As an educator at the American Museum of Natural History, I knew enough about paleontology to tell people that there were no known aquatic dinosaurs. To many peoples’ surprise plesiosaurs, icthyosaurs, mosasaurs (and all the other fossil marine reptiles so often depicted in books and TV shows) – none of them were actually dinosaurs. They were very different types of reptiles, extinct sauropterygians and icthyopterygians (mosasaurs were actually a kind of lizard) that lived in the seas during the Mesozoic Era while dinosaurs ruled the land.
Jurassic Park 3 came out in 2001, which featured a Spinosaurus that swam like a crocodile, and all the people were like, “See! Dinosaurs COULD swim!” But I never said dinosaurs couldn’t swim. There has long been fossil evidence – in the form of trackways that resemble scraped claws against river floors – that dinosaurs probably were pretty good swimmers, and waterways wouldn’t have presented much of a dispersal barrier, just like caribou and other large modern migrating animals who eventually encounter water at some point in their life and have to cross it. Still, there were no fossil remains ever described as belonging to a dinosaur that actually lived in the water.
Since it’s discovery over 100 years ago, Spinosaurus has been known as one of the largest theropods. Theropods are a large group of dinosaurs which includes not only all birds but all the familiar carnivorous species, including Allosaurus and Velociraptor. Spinosaurus was bigger than even the biggest Tyrannosaurus. This week, the journal Science published the paper “Semiaquatic Adaptations in a Giant Predatory Dinosaur“, in which a group of scientists make the case for a long list of adaptations supporting the idea that this huge meat-eating dinosaur with a sail on its back was also well adapted to life in the water.
Finally, a dinosaur that lived in the water and swam and ate fish like a crocodile. I have been waiting my whole life for there to be one, and here it was under our noses the whole time. Good ol’ Spinosaurus. Museum educators will have to adjust their curriculum accordingly.
This is the second post in my series on my favorite examples of convergent evolution, a concept which I defined previously and demonstrated with marsupial and placental wolves. That post featured two modern terrestrial mammalian species with relatively recent evolutionary histories. For the current example, I would like to take us into the oceans and examine a much wider range of organisms as disparate as cartilaginous fish, bony fish, reptiles and mammals — spanning about 530,000,000 years of Earth’s history. Within each of these highly divergent lineages, a particular adaptation for high-speed underwater movement evolved, known as thunniform locomotion – which means literally “swimming like a tuna”.
The thunniform body plan consists of a suite of traits that makes an animal a streamlined swimming apparatus: a rigid anterior section (from the head to about the pelvic region) that becomes much more flexible posteriorly, with a crescent-shaped caudal fin (a.k.a tail fin or fluke). A thunniform swimmer need only to undulate its caudal region (meaning basically its tail) and the result is efficient propulsion through the watery medium with minimal head-wiggling (see Figure1). Thus, the swimmer can disperse over long oceanic distances at high speeds, and keep its eyes toward whatever its pursuing. That this has evolved in so many different types of top marine predators over the eons strongly suggests that the common environment of the ocean presents strong selective pressure for an optimal form of locomotion. In sharks and tuna, thunniform locomotion represents more of a parallel pattern of evolution, which is is when two separate lineages evolve into similar forms independently but had mostly similar starting points (the common ancestor of sharks and tuna was already aquatic). It is even more remarkable, in my opinion, that it developed in both reptiles and mammals, since the common ancestor of these groups lived on land. The convergent evolution of this highly derived body shape thus required totally independent exoduses of terrestrial animals back into the ocean.
Among the many kinds of reptiles that inhabited the world’s oceans during the Mesozoic Era, there were none so well-suited to marine life as the ichthyosaurs (“fish lizards”). The ichthyosaurs are the most highly derived group within the reptilian superorder Ichthyopterygia, which were ocean-going reptiles who had developed their characteristic “fish flippers” very early on in their history. It is generally accepted that the ancestors of the ichthyopterygians were land-dwelling amniotes, however the precise taxonomy of their relationships to other reptiles has been highly contentious and was hotly debated throughout the 20th century. This is because the fossil record is devoid of any transitional forms that could be seen as terrestrial precursors – the oldest ichthyopterygian fossils, which date to the Early Triassic ~250,000,000 years ago, are already so “ichthyosaurian” – and it is extremely difficult to retrace the evolutionary steps taken towards their aquatic lifestyle. However, most modern cladistic analyses indicate that ichthyopterygians were diapsid reptiles along with lepiodsaurs (snakes, lizards and tuatara) and archosaurs (crocodilians, pterosaurs, and avian and non-avian dinosaurs) (Motani et al. 1998).
