Common fat-tailed mouse opossum

Transcription

JANE PICKERING: Good evening, everyone. First, I'd like to thank you all for coming out on this dreadful day. I know it's a little bit better now than it was this afternoon, but it's quite the weather out there. So congratulations on getting here in one piece. My name is Jane Pickering. I'm the executive director of the Harvard Museums of Science and Culture. And I would like to welcome you to this, which is the third in our Evolution Matters series for the Harvard Museum of Natural History. And I'm delighted to welcome our speaker, who has-- moved-- Dr. Zhe-Xi Luo, who is speaking to us today about Jurassic grandmothers from China. Before I get to Dr Luo, I would like to just mention that our next lecture is two weeks from now. Robert Hazen of the Carnegie Geophysical Laboratory is going to talk about the new evidence about complex evolving systems driving phenomena beyond life sciences, such as the diversification of minerals on Earth, and how this impacts the arguments of the anti-science proponents. And this will be on Tuesday, March 12 here in this hall. I know we have some special guests in the audience, some local teachers, who are coming and have a seminar after the lecture. And just to let you guys know that the seminar room is pretty much exactly behind me, right next to the bathroom. So if you go out and just keep going around, then you'll find your seminar room. All the museums classes and exhibits can be found in this brochure, which is on the table. And so please if you don't have one, feel free to pick one up. And if you don't like paper, then sign up your email addresses. Then you'll get everything in your email box about the information about all our programs. And there's also information there if you would like to become a member of the museum and help support out exhibits and our public programs. So speaking of support, I'd like to acknowledge Dr. Herman and Dr. Joan Suit-- who are sitting over there-- who supported the Evolution Matters series. And they have been dedicated sponsors-- thank you-- for the last several years. And thanks to their support, we are able to videotape these lectures. And they are available online. So we just put up the first of this year's series by Dr. Arkhart Abzhanov who did the first-- those of you who were here remember-- which was fantastic, on dinosaur to bird evolution. That's now available, but there are many, many others available. So I encourage you to check that out. So on to tonight's speaker, who's really being welcomed back to Harvard rather than just welcomed to Harvard. Luo is a paleontologist and professor of organismal biology and anatomy at the University of Chicago. And his research specialty is on the evolutionary biology of vertebrates. Tonight, he will focus on the originations of modern mammalian biological adaptations in the Mesozoic Era, often dominated by dinosaurs. And the story isn't quite as simple as that-- and how Jurassic fossils discovered in China shed light on the earliest evolution of placental mammals. He received his Bachelor of Science from Nanjing University of China and then went on and did his PhD in paleontology at the University of California, Berkeley. He completed his postdoctoral training from '89 to '91 right here at the Museum of Comparative Zoology at Harvard. And then left and spent many years as curator of vertebrate paleontology at the Carnegie Museum of Natural History in Pittsburgh. And in 2004, he was named their associate director of research and collections. He is considered one of the world's top experts in Mesozoic animals and is counted among the international team of scientists that has discovered many significant species of early fossil mammals, such as the earliest known swimming mammal and the earliest marsupial. He has received a career award from the National Science Foundation as well as a Humboldt Research Award for senior scientists from the Alexander von Humboldt Foundation. The list goes on, but let's hear about Jurassic grandmothers from China, origins and early evolution of mammals. ZHE-XI LUO: My name is Luo. And in English, it coincides with being short. So every time I talked about early mammals in front of my mentor, Fuzz Crompton, I'm always a short guy. And I shudder. Mammals are very diverse. We have some 5,400 species. And mammals are also very interesting in that we have some 4,000 general fossils. So as a case study of a major clade evolved in the evolutionary history, mammals gave us a great fossil record to work with. Mammals are also very interesting in the ecomorphological diversity. Let me back step a little bit. This is an old exhibition diagram by William King Gregory in the American Museum of Natural History. And it's a tree. It has many different. ecomorphotype maps on top of it. Foremost, we have orders familiar ground-living terrestrial mammals. And of course there are things that can climb the tree. And once you get on the tree, it's not uncommon you can get 40 arboreals. And some of the arboreal forms could be volan or it could even power flyers. You can have tongue feeding termite eaters. And then you can have cursorial herbivores. And you can have heavy animals that are gravid portals, such as elephants, rhinos, so on and so forth. And then you can have carnivore scavengers. You can have carnivore predators. And then finally you have 40 aquatic mammals. And these are called ecomorphotypes. A