Speciation is the process through which species diverge from each other and/or from a common ancestor. There are several definitions of species, most of which focus on reproductive isolation and/or phylogenetic similarities. This can cause some controversy. Speciation can result from geographical separation or ecological specialization. There are stages of speciation in which organisms cluster first into distinct populations before finally becoming different species.
Stearns, Stephen C. and Rolf Hoekstra. Evolution: An Introduction, chapter 12
February 13, 2009
Professor Stephen Stearns: Okay, today I have put up a slide from the first lecture, just as a signpost and landmark to make sure that you're oriented on the path through the course. Basically the material we cover before Spring Break is organized into these large sections, and today we're making the transition from microevolution to macroevolution. And you'll recall that in microevolution we had six lectures on microevolutionary principles, and then we had five lectures, that we just completed last time, on how phenotypes, organisms, how they're designed for reproductive success. We're now going to make the transition into macroevolutionary principles, with the process that connects microevolution to macroevolution, and that is speciation.
So as we now move into macroevolution, today's talk is really mostly about speciation. Then we will, next time, have a talk about phylogenetics and systematics, and then about how you can use the combination of these principles to look at evolutionary trees and to place them on maps, and then to place the evolution of traits onto trees and onto maps, so that you can integrate space and time, and history.
Then we have three ways of looking at the history of life. The first is more abstract, I think a little bit deeper, and that has to do with key events in evolution: origin of life, multi-cellularity, language, things like that. We'll have a review of major events in the geological theater. For those of you that like firecrackers on the Fourth of July that's fun, because that's where we have meteorite impacts and extinctions and stuff like that. And then we'll take a look at the fossil record and what it- the kinds of unique insights that you can get out of fossils that you really can't get out of just looking at living organisms. So those are three different ways to approach the issue of the history of life.
Okay, today speciation. I would hope that by the end of today's lecture and the end of your reading and discussion in section about this, that you would be able to deal with this set of issues. What is a species? How do they originate? What kinds of experimental evidence or observational evidence do we have to back up the claims about how they originate? What's going on during speciation, from the point of view of genetics; if we track the genes through the process of speciation, what do we see? And then there are some special issues: asexual species and cryptic species.
I would like to signal that this is a part of biology in which some of your teaching fellows have special expertise. So if you would like to get to the cutting edge of phylogenetic research, you have got some people in section who can help you, and this department happens to be particularly good at that. There was a front page report in the New York- well on the Science Page of the New York Times, in their Darwin Section, they were reporting on Stephen Smith's Ph.D. thesis. He just got his degree here last year. But he assembled the largest Tree of Life for the plants. So we've got people like that running around, and you ought to tap into them, if you feel the urge. Okay?
Now here is a basic observation, and it is one of those things that just seems so natural to us that it doesn't occur at first sight to ask how it might be that way. But in fact it is a puzzle. We have wild diversity in the world. Please forgive me botanists. I had some help assembling this slide. I would like to have had a few more neat plants in there; there are some incredibly neat plants that could go in there. But if we just look at the world, it's amazingly diverse. Okay? So we look out there and there's lots of different kinds of organisms, and they seem to be in discrete groups, separated from other such groups, and in many ways: appearance, behavior, ecology, genetics.
So why is the world like this? How did that come about? Why is the planet not simply covered with a homogenous layer of primordial slime, rather than these different things? For a long time it wasn't that way really. The prokaryotic world was genetically diverse, but it certainly didn't have this kind of morphological diversity in it. Well those things that are separated are called species, and this is the guy who profoundly influenced this kind of biology.
His name's Ernst Mayr. Ernst died a few years ago at the age of 100-and-something. So he managed to keep active for a very long time. I knew him. He was an impressive guy and he had an almost unprecedented ability to claim that he had already said it.
Now his definition of what a species is--this is from a version of a book that he wrote in 1963, but he had been working on this idea from about 1942, and his idea was that a species is a group of actually or potentially inter-breeding natural populations that are reproductively isolated from other such groups. So it's the capacity for reproducing together that unites the members of a species, and it's the inability or the lack of opportunity to reproduce with individuals of other populations that marks the separation between species; according to Mayr.
Now there were very good reasons for this definition, and it really captures the essence of genetic separation. But there are cases where it doesn't apply. As a matter of fact, I think that biology has a profound capacity to erode the categories invented by humans. [Chuckles]
There are almost always exceptions. To that, I would like to add that Ernst Mayr was a great ornithologist, and that the kinds of patterns you see in species, in birds, were what motivated him to come up with this definition, and if he had been a plant biologist he probably wouldn't have defined things this way. Okay?
Now just something to think about. Does everything really have to belong to a species? Would there be some itch that you needed to be scratched if not everything belonged to a species? Would that make you unhappy for some reason? I'm going to let that one hang in the air.
Okay, now when we look out there and we see that things are separated, there are some pretty straightforward mechanisms that actually separate things, and so this is part of the vocabulary building exercise for speciation science; and it's also a little bit of biology. There are isolating mechanisms that are common, they fall into three categories, and they look a little different in plants and animals.
