Burgess Shale) and often has its gut preserved. There are therefore numerous examples of cololites preserved revealing the kinds of material that Ottoia was eating. In the image above, there are clear examples of hyoliths (cone shaped shells)(A-H), brachiopod shells (I-K), agnostids (L, M) and trilobites (N-P). Image credit: Vannier, J. 2012. Gut Contents as Direct Indicators for Trophic Relationships in the Cambrian Marine Ecosystem. PLoS ONE, 7, 12. CC BY 4.0.

With so many specimens and so much direct evidence, the diet of animals like Ottoia can be fully defined and even compared between different sites. The table above shows the number of specimens with specific gut contents from the Raymond Quarry (RQ+RT) and Walcott Quarry (WQ+WT) showing proportional differences in their diets. Image credit: Vannier, J. 2012. Gut Contents as Direct Indicators for Trophic Relationships in the Cambrian Marine Ecosystem. PLoS ONE, 7, 12. CC BY 4.0.

Looking at the data as a whole, it’s clear that there is a link between the appearance of different animal groups and the appearance of coprolites. Given that an animal can produce faecal material multiple times a day, but only ever one corpse, it’s unsurprising that the appearance of coprolites can marginally predate the appearance of the potential producers in the fossil record. The evolution of coprolites can therefore serve as a proxy for the evolution of guts and entire trophic systems through the Cambrian.

So we now have a better grasp of evolution of faeces through the early Cambrian, what’s the significance? This is all to do with nutrient availability and cycling. Particulate organic carbon (POC) transfers carbon and other nutrients from the surface to deeper waters. This provides resources for epifaunal and infaunal animals (i.e. those living on and in the sea floor sediments respectively). These, in turn, bioturbate (mix up) the sediments, introducing oxygenated waters and creating new ecological niches. This movement of POC also is a driver for the biological pump. By understanding the evolution of faecal material in the early Cambrian, we better understand the evolution of the biological pump and the marine carbon cycle. This would have had global significance and so is an important factor when considering the cause of the Cambrian Radiation.

Dr Julien Kimmig, Staatliches Museum für Naturkunde Karlsruhe.

Episode 175/176: Insect Gigantism

David Marshall — Sat, 02 May 2026 21:20:56 +0000

The Carboniferous period is host to some of the largest arthropods to have ever lived. Giant taxa such as the griffenfly Meganuera and the millipede Arthropleura are almost talismanic and are often depicted in reconstructions of the period. Since many other groups also have giant representatives in the Carboniferous, what is it about this time that allows for arthropods to grow to such large sizes?

Arthropods breathe very differently to how we do with many using a series of branching hollow tubes called trachea for gas exchange throughout the body. This tracheal system uses diffusion and advection to exchange oxygen and carbon dioxide from areas of higher concentration to lower concentrations. In 1995, a study in the journal Nature suggested that elevated oxygen concentrations in the Carboniferous (approximately 30%, as opposed to 21% today) allowed for gigantism in arthropods since oxygen could diffuse deeper/further into their larger bodies.

A recently published study in the same journal is now casting doubt on that interpretation and in this interview, we are joined by one of the authors, insect physiologist Prof. Jon Harrison from Arizona State University. He introduces us to the tracheal system and its link to the size of insects in the Carboniferous.

For some, the Carboniferous is synonymous with large arthropods including the millipede Arthropleura (pictured) and giant “dragonflies” Meganeura (see below). Image courtesy and copyright of Netflix.

MNHN R51142. Lectotype of Meganeura monyi from France held at the Museum national d’Histoire naturelle, Paris. Meganeura had a wingspan of around 70cm. Despite this, they are not the largest insects of all time, with that honour going to the closely-related Meganeuropsis permiana of the early Permian of the USA. Image CC BY 4.0.

Reconstruction of M. monyi by Bob Nicholls.

Instead of lungs, like we have, Insects possess a tracheal system which is a series of branching hollow tubes throughout the body. There are typically eight pairs of openings to this respiratory system (called spiracles) visible on the outer surface of an insect. These can be seen as narrow black ovals on this caterpillar. The tracheal system is key to our understanding of insect gigantism in the Carboniferous.