lot of the classic natural historians have set up orders of mammals that largely go with these ecomorphotypes. But of course, the modern molecular phylogeny have already validated from the genetics perspective most other orders except what we call the insect worth. So mammals are interesting in ecology and in their general morphology. But all these diverse ecomorphotypes share some basic features in reproduction. The first group is called monotremes. And these are mammals. They have mammary glands. And they nurse their young, but their young were born with leathery eggs. And in contrast to the egg laying mammals, we have live birth mammals. The first group of these are marsupials. Marsupials have very short gestation, but rather long nursing in the pouch. And the here is a figurative metaphor of that. The most diverse group within the live birth mammals are placentals. There are some-- depending on how you read the latest phylogenetic literature-- over 5,000 species. Generally speaking, the placentals have very elaborate, more invasive placenta that embed in mother's uterus. As a result, they have a longer gestation and more effective nurturing in the uterus. But all these mammals, in addition to their basic reproductive ecomorphological patterns, they also have bones. They also have teeth. It is really through these osteological and dental features that we can tie the fossil record with the modern mammalian diversity. Normally, this is how we do it. So we have placentals. We have marsupials. And these are united as a theory and characterized by live births. In addition to their mode of reproduction, they also have teeth, jaws, and skeletal features that characterize them. This theory of mammals is grouped together with monotremes, altogether characterized by mammary glands, and also derived by many bone features, so on and so forth. Suppose we find a fossil. Without this soft tissue, it is really these osteological features that help us to place them on the overall mammalian phylogenetic tree, or family tree. Say for example, if we find a fossil called Juramaia, or mother from Jurassic, and its osteology and teeth place it closer to placental than to marsupial, but not so advanced yet to be inside placental. We will call them Eutherians. And if we find another fossil-- say like Sinodelphys and by teeth and the bones-- we are able to associate it closer to modern marsupials than modern placentals. We would place them close to the marsupial line and call them Metatherians. It is really these early fossils that gave us a time estimate about how a particular evolutionary lineage first appeared in the fossil record. But more important than that-- and very interesting to me-- is these fossils would have helped to provide what our ancestors would be like if these are closely related to us. It is a review from this ancestral future, from the early fossils, that we managed to establish the evolutionary pattern, how the modern group arrives in the fossil record. Now these fossils associated with individual lineage gave us the minimal age of a group. But all these extant mammals would also carry with them the molecular sequences that are part of their history. And through those molecular sequences, we can manage provide a molecular time estimate. Usually there is a gap between the minimal fossil age and the rise, or estimate of divergence of time, of a modern group represented by molecules. But the biggest early mammal diversity are not the group associated with placentals or not a group associated with marsupials. They are associated here in between the Therian mammals and monotremes. So these are modern Metatherian and Eutherian groups. There are a large number of groups. And there are some 15 different groups in between Therians and the monotremes. And these groups, from my perspective, are the most interesting mammals. And many of them are from Jurassic. And the modern mammal group in return is nested inside a group called mammalial forms. And these are fossilized mammals from very early age. And their osteological and dental features definitely associate them with the modern groups, but they are not quite there yet. So we call them mammalial forms. And mammalial forms, in return, are descendants from a large cluster of extinct reptilian fossils known as Cynodont. Normally we refer to them as mammal-like reptiles. And of course we will have this osteological feature to construct this phylogeny. And in the more recent months, a much larger version of the study is published by a team funded by National Science Foundation Tree of Life mammal project. But if you take the genealogical tree and place it on the time scale, you will see some very interesting sequences. And the next diagram will follow from here to here, but vertically. We start with Cynodonts that are very common in Triassic. And this is even before the rise of the dinosaurs. Right around the time when dinosaurs start to appear in the fossil record-- in this time, roughly from 220 million to 190 million years-- we start to have these fossil groups come into the fossil record. And they definitely have some part of the modern diagnostic mammalian features. That's why we call them mammalial forms. And if we move a little more recent to the middle Jurassic from 180 million to 165 million, we start to see the fossils that we can definitively place on modern clades that is either associated with modern monotreme to some extent, or