Soon after reptiles started dominating terrestrial environments during the early Mesozoic, they began to conquer the sea as well. But by about 90,000,000 years ago, the last ichthyosaurs died out. Regardless of their ancestry and the length of time they have been extinct, the ichthyosaurs just might represent my favorite historical example of convergent evolution. Figure 2 shows a skeletal and a life restoration of the iconic Ichthyosaurus communis and demonstrates just how well-adapted these animals were to marine life. The overall shape of the body is reminiscent of a streamlined torpedo. The hands and fingers on its front limbs became pectoral fins; other fins included a dorsal fin that is more obvious in fossil impressions as well as a crecsent-shaped caudal fin that makes the tail look very shark-like. You can see how lateral (side-to-side) undulation of its rear end would thrust the animal forward through the water with minimal effort. I included a dolphin skeleton in the figure to highlight how precise the convergence between these two highly divergent vertebrates was. They do undulate on different axes; dolphins thrust their tails up-and-down, not side-to-side, but that seems to be a strongly conserved trait among mammalian swimmers and I think it doesn’t take away from the overall similarities. The evolution of whales and dolphins from their terrestrial ancestors that began ~55 million years ago is another amazing evolutionary story that I won’t get into here. What I wanted to highlight in this post is how closely these air-breathing vertebrates came to resemble one another.
A very good overview of the various forms of undulatory locomotion, including thunniform, can be found here.
One of my favorite evolutionary phenomena to talk (and write) about is convergent evolution. Convergence has occurred whenever phylogenetically distant species, in either or both disparate geographic regions or geological epochs, have evolved very similar phenotypes. This is usually due to similar ecological conditions across the times and places in which the species reside, and in response to similar selective pressures, the different species have in common parallel yet independently evolved adaptations (this concept is illustrated in Figure 1). These similarities are not due to common ancestry and are therefore not homologous. It seems that this would be rare pattern, however the tree of life is full of amazing examples of convergence, and the prevalence of convergence throughout Earth’s history suggests there is at least some degree of determinism in the evolutionary process. I want to write a series of posts which highlight some of my favorite examples of convergent evolution. These posts will try to frame each case in evolutionary, biogeographical, anatomical, and environmental terms.
My first case study in convergent evolution will be the wonderful story of two predatory mammals, which lived at approximately the same time but in separate corners of the globe. One of these is extinct and the other… well it’s complicated. During the Miocene Era which lasted between ~23,000,000 and ~5,200,000 years ago, an environmental change occurred which allowed vast grasslands to open across Eurasia, and with that, the development of a diverse community of crown-toothed herbivorous perissodactyls (“odd-toed” ungulates which include modern-day horses and rhinos) and artiodactyls (“even-toed” ungulates which include modern day cows, sheep and camels) which grazed and browsed across the new prairies. Not surprisingly, there were carnivores which became adapted to chasing down this quick quarry, including canids (the dog family, which includes foxes, jackals and wolves). Most folks are familiar with wolves. Modern wolves (Canis lupus) likely appeared during the early to middle Pleistocene (between ~2,500,000 and 1,000,000 years ago), and are slender, powerful predators adapted to pursuing similarly-sized prey over long distances. They have deep chests and long legs, reminiscent of distance runners, and bone-crunching jaws with pointy canines well-suited for holding onto prey (say, for instance, a fleeing deer), as well as scissor-like carnassial teeth for meat-slicing. No wonder wolves have inhabited a special place in the human psyche – with results that were often bad for wolves. I love wolves, and as a teenager I was mesmerized by reading about their natural history as well as accounts of human-wolf interactions (which often bordered on the fictional), such as “Lobo the King of Currumpaw” by Ernest Thompson Seton in Wild Animals I Have Known (1898) (Figure 2), and especially that of “Thinking Like a Mountain” by Aldo Leopold in A Sand County Almanac (1949), where Leopold recalls the moment he changed from wolf-hater/hunter to one of the pioneers of the environmental and conservation movements. If you ask me, a better appeal for whole-ecosystem conservation has never been written:
We reached the old wolf in time to watch a fierce green fire dying in her eyes. I realized then, and have known ever since, that there was something new to me in those eyes – something known only to her and to the mountain. I was young then, and full of trigger-itch; I thought that because fewer wolves meant more deer, that no wolves would mean hunters’ paradise. But after seeing the green fire die, I sensed that neither the wolf nor the mountain agreed with such a view.