Details of the tracheal system with the spiracles (Sp) leading to the tracheae (T) which then branch into ever smaller tubes called tracheoles. Each cell of the body is able to exchange gasses with the atmosphere directly through neighbouring tracheoles. This system might initially appear fairly rudimentary but there is a lot of plasticity built into it and the use of air sacs (As), for example, allows for the movement of air to be controlled around the body.

Larvae of the fruit fly Drosophila possess just two functional spiracles (orange), each connected to a single large longitudinal trachea (tube shape). These tracheae are capable of delivering all of the insects oxygen needs.

Where the tracheal system is incapable of delivering sufficient oxygen, new growth and branching can be triggered in response to hypoxia (low oxygen) signals released by cells. Here, a green fluorescent protein has been bound to cells suffering from hypoxia. Such experiments have revealed the flexibility of the tracheal system to address the specific respiratory demands of the body.

MicroCT image of a scarab beetle showing air sacs (white) throughout the body and flight muscles (centre, grey) penetrated by tracheae (white hair-like lines) whilst large eggs (solid grey) fill the abdomen. Air sacs are closely associated with flying species but can be used for numerous purposes such as bouyancy control, facilitating large body/appendage sizes, and helping to control ventilation throughout the respiratory system.

Looking at the atmospheric concentration of oxygen throughout the Phanerozoic, it is clear that the Carboniferous and Permian had elevated levels. The leading hypothesis since the 90s has been that this higher concentration allowed for bigger sizes as the oxygen could diffuse deeper into the body of the arthropod. Image from Mills et. al. 2023 CC BY 4.0.

CT scan of the flight muscle of a beetle. Here you can see the tracheae (hair-like shapes) penetrating into flight muscle bundles (negative space). The latest study on this issue reveals that as body size increases, there isn’t a very strong correlation with the tracheolar volume density; we don’t see a major increase in the fraction of the flight muscle occupied by tracheoles. Additionally, no modern insect can be observed near the maximum limit where muscle function begins to be impaired by the presence of the high volume of tracheoles. Therefore if the biggest insects today don’t need to compensate for oxygen diffusion, it can’t be the main controlling factor on body size.

And that’s probably a good thing because large arthropods would wreak havoc (as graphically depicted above).

Prof. Jon Harrison, co-author of Snelling, E.P., Lensink, A.V., Clusella-Trullas, S. et al. Oxygen supply through the tracheolar–muscle system does not constrain insect gigantism. Nature (2026).

Episode 174: A History of Dinosaurs in 50 Fossils

Sophie Pollard — Wed, 15 Apr 2026 19:58:59 +0000

Prof. Paul Barrett of the Natural History Museum, London, recently authored A History of Dinosaurs in 50 Fossils. We took this as an opportunity to get an overview of what we really know about dinosaurs and how it’s even possible to tell their story with just 50 specimens.

In this episode, Paul discusses the history of dinosaur research, the current state of the science and what are still some of the big unknowns.

This imposing lower jaw is one of the original finds used by William Buckland when he named Megalosaurus in 1824.
Credit: ©The Trustees of the Natural History Museum, London

As its remains are so abundant, Allosaurus is often used as a reference for interpreting the anatomy of other less complete theropods.
Credit: ©The Trustees of the Natural History Museum, London

This is the skull of an adult Pachycephalosaurus, with a fully-formed skull dome and knobbly spikes on its snout and skull margins. In younger individuals the dome was much thinner and the spikes were longer and finer. These differences are so striking that some paleontologists originally regarded the youngsters as a different species.
Credit: The Trustees of the Natural History Museum, London

It used to be thought that the plates in Stegosaurus were paired, but complete specimens like ‘Sophie’ show that they alternated from side-to-side along the back. Plates on the neck and at the end of the tail were relatively small while the largest plates were situated over the hips.
Credit: ©The Trustees of the Natural History Museum, London

Dinosaurs dominated land ecosystems for nearly 150 million years. This scene shows one of the most famous examples, from the Upper Jurassic Morrison Formation of the USA, with Stegosaurus (left), Diplodocus (centre) and Allosaurus (right).
Credit: ©Robert Nicholls / Trustees of the Natural History Museum, London