associated with modern Therians to the other. Well by the time we get to late Jurassic, up to 145 million, we start we have groups that can literally identify with Eutherians suck as Juramaia. Cretaceous is also very interesting. But if you go from present date-- back step 220 million years-- the first 1/3 of that 220 million years of the mammalian history are in Jurassic. And by the time we get to Cretaceous, the game is almost over, so thus our theme about grandmothers from the Jurassic today. The fossils from this earliest episode really can help us to frame some questions and address how modern mammalian biology really arose in the fossil history. And this is the work that I started with Fuzz. So I'm particularly happy that Fuzz is able to come and probably offer me some criticism after this. The biggest question is, what makes a mammal? Mammals have hairs. Mammals have mammary glands, sweat glands. These are all integumentary structures associated with the skin. Mammals are also characterized by this very important feature of lactation. Of course in the fossils, we cannot get the fossilized mammary glands. But because lactation has such a fundamental impact on how mammals grow. And we have related growth and dental features that allow us to extrapolate when lactation really originated. And of course, it's a no-brainer that the brains are important for us. Right? Without capacity to think, we wouldn't to be here to sit and enjoy this discussion. Let's look at the hairs and integuments. And I'll start with one fossil. And the name is not important. Let's call it Docodont. It's some extinct mammal from a long time ago. And this particular Docodont has very interesting scales associated with a big fat tail, flattened. And in between those tails, you can clearly recognize impressions that are the hairs. And in addition to this hairy, scaly tail-- and you can also tell that the tail vertebra themselves are flattened. And they have this so-called transverse processes. These are some barbs sticking out to the side. And if you compare the morphology, they are really very similar to modern beavers and modern otters, both of which are aquatic. And it is with this kind of a feature, we are able to understand that this guy definitely flapped this tail in a beaver-like swimming motion. And has very powerful fore limbs. That's good for both digging and also for swimming, such as the case of the modern platypus, which is also semi-aquatic. And then it's teeth have these re-curved, sharp cusp, which is a typically interpreted and is associated with fish feeding. So besides the fact that we've got a cool beaver-like mammal, the important fact is really we have this very ancient extinct group close to mammals, but are not quite within the modern mammal group really fossilized with fur. Basically the mammalian integumentary structure originated well before the modern mammals. And our additional fossil gave us the same information. Well you may say, it's not a big surprise. Mammals have fur. Mammals have these skin-related derived features. Yes, it's not a big surprise, but nonetheless reassuring that we can have fossil records good enough to give us an understanding of those common sense features. Next I'm going to talk about lactation and the growth pattern. I will start with a fossil that's Fuzz Crampton's favorite. This is called Pachygenelus. And this is from the early Jurassic in southern Africa. But the larger group of this actually has much wider distribution in Argentina and in southern Africa. That record of this particular group related to Pachygenelus really goes back to late Triassic as old as 220 million years. And if you'll look at the very interesting pattern of this Pachygenelus fossil, you realize that these PC-- not politically correct-- they are called post-canines. These are the teeth behind the larger canines. And if you look at the pattern of this, you can definitely, in a single fossil, recognize at least three generations of teeth. So this guy had multiple replacements. Besides the fact they replaced quite often, multiple times, they also have an alternating replacement. The fact is probably not as important if you compare it to the modern Therian mammals. And here is an Australian dingo. And it belongs to placental canis, relevant to modern dogs. In a younger individual, we have deciduous teeth. In an older individual, we have these permanent [INAUDIBLE] teeth, which are called molars. When these deciduous teeth are being replaced, it goes sequentially. The premolar, or relatively simpler teeth behind the canine, when they replace they go in a sequential way from this end to this end. And then there are marsupials. And marsupials have one single tooth being replaced. The question we can ask with this early fossil is how we get from the ancestral condition of our near relative to our kind of condition. The reason we care about this is this pattern is related to determinant skull growth and the lactation. Where as this is essentially our ancestral condition. Let me back step a little bit and explain what we call determinant skull growth. I'm going to digress and use an example from the other side of modern vertebrate diversity, and that's the birds and the dinosaurs. Dinosaurs have very rapid early growth. And a modern dinosaur, like ostrich and the emu, facilitated by this intense parental care, they grow really fast early on. But once they get to a certain stage, they