Let’s turn back the clock and take a trip to the the land down under, where after the breakup of the supercontinent Gondwana during the Jurassic Period ~200,000,000 years ago, the resident life forms were stranded at sea on an island landmass known as Australia. Placental mammals had not yet dispersed globally, and the only members of the furred kind in Australia at the time of the continental breakup were the pouched marsupials – and it remained that way until very recently (when humans brought their rats and sheep). During the Miocene, as in Eurasia, grasslands began to open up across Australia, although many regions were still heavily forested. This created many new and unique niches that allowed the resident marsupials to adapt and radiate into a wide variety of forms, such as diprotodonts (which include arboreal possums, familiar and bounding kangaroos, and the extinct large lumbering herbivore Diporotodon), and dasyurids which include the Tasmanian devil and Tasmanian wolf (Figure 3). Wolf? Yes, there is a marsupial wolf. How can that be? The answer is: CONVERGENCE. The scientific name of the Tasmanian wolf is Thylacinus cynocephalus, which means “pouched animal with the head of a dog”. Another more unique and accurate common name is the thylacine.
Thylacines appeared in Australia during the Miocene, and the modern species was present in both Australia and Tasmania by the Pleistocene and into the Holocene, although it was just about absent from mainland Australia by the 1800s and restricted to Tasmania by the 20th century. Just as wolves across Eurasia and North America, thylacines were looked upon as threats to ranchers and were accused of killing chickens and sheep. As a result, they were hunted to extinction and the last one died in a zoo in the Hobart Zoo in Tasmania in 1938. The thylacine was a dog-sized predator, and while its method of prey acquisition is still under debate (whether it was a pursuit predator like wolves or more of an ambush predator like a cougar), together with canids it represents an exceptional case of convergent evolution, which is most evident in a comparison of the cranial anatomy with C. lupus (Figure 4). Both species have an elongated snout with pointed canines and scissor-like molars which are well suited for snapping up, killing and dismembering prey. You’d be hard-pressed to pick one or the other out of a line-up. One of the few cranial traits that can be used as a clue is that the thylacine has more teeth overall, including triangular-shaped molars, thus betraying its marsupial ancestry. But you’d have to be an expert to notice that. While they lived on opposite ends of the world, placental canids and marsupial thylacines developed into similar ecological roles and responded with a convergent evolution toward extremely similar phenotypes.
As mentioned above, while thylacines and wolves evolved into similar forms, they also were met with similar antagonism by humans. Alas, the thylacine died out before the conservation movement took hold. With their much wider distribution across the Northern Hemisphere, placental wolves were able to hold in many areas and were more recently aided by human efforts to bolster their recovery. As Aldo Leopold predicted, the reintroduction of wolves into Yellowstone National Park was a major triumph for conservation, as it returned a keystone species to an ecosystem which was out of equilibrium due to its absence. The return of the thylacine would require either a seance or de-extinctioning through DNA cloning. The probabilities of either of those producing a living thylacine anytime soon are close to zero.
It’s a good week for snake genomics, because PNAS has published both the Burmese python genome (Castoe et al. 2013) and the king cobra genome (Vonk et al. 2013). The related papers come from separate research teams (the python people mostly from Colorado and the cobra cabal the Netherlands), albeit with significant overlap between them. The world of snake molecular biology is a small one, after all.
The python group planted its snake genome flag in the ground more than two years ago with a paper describing their first draft assembly, and I have been eagerly awaiting the results of their full-blown analysis. It was well worth the wait. As the python is known for its feast-or-famine metabolism (the small intestine can grow up to three times in size after gulping a meal half the python’s body mass), the researchers provided a very elegant analysis of the differential expression of genes in digestive organs before and after a meal and show, basically, that different genes are turned on or off either pre- or post-feeding. Very cool.