Although the interior of this titanosaur egg is now filled with brightly coloured agate, it would have originally contained a yolk, egg white and a tiny dinosaur embryo. When this specimen was acquired in the 19th century, it was probably the first dinosaur egg in any scientific collection, but it spent its time at the museum in the mineralogy collection unrecognised for what it really was.
Credit: ©The Trustees of the Natural History Museum, London

One of the most iconic dinosaurs, Triceratops is often shown in combat with Tyrannosaurus. Some fossils provide direct evidence of this behaviour with Tyrannosaurus tooth marks on Triceratops bones.
Credit: ©The Trustees of the Natural History Museum, London

Episode 173: Petrified Forest

Emily Keeble — Tue, 17 Mar 2026 20:51:42 +0000

Petrified Forest National Park in northeastern Arizona, USA is a hub for Triassic palaeontology and has exposures representing 20 million years of the Late Triassic Chinle Formation. Visitors marvel at the colourful fossilised trees from which the park takes its name, but a whole host of animals called these swampy forests home 225 million years ago.

In this episode, we talk to Dr. Adam Marsh, lead palaeontologist at the National Park. We explore the history, geology, and palaeontology of Petrified Forest, along with exciting research centred around specimens from the park. Research is ongoing, with many groups of palaeontologists working on Petrified Forest specimens, and we hear about directions it might go in the future.

Petrified trees and wood are extremely common in the park. Silica from volcanic ash has replaced the wood with quartz. The colours come from the recrystallisation of that quartz and the introduction of different minerals. Image: E. Keeble

The park contains badlands that are bursting with fossils. The sediments in Petrified Forest National Park are full of bentonite, which expands when wet and contracts when dry, pushing fossils to the surface. Image: NPS/A. Marsh

Interviewee Adam Marsh holds a jacket after collection in the park. Image: NPS/A. Marsh

A typical dig in Petrified Forest. Image: NPS/A. Marsh

A quarry – you can see discarded rocks that have been dug out trailing down the hillside. Image: NPS/Adam Marsh

The Puerco River, which gives its name to Puercosuchus (below), rarely flows as a river throughout Petrified Forest and spends most of the year a dry riverbed. Image: NPS/A. Marsh

One of the animals known from Petrified Forest, Puercosuchus, an azendohsaurid allokotosaur, described in Marsh et al. 2022. Image: Megan Sodano

Adam at the top of a ladder in search of rauisuchid skull bones. Sometimes fossils are found in far from ideal places. Skull pieces were initially found at the base of the cliff and followed upwards. Image: NPS/A. Marsh

Jackets containing a phytosaur skull found at the park. Phytosaurs were large crocodile-like reptiles, the teeth of which are very common in the Chinle Formation. Image: NPS/A. Marsh

At the other end of the size scale for fossils, the lower jaw of the caecilian Funcusvermis, described in Kligman et al. 2023. The park has many prolific microvertebrate sites where very small fossils are common. We previously interviewed lead author, Ben Kligman about this discovery in episodes 151/152. Image: NPS/A. Marsh

Adam Marsh sits at a quarry in the park. Image: NPS/A. Marsh

The preservation of fossils in the Chinle Formation can be excellent, as seen here in this Revueltosaurus specimen prepared onsite. Revueltosaurus was a relatively small, heavily armoured pseudosuchian related to aetosaurs. Image: NPS/A. Marsh

Kligman, B.T., Gee, B.M., Marsh, A.D., Nesbitt, S.J., Smith, M.E., Parker, W.G., Stocker, M.R. 2023. Triassic stem caecilian supports dissorophoid origin of living amphibians. Nature 614(7946):102-107. doi: 10.1038/s41586-022-05646-5.

Marsh, A.D., Parker, W.G., Nesbitt, S.J., Kligman, B.T., Stocker, M.R. 2022 Puercosuchus traverorum n. gen. n. sp.: a new malerisaurine azendohsaurid (Archosauromorpha: Allokotosauria) from two monodominant bonebeds in the Chinle Formation (Upper Triassic, Norian) of Arizona. Journal of Paleontology 96(90):1-39. doi:10.1017/jpa.2022.49

Episode 172: Rhynchocephalians

Emily Keeble — Thu, 05 Feb 2026 13:32:20 +0000

Today, there is only one living species of rhynchocephalian: the tuatara of Aotearoa/New Zealand. Despite today’s paucity of species, this was once a diverse group of reptiles, with a wide range of lifestyles from swimming in the ocean to climbing trees. Once highly abundant around the world, reasons for their decline are still debated and may have had to do with competition from their relatives, the squamates, or changing environments.