reach adult body size. For the rest of the longevity, they have this plateaued growth curve. In contrast to dinosaurs and birds, there are many reptiles that would have the slow and the sustained growth. And this little what we generally call the indeterminate growth. Let me redefine a little bit. And this particular diagram is a contrast between a rodent representative of modern mammals and a crocodile. And generally speaking in non-mammalian, non-avian vertebrates, you have the slow early growth, but the growth never quite stops. Whereas in birds and also in modern mammals, facilitated by-- in a mammal's case-- lactation, you have this rapid burst of early growth. But because you grow so fast, you have to taper off somehow. So we have this plateaued growth curve. And this gave us a very interesting body size difference. And if you have good fossil preservation from some fossil sites, you can literally test. And you can distinguish if some samples of fossils actually better indication prolong the juvenile growth and then a long, sustained but slow growth throughout the life. And if we have a case like this, we will call it indeterminate growth. It's not really completely without end, but the sense is it won't go on for quite a while. But if we can do the statistical analysis of a decent fossil sample, we can distinguish a fossil group that has achieved very high, rapid growth. And then their growth, once they reach an early age, would be leveled off. Now here is some real data. And this data are some small rodents, small insectivores, compiled by a very active research lab in the 1970s in Smithsonian. John Eisenberg was the main leader of that. And these are the days of individuals that you can sample from the field. And these are 100 persons of your adult body size. And if you map out of the growth pattern over a time course, you get a sense of many rodents pretty much mature, reach their adult body size, in 30 days. It's very, very fast. And some of the insectivores would go on a bit longer. And they would go 60 days, 70 days. But the bottom line is all mammals grow very fast. Why do they grow fast? We all benefit from mother's milk. And we all benefit from mother's lactation. But once we get up to that adult body size, our body size is fixed. And therefore we have this fixed plateau. It is a really in this early phase-- because we benefit from the lactation-- and we can grow without teeth. And that shaped our basic growth pattern. So everything else being equal-- because we have to plateau off-- we have early termination of the possibility to replace our teeth. Therefore this lactation really gave us a different time table for growth. And if we are able to establish the so-called limited dental replacement called diphyodont dental replacement, we are able to infer that this particular fossil must have had lactation. So that's my reasoning. And here is a fossil. And it's called it Brasilodon. It's from late Triassic Brazil. And this is a young individual. This is an older individual. That's the time course. And you can tell that the front part of the teeth are still reptilian. Whereas the backside of the teeth-- and they are having the sequential replacement. There is no longer this rapid replacement of teeth. Therefore, it's more mammal-like, if not yet fully mammal-like. Now if we go along in the evolutionary tree to another, more derived form, called Sinoconodon. And this is one of those mothers from early Jurassic China. This is actually a work that I did with Fuzz together. And this is a very young individual. This is a very old individual. I'm not going to quiz you about this very complex diagram, but it's suffice to focus on this anterior part. It's called canine. And even in the youngest individual, we have replacement of canine. If you map it out, you'll find the following pattern. And that is the anterior teeth are replaced multiple times. And we can actually know for sure, at least, the Sinoconodon has four generations of dental replacement. It will replace early. It will replace late. Basically, it keeps on replacing and never fully stops. And that is a reptilian pattern. Now if you compare this fossil, Sinoconodon, with another fossil that is relatively well-known called Morganucodon. You will find the size pattern is very different between the two species. Essentially, Morganucodon has a very narrow growth band. And it fits the prediction of the general model. If you have a sample of a fossil-- very few babies, relatively few adults-- and they have a very narrow size range. In contrast to that, Sinoconodon has a much wider range. And in fact for the range of Sinoconodon we have, we estimate it can grow almost like 44 of the body mass. But the key is the small Sinoconodon replace their teeth. The larger Sinoconodon would replace their teeth. And essentially, minus the fact that Sinoconodon looks like a mammal with the jaw. It looks like a mammal with ear, but its dental replacement is reptilian. So what does this tell us? It'll basically tell us that in these near mammals, we start to have more mammal-like condition. There is a reduced replacement of teeth. And in Sinoconodon, we already have-- very likely-- a single generational of most of the teeth in the jaw. But the anterior teeth are still replaced multiple times, early and also late. It is a really in Morganucodon we start to have no replacement of molars, therefore achieves the truly modern mammal-like diphyodont