The authors of the python paper also investigated how snakes evolved their iconic and constrained morphology – because, after all, snakes lack legs and are equipped with a feeding apparatus equivalent to a human being swallowing a 16lb Thanksgiving turkey whole, and the molecular bases of these adaptations are unresolved. They analyzed a large number of genes shared across vertebrates – called orthologs – and found that during the evolution of the vertebrate lineage leading to snakes, many genes associated with skull and spinal development, metabolism and other functions experienced a faster rate of evolution than in other lineages. Also very cool.
Next up: king cobra. Why do we need both a python AND a cobra genome? The answer lies in the difference between these two snakes. One of them – and I hope you guessed cobra – is venomous. So it is no surprise that much of the justification for sequencing the king cobra genome included a need to understand the evolutionary origins and maintenance of the genes that control venom production. The cobra genome contains a multitude of protein families that underwent a significant expansion during cobra evolution that resulted in what we see today – a highly potent mixture of toxins designed to ensure certain death to a chosen prey item.
The two genomes differ vastly in their qualities of assembly. Through a mixture of various sequencing methods, the python team was able to get an N50 value of 207kb, meaning 50% of the assembled chunks of contiguous sequence were at least 207,000 base pairs of DNA in length. That assures that the research team would be able to recover the majority of genes – exons and introns and all. The cobra genome by contrast has an N50 value of only 3,982 base pairs, meaning that some of the genes may be fragmentary and the length of many introns will remain unresolved. However, I think that using N50 as a the gold standard of genome assembly “quality” is misleading. Sequencing strategies that significantly raise N50 values cost more money. In this day and age of modern biology, where small labs or groups of researchers conjure up whatever resources they can for an in-house genome sequencing project, the most affordable strategy for however you wish to address your biological questions will probably suffice. Both of these snake genome papers make the cut in that regard, and they are a significant contribution to the field of reptilian genomics.
- Python Genome Helps Explain Snake’s Extreme Eating Ability – Huffington Post (huffingtonpost.com)
- How Snakes Got Their Extreme Makeovers (news.sciencemag.org)
The “Anolisphere” is all abuzz with a recent paper by Nicholson et al. published in the journal Zootaxa with the inviting title, “It is Time for a New Classification of Anoles”. The authors propose two important and potentially disruptive changes to some widely-accepted features of the study of Anolis lizards:
1) The currently recognized genus Anolis is actually comprised of eight ancient lineages that are so divergent it is justified to break Anolis into eight genera (Dactyloa, Deiroptyx, Chamaelinorops, Anolis, Xiphosurus, Ctenonotus, Audantia, and Norops).
2) The widely accepted “ecomorph” hypothesis, which states that several independent lineages of Anolis lizards underwent convergent adaptive radiations on Caribbean islands (I wrote about this in a previous post), is not supported by the evidence. The authors therefore suggest a looser-fitting “ecomode”model, which simply describes similar ecological adaptations without having to invoke any deterministic process.
Jonathan Losos of Harvard University has referred to this paper as “undoubtedly the most important paper on anoles to be published in the last several years”. Nonetheless, these intriguing ideas will require long and heated debate before they are to be accepted by the scientific community as a whole. The editors of the anole-themed blog Anole Annals have already staked out a position as being strongly opposed to the changes. It has already been suggested that the Anolis community need not be compelled to formally adopt the new classification, and that it would be disruptive and confusing for future researchers, given the long body of work in which Anolis is referred to as a single large genus. There are also mounted defenses of the new paper as well. Check out what’s happening on the blog this week.
In my opinion, we should keep the single-genus status of Anolis, and here’s why. Pretty much everyone in evolutionary biology agrees that Linnean classifications (the binomial system which assigns genus and species names, as well as higher order classifications such as family, class, phylum, etc.) should reflect the monophyly of groups. This means that if you are describing a new genus, or family, or class, the number of lineages nested within the new group is largely subjective and the only real criterion is that all the lineages have descended from a common ancestor.