Rhynchocephalians are related to lizards and snakes within Lepidosauria, but despite their outward appearance, are not lizards themselves and have a number of differences that make them distinct. In this interview, we speak to Dr. Victor Beccari, an expert in rhynchocephalians, and discuss this and more about this fascinating group of reptiles.

The tuatara (Sphenodon punctatus) is the only living rhynchocephalian. It is endemic to just 32 islands off of North Island, New Zealand, having gone extinct on the mainland. This was likely due to the introduction of new predators such as rats. They are a protected species classified under the New Zealand Threat Classification System (NZTCS) as ‘relict’ and ‘at risk’. This recognises that whilst their population is currently stable, their current range is less than 10% of what it historically was. This makes them more susceptible to the impacts of climate change or invasive species.

Rhynchocephalians from the Solnhofen Archipelago, showcasing their size and morphological diversity. A) Sphenodraco scandentis; B, Homoeosaurus maximiliani; C, Kallimodon pulchellus; D, Oenosaurus aff. muehlheimensis; E, Pleurosaurus ginsburgi. (From Beccari et al. 2025a; Fig. 12).

Fossils from the Solnhofen Archipelago in modern day Germany are often beautifully preserved and reveal this past diversity, but are flattened, making some aspects of studying morphology challenging.

Victor Beccari, our interviewee, at the Palaeontological Museum of Munich, looking at the holotype of Kallimodon pulchellus, a well preserved Solnhofen rhynchocephalian figured above.

3D models of Pleurosaurus skulls. From top to bottom, Pleurosaurus ginsburgi, Pleurosaurus cf. P. ginsburgi, and Pleurosaurus goldfussi from Beccari et al. 2025b.

Pleurosaurus was an unusual genus of marine rhynchocephalian with an elongate body and tail as adaptations for swimming. CT scans allow for the reconstruction of skulls from disarticulated remains, giving us a more complete picture of what the animal would have looked like in life. Some bones have been taphonomically distorted (squashed during the fossilisation process) and so may have looked slightly different. This image shows two species of Pleurosaurus, with a partial juvenile skull in the centre that tentatively belonged to P. ginsburgi. It had an unworn dentition, so its teeth had not been used much, and unfused neural arch pedicles in the presacral vertebrae, adding weight to its being a young juvenile and not a small adult.

Ecomorphological analysis of Jurassic rhynchocephalians, compared to extant limbed lizards. The shaded areas show the ecomorphospaces of lizards, whereas the points represent rhynchocephalians (from Beccari et al., 2025a; Fig. 14).

The different colours of the shapes in the ecomorphospace refer to the life habit of the lizards, whether they are arboreal (climbing), saxicolous (living amongst rocks), terrestrial (living on the ground), semi-arboreal, or semi-aquatic (living some of their life in the water). These are compared to extinct rhynchocephalians and Sphenodon (tuatara) to hypothesise how each species was living. It is assumed that if an animal falls within the space of the arboreal lizards, for example, that it was likely arboreal, although there is considerable overlap with the hulls of each ecology. Sphenodraco, the new genus described in Beccari 2025a falls within the arboreal ecomorphospace every time.

Reconstruction of the first arboreal rhynchocephalian, Sphenodraco scandentis, in the foreground, with the pterosaur Ramphorhynchus in the background by Gabriel Ugueto (Beccari et al., 2025a; Fig. 16).

Signs that Sphenodraco was arboreal include its elongated limbs, recurved claws, and limb and manus (hand) proportions also seen in exclusively arboreal lizards.