The first category is pre-mating or pre-zygotic isolation, and in plants that's often determined--not always but often--by pollinator behavior and flowering times, and in animals it is determined by mating behavior and gamete release. So here we have actually a place where sexual selection, from the last lecture, connects into speciation, because who you choose as your mate is going to determine whether there's an opportunity for inter-specific hybridization or not.
So this is actually an issue about mate choice, which is kind of at a higher hierarchical level than the issue of should I choose my mate for resources or good genes or sexy sons or something like that? Should I choose a mate, because in mating with that individual I will form hybrids or I will form members of my own species? In a minute you'll see some pollinator behavior slides; so just to anticipate that.
It is quite possible for plants to be separated just by flowering time. Imagine a valley that's running east-west, in a temperate area of the planet, and it is getting sun earlier on one slope than on the other, simply because of the angle of the sun to the earth. Everything's going to speed up on the warm slope and slow down on the cold slope, and you're going to get different flowering times for purely biophysical reasons, on the two slopes. So initially a plant population might come in after maybe the glacier has melted off and colonize those two slopes, and they wouldn't be able to avoid having different flowering times, and therefore starting to separate genetically on the two slopes. So there are all kinds of reasons for separation to start.
Now the second class of isolating mechanisms occurs a bit later; it's post-pollination or after insemination. And in plants there are--and actually in many algae and ciliates--there are self-incompatibility mechanisms. These are actually put in place through recognition molecules. And there is an interesting--gamete biology is very interesting. There's a lot of mate choice which is going on at the level of the sperm finding the egg and the egg sitting there and deciding, "Do I want this one or not?" Throw that one away and take another one. And these are mediated by well understood enzymatic reactions and protein structures; bindin and lysine. The self-incompatibility argument in plants basically has to do with the costs of inbreeding, and the costs of selfing.
And the arena in which that takes place is after the pollen grain lands on the style of the flower and starts to grow down the style, towards the ovary of the flower, the tissues that are in the style are checking out that particular pollen grain, which is making basically a little plant-like thing, as it grows down towards the eggs, and that is where the self-recognition is implemented. If its pollen is coming from that same plant, it'll get killed at that point.
So there are post-pollination and post-insemination mechanisms of isolation. Now after fertilization occurs, post-zygotic isolating mechanisms basically have to do with viability, survival and fertility. So hybrid inviability or infertility is a common thing, once two species have diverged to a certain extent.
And of course the commonest one that you're probably all familiar with is the mule; the mule is a cross between a donkey and a horse, and mules are sterile. So hybridization between donkeys and horses isn't going to go anywhere evolutionarily because there are never any grandchildren. I believe the same is true for ligers, crosses between lions and tigers. I'm getting nods from the zoological aficionados in the audience. So thank you for the confirmation. [In fact female ligers do breed successfully with male lions. Male ligers are sterile.]
However, at this stage, often fairly distantly related plants are able to hybridize. It's usually not across the boundaries of genera or families, but sometimes you can get plants from two genera hybridizing with each other, which may be an indication that the classification has been wrong. But plants generally hybridize a lot more easily than animals do, and it may have something to do with the differences in their developmental complexity. Usually the reasons for hybrid inviability in animals are developmental abnormalities.
Okay, so I had signaled that we were going to look at a pre-zygotic example. So here is species separation by pollinator recognition. Here are two columbine species, and they can overlap geographically. You'll notice that they really have very different colors, and that is actually related to the brains of the things that pollinate them. Yes, moths do have brains. Okay? So hummingbirds are notorious among pollination biologists for loving red and yellow, and if you see a white flower, you can normally--it's a pretty good bet that it's going to be pollinated by a moth or a bat.
And in fact if you were to look at these things in the ultraviolet, you would probably see that there were, in the ultraviolet, interesting markings on them, and often flowers that have ultraviolet markings on them are bee pollinated, because bees see particularly well in the ultraviolet. So these are two species that are separated by pollinator behavior, and I invite you to play through a scenario in your mind of how their ancestor had the same color and then how they might have gotten different. Okay? It would've been a gradual step-wise process, as they specialized on a particular pollinator.
Okay, so species can be separated by various mechanisms and at various stages in the process of reproduction. Mayr gave us a biological species concept; that's the biological species concept, isolation. And there are plenty of others now. And I've put this list up here, not for you to memorize, but more or less--I do invite you to go through and sample these species concepts--but to show you that actually to try to state logically what we mean when we say the word species has been a difficult task for phylogenetic biologists to deal with, and the fact that I can put a list like this up there is a measure of the difficulty, and a measure of the controversy and the disagreement that has occurred over this issue.
What each of these definitions does is try to define a species from a slightly different biological starting point. And what they're trying to do essentially is get the most general and useful definition that they can. And in some way they all fail, which is kind of neat, because it means there's still work to be done, but in some way they're also all partially correct.