When basing taxonomy on the estimation of phylogenetic trees (which is pretty much how it’s done these days since tree building is how one establishes monophyly), your taxonomies are only going to be as good as the trees you use. Granted, the tree in the Nicholson et al. paper has strong statistical support, but that only really means that it does a good job of describing the data used to construct it in the first place. If down the line there is more or better data, a new tree may have better support, and then new taxonomies will have to be proposed. Since our understanding of taxonomy is dependent on phylogenetic information that is subject to change, and since we can say with certainty that Anolis in the broad sense represents a monphyletic group, AND much of the research out there treats it as such, I think we’re better off just keeping Anolis the way it is. But up to this point, I’m a consumer of systematics and not an authority on this by any means.
The question remains, can a single genus be ~130 million years old? And the answer is absolutely yes. While our genus, Homo, is only just over 2 million years old, Ginkgo is probably about 200 million years old. And they’re both considered genera! My point is that Linnean classification, while useful, really serves a purpose in the sense that someone knows what museum drawer to put these things in. And I think Anolis should be one drawer (with almost 400 species, that’s a big drawer!).
In any case, under the newly proposed taxonomy, the species I study, A. carolinensis, will remain in Anolis. So I have little to lose, other than a divorce from hundreds of other fascinating possibly former congeners.
Our paper on green anole phylolgeography has been published, so I thought I would justify the study in the first place and briefly synopsize our major findings and their implications. There’s a lot to talk about, so it will be completed across two posts.
First, some background….
As you may know, Anolis carolinensis is the scientific name for the green anole, which is a smallish lizard that lives in the southeastern United States. Last year, its complete genome sequence was published. At the time, it was the only reptile to have a fully sequenced genome (although this list is ever growing) and the phylogenetic gap it filled among sequenced vertebrates created a real demand. In addition, Anolis carolinensis is the lone species — out of a genus of around 400 — which occurs naturally in North America, with the bulk of Anolis biodiversity situated in South America and the islands of the Caribbean. On these islands, several independent yet convergent adaptive radiations led to the evolution of the same “ecomorphs” on different islands (a nice blog-post-sized review can be found on this Map of Life page). It’s a lovely story.
What does the Anolis genome offer? The chance to understand the genetic basis of adaptation (discussed by the eminent evolutionary biologist Jonathan Losos in this blog post on the Anole Annals), which is a very exciting prospect indeed. Using the genome as a resource, biologists may be able to pinpoint exactly which parts of the genome are affected by natural selection as new species form, or as different populations adapt to new surroundings. On another note, our lab wanted to look at how natural selection shapes the structure of a genome by limiting the activity of transposable elements, and the Anolis genome offered a unique opportunity to address this important question in comparative genomics. With regard to A. carolinensis specifically, it lives in a wide variety of habitats from subtropical Florida (FL) to more temperate Tennessee (TN), where anoles are subjected to freezing winters. The genome opens considerable opportunities for investigators to learn the genetic basis of each population’s unique set of adaptations across this landscape.
There was one problem, and it is familiar to anyone who studies population genetics (an esoteric bunch, indeed, however this is a major field in biology): if you want to know how much of the genetic variation in a population is affected by natural selection, you have to first have an estimate of the total amount of expected genetic variation (a parameter known as θ, or Theta). One thing that significantly affects θ is population size (usually denoted as N; the other thing affecting θ is the mutation rate u, so that θ=4Nu): a large population will have more genetic variation. But in order to measure the size of the population, you need to know its structure. What does that mean? Well, in a wide-ranging species (like A. carolinensis), it is very unlikely that individuals living in, say North Carolina are able to mate and share genes with those from Texas. Through a process called genetic drift (which is the random changing of gene frequencies due to finite numbers), these separated populations will resemble each other less and less over time. Understanding structure allows biologists to measure the amount of gene flow across landscapes, which can significantly alter local population sizes. Once you have an idea of the population structure, only then can you estimate the population size and measure natural selection at the genetic level.
In addition to estimating population sizes, looking at how genetic variation is distributed geographically allows researchers to make inferences about the evolutionary history of a species. For instance, if you were to observe that certain fixed genetic differences occur on either side of a large river, it may be due to the fact that the river provides a dispersal barrier. If you know how old the river is, you may be able to estimate how long the populations have been separated. Applying these methods to the understanding of the history and formation of species is a field in biology known as phylogeography. Phylogeography had its origins in the late 1970s, and as it became more practical to study genetic variation (especially after the advent of PCR and DNA sequencing), it exploded in the 1990s and 2000s, and now has at least one technical journal (aptly titled Molecular Ecology) solely dedicated to publishing studies using its methods. The phylogeography of a species is of special importance with regard to θ, because if the history of the population includes exponential growth (as it would when a receding glacier reveals virgin habitat) it will skew the θ estimates downward.