References:

Beccari, V., Guillaume, A. R. D., Jones, M. E. H., Villa, A., Cooper, N., Regnault, S., & Rauhut, O. W. M. (2025a). An arboreal rhynchocephalian (Lepidosauria: Rhynchocephalia) from the Late Jurassic of Germany, and the importance of the appendicular skeleton for ecomorphology in lepidosaurs. Zoological Journal of the Linnean Society, 204(3), 39. https://doi.org/10.1093/zoolinnean/zlaf073

Beccari, V., Villa, A., Jones, M. E. H., Ferreira, G. S., Glaw, F., & Rauhut, O. W. M. (2025b). A juvenile pleurosaurid (Lepidosauria: Rhynchocephalia) from the Tithonian of the Mörnsheim Formation, Germany. The Anatomical Record, 308(3), 844–867. https://doi.org/10.1002/ar.25545

Episode 171: Freshwater Mosasaurs

David Marshall — Fri, 12 Dec 2025 02:41:32 +0000

We’ve been given exclusive access to a brand new study examining the chemistry of a mosasaur tooth found within the Late Cretaceous Hell Creek Formation, North Dakota. The remarkable circumstances of how this tooth was discovered meant that multiple lines of chemical evidence could be reliably gathered, each acting as a powerful palaeoenvironmental proxy providing clues as to how and where this giant aquatic predator lived. The results of the study now mean that a revision of mosasaur palaeoecology is required and that food webs in one of the world’s most famous fossil deposits might need redrawing.

Joining us in this episode are the lead authors Nathan Van Vranken (Eastern West Virginia Community and Technical College) and Melanie During (Uppsala University & Vrije Universiteit Amsterdam). Their study “King of the Riverside”, a multi-proxy approach offers a new perspective on mosasaurs before their extinction is open access and in BMC Zoology available now!

Map of the exposures of the Late Cretaceous (Maastrichtian) Hell Creek Formation in North Dakota. The blue pin represent the site (NDGS L4327) in Morton County where the mosasaur tooth from this study was discovered.

This site is interpreted to have been the flood plain of a river and had only ever produced terrestrial or freshwater fossils including dinosaurs, turtles and crocodilians. Image: Volunteers digging with the North Dakota Geological Survey. (Why not volunteer yourself?)

The Tyrannosaurus rex tooth (NDGS 15125) collected by Trissa Ford. This fossil was is pretty poor shape for a tooth and so great care was taken to collect it safely, with the whole football-sized block of rock being wrapped in plaster and destined for the prep labs.

As the T. rex tooth was being excavated, a mosasaur tooth (NDGS 12217) was fortuitously discovered. Had the former tooth not first been found and collected in such a way, might the latter tooth have been missed entirely?

Details of the mosasaur tooth, identified as belonging to a prognathodontine, and its approximate position relative to the T. rex tooth. But what is a mosasaur tooth even doing in this freshwater deposit?

Not only that, but when the T. rex tooth finally went in for work in the prep lab, another fossil was discovered: the jaw of a crocodilian (NDGS 18199). Incredibly, the jaw still had teeth embedded. So that’s three teeth from the same rock.

If that wasn’t enough, hadrosaur remains are abundant at the site, so there was no difficulty in getting hold of one of their teeth. That’s three teeth in immediate association and one that’s very close. Having the same kind of structure from different organisms all found in the same location is a geochemist’s dream!

Melanie was able to mine material from each fossil and examine the ratios of isotopes of different elements within. Each was able to act as a different kind of palaeoenvironmental proxy, allowing her to determine the conditions in which each fossil lived.

The ratio of strontium isotopes in each fossil can show the influence of freshwater. The ratios of strontium isotopes in marine environments is very predictable, but in freshwater, the signal is a lot more varied. Here, the fossils from the Morton County site (blue) -including the mosasaur- are all well removed (bottom of the graph) from the narrow range expected of marine fossils (top of graph). Even the fossil of Mosasaurus dekayi from an older site, filled with marine fossils, shows a likely freshwater influence. This could be evidence for a halocline where the old Western Interior Seaway (WIS), having lost its connection to the open ocean, now has a layer of freshwater on top. The marine organisms (ammonites and sharks) can remain in the lower saltier water, but the mosasaurs would need to travel through freshwater to breathe air.

Plots for carbon and oxygen isotopes. Carbon isotopes (left plot, X axis) can be used as indicators of an organisms position in a food web with¹³C accumulating the further up the chain you go. The higher values (to the right of this plot) show that the mosasaur (E) from Morton County (blue shaded area) was certainly no lower in the food chain than any other animal from that site. Unfortunately the sample sizes are too low to make any kind of claim that it sat atop of the food chain. Oxygen isotopes can also be used to indicate whether an organism lives in environments with marine or freshwater (right plot) here, we see that the mosasaur (E) closely plotted to all the other freshwater animals from the Morton County site, each with relatively little ¹⁸O compared to true marine organisms.

