Sunday, 17 September 2017

Killing-off Scottish Mammaliamorphs


With all that’s been going on in the last couple of months (including attending #2017SVP, Friends of  Hugh Miller’s conference, and #SVPCA2017), I’ve dropped the ball on this blog a little. But this week past, I had a new paper out with my colleagues, and I’d like to tell you a little more of the story behind it.

When I started my PhD on the fossil material of the Isle of Skye, one thing really surprised me. There is this whole group of cynodonts (the group that includes mammals and their nearest relatives) that no one ever talks about. They are called tritylodontids (Tritylodontidae), and despite being one of the most successful clades of cynodont – lasting from the Late Triassic to the Early Cretaceous - they are often overlooked by palaeontologists. We find one of them, called Stereognathus, on the Isle of Skye.

Perhaps the best piece of tritylodontid palaeoart, Stereognathus by Mark Witton - find out about him and support his amazing work.
Why should we care about these extinct animals? Well, they’re considered by most to be the sister-group to stem-mammals, and share many similarities in their skeleton. They would have had fur and whiskers, and you might mistake one for a mammal if you didn’t look too closely. But there are some interesting differences. Tritylodontids kept the so-called “reptilian” jaw joint, between the articular and quadrate bones (not the dentary squamosal contact as in mammals). They also didn’t go through the same extreme reduction in body size we see in the first mammals. In fact some tritylodontids were pretty big, like Kayentatherium from the Early Jurassic, which was about a metre long. Unlike the earliest mammals, tritylodontids appear to have been almost exclusively herbivorous. 

All in all, they were occupying quite a unique ecological space in the Mesozoic. 

The first tritylodontid to be described was the Middle Jurassic genus, Stereognathus, named by Charlesworth in 1855 and figured by Richard Owen shortly after. Unfortunately though, this genus is represented by only a few molar teeth in a fragment of jaw, a bunch of isolated teeth, and some individual bits of limb that might belong to Stereognathus, but it’s uncertain. Most of this material came from sites in Oxfordshire, England.

Owen's original figures from 1857, scanned courtesy of BHL
Charlesworth called this first Stereognathus species, S. ooliticus, after the rocks they were found in, the Great Oolite Group. More than 100 years later, Robert Savage and Michael Waldman found teeth they identified as belonging to Stereognathus on the Isle of Skye. They named it S.hebridicus, after the Inner Hebrides islands, to which Skye belongs. Unfortunately, their only diagnostic feature was that S. hebridicus was bigger than S. ooliticus. More recently, field work has recovered more S. hebridicus teeth. I decided to set out with my co-authors to test this size difference, and see if we could find other details of the shape of the teeth that would support S. hebridicus as a separate, Scottish species.

A photo of the most intact Stereognathus specimen, found and photographed by Andrzej Wolniewicz during fieldwork on the Isle of Skye.
We microCT scanned all the new material, and the two holotypes of S. ooliticus and S. hebridicus. We measured every tooth of Stereognathus we could, many being too broken up or worn to give reliable measurements. Even the original S. ooliticus specimen figured by Owen was in a poor state: comparing it to his original drawings, it seems over 150 years of handling hasn’t been kind. 

We didn’t find any good evidence that S. hebridicus is a separate species from S. ooliticus. Although the holotypes are radically different in size, when you take all of the rest of the teeth we have and plot them (below), they fall along a line of size change you’d see from baby to adult. In other words, size differences are explained by ontogeny: development from earliest stage of life to maturity.

All of the measureable upper postcanines of Stereognathus. Empty diamonds and squares indicate less certain measurements (due to breakage or wear). I expect field work over the coming years will fill this out to show a full growth series of size.
To my horror, I’ve sunk the only Scottish tritylodontid! Unless new fossils tell us differently, it looks like S. hebridicus is no more. This makes it a junior synonym to S. ooliticus. It’s a shame to see our Scottish species go, but I’m glad that it was the new, more physically intact fossil material from Skye that allowed us to make a proper reassessment of the genus. 