What was known about the phylogeography of green anoles? Well, not much. It was established that A. carolinensis arrived from Cuba sometime near the Pliocene-Pleistocene boundary about ~3 million years ago — an “Out of Cuba” hypothesis. The last study that looked at genetic variation in the species was Wade and Echternacht (1983), which relied on the differential electrophoretic movements of proteins called allozymes from anoles collected at seven localities. The major findings were: (1) 17 out of 25 allozymes completely lacked variation and (2) South FL was the home of the most divergent anole population. I like this paper, but it didn’t delve deep enough to make any robust inferences about green anole evolutionary history. This is nothing against the authors, it’s just that the methodology and technology weren’t strong enough yet.
Since then, as major advancements were made in the application of robust evolutionary models to more easily attainable DNA sequences for phylogeographic inference, the green anole escaped the scrutiny of phylogeographers, even as more and more co-distributed taxa were looked upon. It seems as if this widespread, abundant and even iconic lizard had become completely overlooked. It wasn’t as if samples were hard to come by. In fact, if you were to search HerpNet for museum samples of green anoles, you would receive 9,548 records! We had done our own fieldwork as well, and several questions were left wide open for us to address: (1) what is the structure of green anole populations across its range and just how divergent are the major lineages; (2) are the often-cited common dispersal barriers associated with co-distributed taxa also correlated with the distribution of genetic variation in green anoles; and (3) how did the glacial history during the most recent ice ages effect green anoles living at higher latitudes? So we decided to give it a try.
I’ll talk about the paper in Part 2.
Tennessee Governor Bill Haslam (R) recently failed to block a law passed by his state legislature that provides legal protections for science teachers who criticize scientifically tested principles such as evolution and global warming in their classrooms. I’m sure there exists across the internet numerous angry rants about this as an assault against scientific reasoning and the proper education of our citizenry, and I fall heavily on the side of that argument, but I won’t get into it here.
I will chime in with one thing, however.
The bill states “some scientific subjects, including, but not limited to, biological evolution, the chemical origins of life, global warming, and human cloning, can cause controversy”. What this “teach the controversy” argument really means is “create controversy”, because the scientific subjects listed above are like comparing apples and oranges. Evolution and global warming are hypotheses that have not yet been falsified and thus are considered theories. This means that for both, there has to date been almost no evidence leading the scientific community to believe an alternative explanation. The consensus among scientists is: life evolves and the Earth’s climate is getting warmer. There really is no controversy, unless you ask certain non-scientists who argue against these theories on moral grounds.
That is like taking your car to the mechanic, and he/she says: “I looked at your car and determined that the reason your steering wheel is drifting is because your tires are unevenly worn”. Then, you take your car to your uncle and he says “ah, the roads are just crooked”. Who would you believe: the person with the expertise and training to examine the situation and make a determination as to the cause of his/her observations, or the non-expert who is just telling you what he thinks?
Furthermore, one may believe that the act of cloning a human is immoral. Fine. But human cloning is not a theory, it is an act, so one is allowed to make a logical argument against its practice. To argue that scientific theory is immoral just doesn’t make any sense. Are Tennessee teachers going to tell students that it’s okay to think that scientists should maybe “lay off the evolution thing”, as in “HEY, STOP DOING EVOLUTION”? Scientists would respond, “uh, we’re not doing anything, as far as we can tell the things are just evolving on their own, we are just watching”.
The reasoning behind it is suspect. Which makes me worry that powers greater than you or me or the average Tennessee parent were at hand. Which, in turn, makes me fear for the future of our republic.
- Anti-Evolution ‘Monkey Bill’ In Tennessee (1oneday.wordpress.com)
- Tennessee Passes ‘Monkey Bill’ To Teach The ‘Controversy’ On Evolution And Climate Science – Brad Johnson – ThinkProgress (richarddawkins.net)
- A Question for Education ‘Reformers’ About Tennessee’s “Monkey Bill” (mikethemadbiologist.com)