The chemical evidence all points towards the conclusion that mosasaurs were capable of living in freshwater environments. What does this mean for our understanding of both mosasaurs and of the Hell Creek Formation? Image: Cristopher DiPiazza.

This study reveals how powerful geochemistry can be for interpreting past ecologies and environments. It also shows how important fieldwork and a little bit of luck can be in the scientific process. Image: Melanie During.

Dr Melanie During with Tyrannosaurus rex tooth (NDGS 15125) and mosasaur tooth (NDGS 12217).

Nathan Van Vranken in his teaching lab at Eastern West Virginia Community and Technical College.

Society of Vertebrate Paleontology Annual Meeting 2025

Emily Keeble — Wed, 03 Dec 2025 21:06:02 +0000

For 2025, SVP’s Annual Meeting was held in Birmingham, UK, giving the Palaeocast crew easy access to one of the world’s biggest palaeontological conferences. Emily and Filippo were able to record interviews with many of the delegates and the organisers of the event.

Since it’s not always possible to find the time and money to attend a conference in another country, we hope that such reporting can give you a sense of what it’s like to attend and speak to just some of the hundreds of people in attendance.

The welcome reception of the conference was held in the Lapworth Museum of Geology. This provided palaeontologists with the perfect backdrop for mingling and discussing many of the fossils on display.

Amongst all the scientific presentations and posters, there were exhibitors promoting their journals, societies and services to the researchers present. Amongst them were a lot of artists including model makers 10 Tons (who we interviewed way back in Episode 98: 10 Tons).

Even amongst all the amazing new discoveries discussed at SVP, 10 Tons’ ‘Time Traveller’s Guide to Gastronomy’, featuring models of extinct organisms served as food, was probably one of the most talked about things at the whole event. Would you have dared to try Anomalocaris sushi or Ichthyosaur steaks?

At the end of the conference, the ‘Awards Banquet’ was held. This is a huge event in which delegates are wined and dined before a an awards presentation is given. Served at this meal was Anomalocaris sushi…

This years award winners. These typically range from best student poster all the way through to lifetime achievement awards. Past interviewees amongst the winners included Romer-Simpson Lifetime Achievement Award winner Prof. Michael Benton (Episode 64: When life nearly died) and Lanzendoft-National Geographic PaleoArt Prize winner Bob Nicholls (Episode 80: Paleocreations).

Dr Emily Keeble amongst colleagues at the meal. It’s a great reminder that conferences are the perfect place to exchange ideas, to meet with potential new collaborators and to catch up with old friends.

Episode 170: Cariocecus bocagei

David Marshall — Mon, 15 Sep 2025 22:29:43 +0000

The iguanodontians were an incredibly successful group within the Cretaceous. They could reach incredible sizes, with the largest species even matching the proportions of some sauropods, and they also had an incredible palaeogeographic range, meaning that their remains are found all over the world today. In the late Jurassic, they were a lot less diverse and much smaller, so the late Jurassic and early Cretaceous are key times for understanding the evolution of this group.

Cariocecus bocagei is a newly described iguanodontian from the Early Cretaceous of Praia do Areia do Mastro, Portugal. Whilst it is only know from a partial skull, there are numerous useful characteristics that help identify it as a new species and fill in our understanding of iguanodontians in this important time in their evolution.

Joining us in this episode is Dr Filippo Bertozzo of the Royal Belgian Institute of Natural Sciences.
(Video interview available here: https://youtu.be/2fviWUan97s)

Cariocecus bocagei was discovered in the Lower Cretaceous Papo Seco Formation of Praia do Areia do Mastro, Portugal.

The block of rock was collected for the teeth that were externally visible, but after painstaking preparation, numerous different bones from of the right side of the skull, parts of the skull roof, and a nearly complete braincase were recovered.

The bones of the skull were studied using a CT scanner and revealed exciting new details about its cranial nerves and allowing for the most detailed reconstruction of the soft tissues of a dinosaur’s inner ear ever made.