Who knows, perhaps there is still another species of tritylodontid to be found on the rocky shores of the misty Isle? I'll keep looking and get back to you...

______________________________________________

Panciroli E, Walsh S, Fraser NC, Brusatte SL, and Corfe I. 2017. A reassessment of the postcanine dentition and systematics of the tritylodontid Stereognathus (Cynodontia, Triltylodontidae, Mammaliamorpha) from the Middle Jurassic of the United Kingdom. Journal of Vertebrate Paleontology.

Waldman M, and Savage RJG. 1972. The first Jurassic mammal from Scotland. Journal of the Geological Society of London.

______________________________________________

Monday, 8 May 2017

Reconstructing Wareolestes

I recently blogged about the paper I published with colleagues on a little lower jaw from the Isle of Skye, belonging to a Middle Jurassic mammaliaform called Wareolestes rex. Like many people, I believe palaeoart is a vitally important part of palaeontology, particularly as it forms a quick visual bridge between palaeontologists and the public (and other scientists!) So I'd like to share the palaeoart reconstruction I did for Wareolestes, and talk about the science behind it, and the informed speculations.

My reconstrction of Wareolestes rex. The only bit we have is the jaw, so much of the rest is informed speculation - informed by scientific evidence and living relatives. It was done in pencil and watercolour.
Mesozoic mammal art is often not especially inspiring. This is partly because, until recently, most Mesozoic mammal fossil material comprised mostly teeth, which doesn't make for exciting reconstruction. As a result, artists often choose to have their early mammals snarling to show off their teeth - the only certain part of the art! Also, there is a strange notion that small mammals are boring; but a quick google search will soon show you that they come in a huge range of sizes, shapes, colours and behaviours: inspiration for great art. However, palaeoartists have shied away from speculating about fur, whiskers, ears and eyes, erring on the side of caution and producing identikit snarling, splayed mice.

I wanted to try and do a reconstruction in an informed, but moderately speculative way. So I started work on reconstructing Wareolestes while I was still writing the paper, hoping to make informed scientific decisions based on the fossil record, and then spice it up by examining the appearance of small modern mammals.

There is a bit of a tradition among Mesozoic mammal workers of comparing our fossil beasties to the modern American opossum (Didelphis). I therefore decided to use them as the inspiration for the reconstruction. Google 'opossum', and you'll find this animal has a penchant for bearing its teeth. While I shy away from those snarling Mesozoic mammals on the whole, our Wareolestes fossil is a lower jaw with teeth, so this open-mouthed expression was fitting in this instance. Forgive me for breaking my own rules; in future reconstructions of fossils (especially those with known postcranial skeletons) I'll avoid teeth-bearing if I can help it.
The opposum Didelphis is a wonderfully charismatic creature, and often photographed bearing its teeth - handy if your fossil is a jaw with teeth and nothing else.

Our digital reconstructions (from microCT scans) were the base for drawing the mouth and teeth. As not all of the teeth were present in the Scottish fossil, I modelled the other premolars, canines, and incisors on two closely related genera: Dinnetherium and Megazostrodon. On reflection, I think the front of the snout should have been a little longer... but at this point we don't know for sure. I based the rest of the skull on Morganucodon, as Wareolestes is a morganucodontid. However, I made it more robust, because Wareolestes was larger and chunkier than it's geologically older relative.

Amazing skull and muscle reconstruction of Morganucodon by Lautenschlager et al 2016

So far, so good. But from here on, things get more speculative. We don't have much in the way of preserved fur or skin, except for a few exceptional specimens from China (but not of this genus). Undoubtedly Wareolestes and other Mesozoic mammals had fur, inherited from their non-mammal ancestors. They almost certainly had whiskers: we know this thanks to evidence in the fossil record for innervation in the snout. Whiskers probably developed in earlier non-mammalian cynodonts, which would have used them to sense their way through burrows. Therefore, whiskers are very likely to have existed in the earliest insectivorous, nocturnal mammals, being used to sense their environment and hunt for food.

I chose the shrew and the Solenodon (see below) as inspiration for the whiskers. Their whiskers extend quite far up the face, and point in multiple directions.