dental replacement. That usually facilitated in extant mammals by lactation. So if you take this logic one step further, essentially between Sinoconodon and Morganucodon, we caught the evolutionary moment when we started to have the earliest fossil that can tell us reliably lactation has arisen. I wonder if we have dental care professionals here. You will wonder every time you go into the dentist's office, they have all these fancy gadgets that make noise. You'll open your mouth, half sedated. And they do this on your teeth. It's suffering, right? Well the bottom line is we have limited dental replacement. If we get stuck with that permanent dentition, we have to take care of it. So essentially next time you tell your dentist, it's really that moment 220 million years ago. Evolution took a turn for the better. So you get to suffer. And he has a job. Well let's move on to the brain. We have this very nice fossil. And this is prepared by the very best fossil preparer, Bill Amaral, who worked here in the MCZ for many years. He just recently retired. Besides the fact we can figure out a many anatomical details, this is the one we know about this particular fossil. And it has a relatively large brain case. And we named it Hadrocodium. Hadro means full. Codium means big head. And so this is also a study that I did with Fuzz. Now if we scale the size of the brain against the estimated body mass, we have this EQ. It's not emotional quotient. It's encephalization quotient. It's a metric that estimates body size after you corrected for body mass. And so the EQ is 0.49, in contrast to a modern opossum that has an EQ of 0.34. So this guy is brainy compared to modern opossums. Here is how we get it. Here is a CT scan surface model. And in the inside, we can see quite a bit of anatomical structure. With quite bit of struggle from my co-author Ted Macrini, we were able to extract this brain endocast. And you can see many features such as an olfactory bulb and a relatively large cerebral hemisphere. And not just a large cerebral hemisphere, it's really the side of the cerebral hemisphere homologous to the brain cortical area in modern mammals for sensory touch and motor coordination that got significantly bigger. So what does this mean? Well, let's pull back and do some comparison first. On the left is our distant relative from early Triassic called Thrinaxodon. It has been endocast. First we can tell that the relatively modern mammals have very distinctive olfactory bulbs. And the biggest difference is modern mammals have this large cerebral hemisphere. And modern mammals also have a cerebellar hemisphere that is absent in the distant Cynodonts. But when you add on top of one another, you can tell that this guy is not very brainy. And this guy is brainier. And so the question is how do we get from our Triassic ancestor, or ancestor-like condition, to the modern mammals? Well here is how we do it. We use the evolutionary tree as a road map. And we essentially map this transitional state one state at a time on the tree to see how it is distributed. And its distribution pattern will give us a possibility to infer how the process of evolution actually occurred. Now with the rise of a mammalial forms-- these are by jaw, by teeth, characterized as very mammal-like-- we start to have this divergence of the brain hemisphere. And once we get into modern mammals, and we start to have very distinctive mid-brain areas differentiated, it is a really in the modern Therian. Mammals that we start to have this cerebellar hemisphere. Let's focus on next this general anatomical region. In modern mammals, this general area of the brain, the neocortex, is responsible for sensory touch. And you can actually map these somatosensory fields. And so you have these representative animals. And the signal should come from a tactile sense of the hair. We know that the early mammals got hair. So that's not a worry. Here is a diagram-- I'm sorry this is washed out. But you've probably seen this multiple times. It's a cut out section of the brain in the cerebral hemisphere. And it represents the sensory touch of the surface. But if you look at this Morganucodon, you get a sense that the brain cast has become larger. It's not just become over-all large. It's really in this limited cortical area that a modern mammal responsible for somatosensory fields that becomes larger. So what this means is in the early evolution of mammals, we start to have a large olfactory bulbs, large cerebral cortical area responsible for tactile sense. And this is before we get to the modern mammals range of a brain size. Again this EQ is relative brain mass after being corrected against the body size. And in Hadrocodium, we start to have the relative brain size really within the range of modern mammal variation. And after that, the story is largely over about the brain evolution. So we talked about what we can do with the fossil from the Triassic and the Jurassic. Well let's forward relatively rapidly to talk about the split of modern groups-- modern groups of placentals and marsupials. Here is the fossil we call Juramaia. It's 160 million years. And I'm happy to report that we now already have a second fossil of this that we are working on. But before I go into that, let me go back to our breakfast egg. And here is the embryo. And the yellow part is the yolk sac. Much of it is allantoic sac. And here is the amniotic fluid sac. Essentially to live