Cariocecus bocagei was confirmed as a new species by the fusion of the maxillo-jugal complex and the tri-lobed shape of the supraoccipital. The ossification of the maxilla and jugal is interesting considering that other bones of the skull are unosssified. This means that the individual was likely a sub-adult, having not reach full skeletal maturity, but for whom the fusion of these bones was an important trait, possibly to aid chewing and processing of harder materials.
Image: 3D reconstruction of Cariocecus bocagei by Fabio Manucci.

Cariocecus was also in possession of a uniquely-positioned supraorbital (the spike-shaped bone positioned close to the eye). This structure is present in modern crocodiles and birds and is associated with support of the ‘supraorbital membrane’, essentially a soft-tissue ‘roof’ to the top of the orbit. Like in birds such as eagles, this supraorbital bone would have given Cariocecus a prominent ‘eyebrows’ which could have benefited/protected the eye in many ways.
Image: Life reconstruction of Cariocecus bocagei by Fabio Manucci.

Strict consensus tree showing Cariocecus at the base of Hadrosauroidea. Cariocecus is closest related to the British species Brighstoneus simmonsi and Comptonatus chasei, showing the close ties between the European hadrosauroids in the Early Cretaceous and perhaps a European/African origin for the group.

Reconstruction of Cariocecus bocagei by Victor Feijó de Carvalho.

Dr Filippo Bertozzo with part of SHN.832, the holotype of Cariocecus bocagei.

Episode 168/169: Grasslands

David Marshall — Mon, 01 Sep 2025 20:47:00 +0000

Grassy biomes, including grasslands, savannahs and crops, cover over 40% of all land on Earth. They play a significant role in carbon and silica cycles and have a large impact upon the climate. Grasslands (grass-dominated ecosystems) have shaped the evolution of numerous groups of organisms, most obviously grazing mammals, and can support a huge amount of biodiversity. Humans evolved in the savannas and through domestication of grasses formed agriculture, leading to a modern diet dominated by grasses such as oats, rice, wheat and corn.

As anthropogenic climate change threatens large scale uncertainty, it’s vital that we understand the controls that govern the success of this fundamentally important group. It is only by studying the evolutionary history of grasses that we might be able to predict how they will fare in future.

Joining us in this episode to speak about the challenges of piecing together the evolutionary history of grasses from a relatively poor fossil record is Prof. Caroline Strömberg of the University of Washington.

Grasses are monocotyledons (flowering plants whose seeds contain one embryonic leaf and that typically grow long linear leaves with parallel venation) belonging to the family Poaceae which contains around 11,000 species.

Grasses such as rice, wheat, oats, corn and sugar (pictured) form a vast proportion of the food consumed across the world. They can also be used for producing cattle feed or grazed directly. Grasses such as bamboo can be used for construction and others as biofuels.

Grasslands are ecosystems dominated by grasses. They don’t have to be monospecific, but can contain different types of grass as well as other plants too. Grasslands can host a surprising amount of biodiversity with some beating tropical rainforests for the number of species they can support. They cover over 40% of the Earth’s land having a huge impact on the climate, mineral cycles and other organisms.

From their evolution in the Early Cretaceous (no younger than 100 million years ago) to the first definitive body fossils in the Miocene (no older than 23 million years ago) the majority of the evolutionary history of grasses is without a reliable body fossil record. This is due to biases in the fossil record making grasses less likely to be preserved. This means that various proxies need to be used in order to understand the evolutionary history of grasses.

One line of evidence are the ‘phytoliths’ which form within the epidermal cells of grass. These are made of silica, so have a relatively good preservation potential. The shape of these phytoliths are known to carry a phylogenetic signal, meaning it’s possible to match different phytolith shapes to different groups of grasses.

Whilst it’s not possible to reconstruct what an individual plant would have looked like from a single phytolith, they can be used to collectively build a picture of which groups of grasses were present in the area surrounding a depositional environment. This can still be incredibly informative and much of the understanding of grass evolution comes from such microscopic fossils.
Image: Eocene bamboo grass rondel phytolith from Nebraska. Phytolith = ca. 12 µm across. Credit: C. Strömberg.