Beautiful shrew, showing off those sensitive whiskers. (By David Chapman, from the Cornwall Mammal Group)

The nose and ears of Mesozoic mammals, being entirely composed of soft tissue, are also impossible to reconstruct without speculation. Mesozoic mammals had well-developed olfactory bulbs, so they had a good sense of smell. I went for an opossum-like nose, simply because I liked the look of it.

This sleeping opposum's nose might be the cutest thing EVER. Will the squeeing ever stop? (I got this off pintrest, contact me if you can ID the source)
At this point in their evolution, mammals still had their post-dentary bones attached to the inside of the jaw. These bones would later reduce and detach, becoming incorporated into the middle ear. This allowed mammals to develop exquisite hearing, especially at higher frequencies. So what did they hear when the bones were still attached to the jaw? The answer is: we don't know. The postdentary bones were certainly used in hearing though. I decided a small proto-ear was fitting, based on a slightly crumpled version of a Solenodon ear. I kept them simple, small, and placed low on the head.

Close up of a Hispaniolan Solenodon (Source)
The Solenodon is a small, nocturnal insectivorous mammal found on some Caribbean Islands. It is weird on so many levels, not just because it has venomous saliva, but also as it is the only genus surviving in its family, the Solenodontidae. Phylogeneticists trace their origins back to the Cretaceous, making this an altogether unique animal and a good analogue to find inspiration for the life appearance of Mesozoic mammals. Because of this, I chose the dark back and upper face of the Solenodon as inspiration for the colour pattern on Wareolestes. However, I then decided to add a little cheek and eye patterning.

Finally, the eyes. It's hard to say how large the eye would have been relative to the head. I decided to go for something I know and love: the eye of a rat. This is a total bias on my part, because as a many-time rat-owner, I was always delighted by those dark chocolate beady eyes greeting me each morning, staring with demented twitchiness and tiny black pupils pointing in opposite directions. Mental. However it does leave me open to all the "aren't Mesozoic mammals just a bunch of rats anyway?" comments people just can't restrain themselves from making... sigh*.

So there you have it. This was my process in creating a Wareolestes rex reconstruction. I played it kind of safe - I could have speculated about the rest of the body, but decided it was a step too far. This is my first proper go at palaeoart, so theres still a lot to learn. Hopefully in the coming field seasons we'll find more of the skeleton and I'll be able to revise the image based on more evidence. In the meantime, I'd love to hear you opinions, ideas, and comments - get in touch on twitter: @gsciencelady

*Wareolestes is of course not a rodent, as rodents didn't evolve for another 100 million years. The resemblence is superficial, anatomically they are totally different.


References
 
Benoit J., Manger P. and Rubidge BS. 2016 Palaeoneurological clues to the evolution of defining mammalian soft tissue traits. Scientific Reports.

Lautenschlager S., Gill P, Luo Z-X., Fagan MJ., and Rayfield EJ. 2016 Morphological evolution of the mammalian jaw adductor complex. Biological Reviews.

Panciroli E., Benson RBJ., and Walsh S. 2017 The dentary of Wareolestes rex (Megazostrodontidae): a new specimen from Scotland and implications for morganucodontan tooth replacement. doi: 10.1002/spp2.1079

Thursday, 4 May 2017

Wareolestes bears its teeth


I've been lucky enough to see my first scientific paper come out this year (first of many I hope!) with my wonderful colleagues, Stig Walsh (National Museum of Scotland) and Roger Benson (Oxford). We describe a lower jaw found in the last couple of years during our ongoing palaeontological fieldwork on the Isle of Skye, Scotland. It is from the Kilmaluag Formation: which is Bathonian (Middle Jurassic) in age. This lower jaw, or dentary, is from a relatively large Mesozoic mammal living around the lagoons of Jurassic Skye. What’s more, it retains replacement teeth inside the jaw. This is the first evidence for mammalian milk-production in the Scottish fossil record.