within the egg shell comfortably, we have to bring our groceries with us. We have to have a nice hot bath. And then not enough, we carry our porta potty with us inside the egg shell so we can live very comfortably. Now to take those homologous parts to understand the placenta of modern mammals. Marsupials largely have a yolk sac placenta. So this is a general feature of marsupial placentas. And this placenta is not terribly integrated into the endometrium of the uterus. And then placental mammals also got the placenta, but it's more sophisticated. It's from the amniotic part of the extraembryonic membrane. And this placenta structure deeply embedded into the uterus. Therefore we have much more effective parental care in utero. Now if we look at the broader pictures of placental diversity, no matter which group you are looking at, they are all having largely the same type of placenta. Marsupials are different. The parental care is much more limited within the placenta. And usually marsupials also have much shorter pregnancies. And after the precocial fetus are born, they are sucked onto the mother's nipple inside the pouch where they have the rest of the development. All the marsupials tend to have the same general placental design except the bandicoots, which are more like the modern placentals. So there is always exception. But all of these features are also associated with teeth, with jaws. And it's by really looking at the teeth and the jaws we are able to map some of the earlier fossils onto the overall placental and the marsupial trees. This also matters from another perspective. And that is we now understand mapping the genome or the complete genetic makeup. It is very important. So after 2005, multiple placental mammals have been mapped for their entire genome. And around 2006, at least two of the marsupials have also been mapped for the genome. And when you go down on the evolutionary tree, the first version of completed genome is already mapped for platypus in 2007. But eventually in order to compare different genomes, we have one roadmap. And that's our evolutionary tree. And our evolutionary tree, in a sense, is also our legacy. That's part of the reason why our interest in building this evolutionary tree. But not just with the molecules, but also with fossils too. Let me revisit an age-old issue about the divergence time of modern mammal groups. This is the best study that had most extensive sampling. I know you cannot read the details. In this daisy dial, there are 400 different families. But what's important is to bear in your mind-- and this black line is geological time scale. This is the present time. This is the middle Jurassic. This red circle represents the Cretaceous and the Paleogene boundary. That's the 65 million year event when dinosaurs got wiped out. When you map all this evolutionary lineage according to this study, you have many groups that go before the end of the Cretaceous. And here is the study by the NSF Tree of Life mammal project. The molecular team, they have sampled the largest concatenated molecular data set. And many of the groups would go back into the Cretaceous by this molecular study. Of course the morphological evidence do not always support such a claim, excepting of one case. And that's the split of marsupials and placentals. The molecular estimate of marsupial placental split goes roughly the middle Jurassic, roughly from 140 million to about 180 million years. And no matter which study you are looking at, it's roughly in this time window. The discovery of Juramaia definitely plays very closely with this molecular time estimate. However, for the rest of the late Cretaceous fossils, we still have a tremendous gap between the morphological estimate of major groups than the molecular time estimates. And here is one of the early fossils called Eomaia. This is now from Jurassic. This is from early Cretaceous. And this is a Sinodelphys. And this is punitively related to the marsupial clade-- at least, I think so-- when you map this fossil record onto the time scale and following evolutionary genealogical tree. About 30 years ago we were largely working with the fossils from this time interval. About 10 years ago, we start to have Eomaia Sinodelphys. And we map the fossil record to this level. And in the more recent time, we can definitely push the split between placental on one side and marsupial on the other down to this time interval, which will fall comfortably within the molecular estimate of these groups divergence. So it's good to have concordance, but we do not always have concordance in a sense that we still have this unresolvable issue about when the modern placental groups-- not all mammals, just placental groups-- start to appear on Earth. And with this kind of a fossil, we can also answer some ecological functional questions in a limited way. And here is a reconstruction of Sinodelphys. Let's focus on the wrist. And by looking at of the wrist, we can tell that up in the form closely related to marsupials, there are serving bonds that are far more robust than the form that are closely related to placentals. And it's also very interesting the shape of the claw and of the proportion of the fingers can also gave us an inference about how this animal actually moved around. Well let's back step a little bit to follow this line of discussion. And here is the modern opossum groups. And these are largely South American extant opossums. And