In fact, phytoliths and epidermal remains have even been discovered in between the teeth of a basal hadrosaurid, Equijubus normani, from the Early Cretaceous of China. These fossils provide direct evidence that grass formed part of this dinosaur’s diet. Other such direct evidence of dinosaurs eating grass comes from coprolites in the Late Cretaceous.
Credit: Yan Wu, Hai-Lu You, Xiao-Qiang Li, Dinosaur-associated Poaceae epidermis and phytoliths from the Early Cretaceous of China, National Science Review, Volume 5, Issue 5, September 2018, Pages 721–727, https://doi.org/10.1093/nsr/nwx145 Licence: CC BY 4.0.

These epidermal fragments and phytoliths are the oldest-known grass fossils and were, unsurprisingly, from one of the earliest-diverging groups of grasses.
Image: (a, b) Silicified epidermal pieces. (c-h) Three slightly bilobate phytoliths with (c-f) representing one phytolith in four different views. LC, long cell; SC, short cell; SCP, short-cell pair; ST, stoma.
Credit: Yan Wu, Hai-Lu You, Xiao-Qiang Li, Dinosaur-associated Poaceae epidermis and phytoliths from the Early Cretaceous of China, National Science Review, Volume 5, Issue 5, September 2018, Pages 721–727, https://doi.org/10.1093/nsr/nwx145 Licence: CC BY 4.0.

The spread of grasslands can also be observed in the fossil record of other organisms. Adaptations to life in such open habitats includes the possession of long legs, large body size, tooth morphology and large chewing muscles. The evolution of ‘hypsodont’ teeth (those with high crowns) was required to deal with the erosion from processing grasses containing hard silica phytoliths and so are good evidence for a grazing lifestyle and thus the presence of grasslands.

Another powerful proxy are carbon isotopes. The ratios of ¹²C and ¹³C within a plant is indicative of which type of photosynthesis it uses, whether C₃ (orange) or C₄ (blue), with the latter type able to fix more of the heavier ¹³C from the atmosphere. C₃ photosynthesis conveys an advantage for plants at higher latitudes/altitudes, in lower temperatures and in higher atmospheric concentrations of CO₂ whereas C₄ photosynthesis is more efficient for plants in the opposite climates. C₄ photosynthesis is able to conserve more water, so C₄ plants have a distinct advantage in the tropical lowlands where seasonal rainfall and aridity become an issue. Within grasses, C₄ photosynthesis independently evolved at least twenty times and all within the PACMAD group of grasses. Therefore, if elevated amounts of ¹³C are detected in an analysis, then it likely points to the increased presence of grasses from the PACMAD group and a warmer subtropical climatic regime. Furthermore, since grasses are primary producers, this isotopic signal is picked up by herbivores and carried up the food chain. It’s therefore possible to estimate the proportion of C₃ and C₄ plants supporting your own diet from the carbon isotopes making up your body. Credit: Gallaher et al. 2019.

Using phylogenetics (the study of evolutionary relationships), it’s possible to observe large scale evolutionary patterns. The tree above, when colour-coded to show the habitats of modern species reveals grasses with the earliest divergence dates to mostly live in the forest understory (dark blue), whereas bamboos occupy forest margins (light blue). By comparing the tree above with the previous one showing C₃/C₄ photosynthesis, it’s apparent that most C₄ grasses are open habitat (red), whereas only one group outside of the PACMAD group, the Pooideae, are open habitat. Credit: Gallaher et al. 2019.

By collating the lines of direct and indirect evidence, a wider picture of the evolution of grasses and grasslands can be produced. There are big delays between the evolution of grass, the first record of open habitat grasses and the formation of the first grasslands. Surprisingly, the timing of these events also differs between continents.

Interestingly, the evolution of hypsodonty (high crowned teeth) occurred tens of millions of years before the first evidence of grasslands in South America. Untangling the controls behind these patterns is key to understanding how grasses and grasslands are likely to respond to future change.

With With anthropogenic climate change delivering future uncertainty, food crops may be at risk. How might C₃ crops, such as rice (pictured) fare with increased temperatures and aridity? Would it be better to plant C₄ plants in such areas or could a C₄ rice be engineered? By studying the evolution of grass and grasslands, we are better informed to be able to tackle such huge issues.