The surface of the jaw was somewhat abraded by wave action, but microCT scanning revealed exceptional details (see below). The fossil is now part of the collection at the National Museum of Scotland: NMS G.2016.34.1
The Middle Jurassic was a pivotal time for most major groups of animals: the dinosaurs exploded in diversity, as did Mesozoic mammals. Unfortunately for palaeontologists, globally there is a real paucity of Middle Jurassic rocks. This makes the sediments of Skye internationally important. On the island, myself and my colleagues are uncovering the secrets of Middle Jurassic mammals and other small vertebrates from this elusive time period.

We identified the jaw as belonging to Wareolestes rex. Scampering between the copious turtles and crocodiles near the water’s edge, or perhaps living further inland and being washed by river into one of these lagoons upon death, Wareolestes is one of the many mammals that lived alongside their infamous dinosaurian cousins at this time in Earth's history.

Wareolestes was previously only known from a couple of isolated molar teeth from England. Palaeontologists disagreed about whether the holotype (the tooth that defines this species) was an upper or lower molar, and whether it was left or right. Now that we have a dentary with teeth in it, we can compare it with the holotype and say that the holotype is a lower left.

The holotype of Wareolestes rex, NHMUK PV M36525.

Wareolestes rex means 'Ware's brigand king', and was named in 1978 by Eric Freeman as one of several species recovered from productive fossil mammal beds in Oxfordshire. Productive is a relative term for mammal palaeontologists. We aren’t talking about whole skeletons, or even skulls. We’re talking teeth – lots and lots of them. Sifting through them, many new species of mammal were discovered and named. You can flick through hundreds of them in the collections at Oxford Museum of Natural History and the London Natural History Museum, each mounted on a pinhead and placed carefully in its own miniature test tube.

This Scottish jaw is not only the most intact fossil of this species yet found, but only the third species of Mesozoic mammal described from Scotland. (There are others: watch this space, as I'll be publishing on them in the near future).
The jaw was microCT scanned, revealing beautiful detail. This is a lingual view. Inside, several unerupted adult teeth tell us this animal was a sub-adult, and that it had the modern pattern of mammal tooth replacement.
Our Scottish Wareolestes rex fossil would have once held five premolar teeth and three molars. Only the three molars remain in place above the gumline. The front of the jaw with canines and incisors (the puncturing and biting teeth) is missing. The part of the jaw that hinged against the skull is also missing at the back. It may seem quite paltry: only 2cm long, missing lots of bits, worn on one side by the crashing waves... Yet this is unusually complete for a Mesozoic mammal, let alone one from the UK, or the Middle Jurassic.

When we microCT-scanned it, we discovered the replacement teeth were still sitting inside the jaw. They are all more or less level, suggesting the teeth were replaced at approximately the same time. Sadly, the wave-damage means we can’t be more definitive about the sequence of replacement, but hopefully future finds will confirm this.

This is my palaeoartistic reconstruction - check out tomorrow's blog for more on how it was made.
This is my palaeoartistic reconstruction of Wareolestes rex, based on our jaw fossil from Skye. (Come back soon for my blog on Reconstructing Wareolestes, where I discuss the science and speculation behind my reconstruction of our Jurassic Scottish beastie). While it is sad to think it never made it to adulthood like its siblings, because this wee beastie was preserved in the fossil record, we can learn more about Mesozoic mammal evolution. It also adds to our picture of great diversity of Middle Jurassic ecosystems on the Isle of Skye.

My 3D print out of Wareolestes rex.
All of the microCT scan data, plus 3D models of the dentary, are available on Data Dryad - DOI: doi:10.5061/dryad.5n36j - so you can download a model and print your own copy. Mesozoic mammals: every home should have one.

----------------------------------------

References 

Freeman EF. 1979 A Middle Jurassic mammal bed from Oxfordshire. Palaeontology, 22, 135-166.

Panciroli E., Benson RBJ., and Walsh S. 2017 The dentary of Wareolestes rex (Megazostrodontidae): a new specimen from Scotland and implications for morganucodontan tooth replacement. doi: 10.1002/spp2.1079