we have [INAUDIBLE] possum. And that is arboreal. and we have Mouse opossum that is also arboreal. But we have Gray opossum that is terrestrial. And like our backyard Virginia possum, it's a large terrestrial, but it's also capable of climbing. And among the terrestrial opossums, we can have swimmers. Now it turns out all this habitat preference or niche specialization are also represented by the finger proportion. How so? So let's say here is [INAUDIBLE]. It's one of the opossums. And it's definitely known as ground living. And here is our familiar Virginia possum, Didelphis. It's largely ground living, but it's also capable of climbing. And these opossums are full time tree living. So in the specialized forms that stay more time on the tree because they have to grab and hold the finger proportions are different. Essentially the farther end of the fingers tend to be a longer. And the same pattern can be also detected in modern placental mammals. If you map the earliest marsupials and the earliest placentals from China, their finger proportion definitely falls onto the climber side than on the terrestrial side. All right so we get some level of inference about how they moved around. Now here it's a bit more complicated. Again, we are focusing on the hand bones called metacarpals. And we're focusing on the finger bones called proximal phalanx and the intermediate phalanx. So you map this bone along this axis. You map this bone on this axis. And then you map this bone on this axis. Essentially this plot shows you how, by finger proportion, you can allocate the habitat preference of this modern primate and the marsupials in their distribution. So that's how we understand . We can use the finger proportion to distinguish different habitat preference. Let's bring the fossils into this discussion. And this Triconodon's multituberculates. These are extinct fossil groups. These lived from Jurassic to Cretaceous. The majority of these, so far as we know, they're more like terrestrial mammals. But by the time we get to a near-mammal relative called [INAUDIBLE]. They're also more likely-- or the majority known so far-- are also terrestrial. But by the time we get to the modern placental relatives, marsupial relatives, they are definite closer to the tree living primates and tree living marsupials. So associated with the rise of our modern group-- that is placental on one side, marsupial on the other-- there is definitely a shift on locomotory adaptation and the habitat preference. Essentially somewhere around the rise of modern mammal group, of Therian, we got into the tree. Or we had a better capacity to get into the tree. Now this specialization is not isolated. We used to have the stereotype before the Cretaceous Paleogene extinction, somehow we have very limited diversity of modern mammals. And we also have a just-so story to go with it because these early mammals lived in the time of dinosaur. They shared the same ecosystem with the dinosaur. And suppressed by the dinosaur, they just failed to gain the great ecological possibility that it can realize until the dinosaurs are all gone. So that's what we understood then. But this is what we understand now. Around 2006, we start to have semi-aquatic forms discovered in the fossil interpreted from some of the older fossil records. And the start from 2001 to 2005, we get large enough of a mammal. We know for a fact it can eat a dinosaur, but it's generally capable of carnivory. And that was considered unlikely before the recent time. And in 2005, we discovered this form called [INAUDIBLE]. And it has a very unusual specialization otherwise only known in aardvarks and armadillos. That is they have these peculiar teeth that's usually associated with tongue feeding of colonial insects. Highly specialized, it used to be considered as only possible with extant mammals. Yet in a totally unrelated Mesozoic group, we found the same pattern. Of course we have, now, multiple forms that can climb the tree. And once you can climb the tree, it's only a jump away from gliding. This is what we used to know. And this is what we now know for our last 20 years of hard working by paleontologists going to all corners of the world finding the best fossil you can. And also in the case of China, the great ingenuity of all these peasants tried to do better for themselves, dug out all these great fossils. So what our Jurassic ancestor can tell us? We can tell that they are not just a bunch of Mesozoic road kills. They are flattened fossils. But they can definitely tell us about how the integumentary structure of mammals that characterized our whole group arise in the fossil record. And they can definitely tell us about the general growth pattern. We can, with confidence, tie to the most fundamental mammalian adaptation. That's mammary glands and lactation. And we can also tell the general pattern how in the earliest phase of mammalian evolution, we start to have a larger brain. And we can now also tell that no later than middle Jurassic, we start to have the three main branches of extant mammal lineages. And that's placentals, marsupials, and the monotreme. And the rise of the Therian mammals, the placentals and the marsupials, is definitely accompanied by some very interesting ecological diversification. It's more than just luck. It's really with good fingers that we manage to hang on to this next episode of evolutionary diversification in Jurassic. Thank you.