So I told you about Mayr's concept. The recognition concept basically is based on mating systems. So species are things that will decide to mate with each other; that's Hugh Patterson's. The phylogenetic species concept--there are various ways of looking at the Tree of Life and deciding what on it constitutes a species. So Joel Cracraft said it's a cluster or organisms that are distinct from other such clusters, within which there's a pattern of ancestry or descent. So he says well essentially it's the stuff that's out there at the tip, that's closely associated.
Back in '78 Wiley said that a species is a lineage of ancestor-descent populations that maintains its identity from other such lineages and has its own evolutionary tendencies and historical fate. I don't know why he had to put the last phrases in; it kind of messes up the idea a little bit. But essentially if you're looking at a phylogenetic tree, what he's saying is there's a start and a finish on a branch. So it has duration through time; that kind of thing.
And then we'll come back to the genetic cluster species concept, which is one that Jim Mallet put forth. Basically it says if you can get a lot of genetic data on the things that you're looking at, and you plot them out into gene space somehow, and you find that they form distinct clusters, then each separate cluster is a separate species.
Now each of these is useful in a somewhat different context. And, as I said, I think that they all get at parts of what we mean when we say the word species. And I don't want you to forget the question I have hanging in the air, which is do we have to make everything into a species? Okay?
So, concepts and criteria. This I know looks a little weird. Okay? Why do they distinguish between concepts and criteria? Because after all criteria are concepts. Okay? This is because they were trying to achieve kind of a local clarity in what they were talking about. So the concepts are more or less general statements about how you might go about thinking about what a species is, and the criteria are rules of thumb to decide whether a thing is a species or not. So you can think of the concepts as being abstract and the criteria being practical.
So we've got these concepts. And here are a few criteria that could be applied to any concept. So one would be, why are the things separated? Is there initial separation? And of course if we want to go into it, we can look at the causes, but maybe the most important thing is just that they're separated, not why.
We can look at whether the species is cohesive in some sense; there are different definitions of cohesion, but one is, is there genetic mixture? Are they breeding with each other? So cohesion has something to do with the biological species concept. Another is, is that--are the organisms in the populations that we're looking at monophyletic? I'm going to come to the definition of monophyly today, and I'm going to repeat it next time, so that it gets hammered in a little bit. But basically that means do they all share a common ancestor?
And then there's the issue of distinguishability; can we actually tell them apart? Now when we get down to cryptic species, at the end of the lecture, you'll see that they are indistinguishable, except at the genetic level. We can't look at them and see any difference. But for many things these will be useful criteria.
So, monophyly. The things in monophyletic groups share a common ancestor, and that common ancestor is not the ancestor of any other group, and there aren't any things that are descended from that ancestor that aren't in this group. Okay? So it's really everything that came from that common ancestor. So birds all appear to have shared a common ancestor that split off from basically a group of dinosaurs, back in the Cretaceous; possibly a little earlier, in the Jurassic. Mammals all split off earlier from a group of organisms whose later descendents then included the dinosaurs, but also many other things in that grade.
In contrast to a monophyletic group, a paraphyletic group is a group that doesn't contain all the things that are descended from the most recent common ancestor of its members. Okay? So yes, everything in here has a common ancestor, but hey, there's some other things that descended from that common ancestor that are out here.
So, for example, if you call fish a natural group, you're making a mistake because the tetrapods are descended from the same ancestors as the ancestors of everything you want to call a fish. So the amphibians, the reptiles--I'll come to the reptiles in a minute [chuckles]--the birds and the mammals are also things that we ought to call fish--okay?--by that definition.
So we need a different word to reflect the history of relationship. In the reptiles, the birds and the mammals are missing. Okay? So if we say reptile, and we want to really refer accurately to a Tree of Life, a good phylogenetic tree, then that word should actually also mean, oh, we mean the birds and the mammals too. The thing that's going on here is that our everyday language culturally evolved before our scientific discoveries demonstrated what the natural relationships of groups were, and therefore we have embedded in our everyday language mistakes. So that's why these distinctions have been drawn. They are pointers to mistakes. So if somebody says paraphyletic, it's like that's a mistake.
Then we have polyphyletic, which is another kind of a mistake, and a group is polyphyletic--the word referring to that group is a mistake--if the things in it are descended from several ancestors that are also the ancestors of things that are classified into other groups. So, for example, all the stuff that Linniaeus called worms was highly polyphyletic. It included the mollusks. It included therefore the octopuses and the squid.
If you were to refer to the Old World Euphorbiaceae and the New World Cactaceae, among the plants, as being in the same group because they look the same, you would be wrong, because they both have--they're actually fairly distantly related from each other and they have converged on Cactus-like forms, and they have closest relatives that don't look anything like a Cactus. So if you ever make a group like that, that would be polyphyletic.
Are there any questions about these three distinctions, or have I managed to get across those differences clearly enough? Because this is--it's important; it's simple but it's important.
Okay, species concepts and criteria were reconciled by Kevin De Queiroz, and basically what Kevin did was more or less arrive on a scene where people had been embroiled in controversy over species concepts, and he wanted to kind of stop all of the squabbling about the concepts and try to move the research on by more or less trying to resolve that, and do so in a constructive way.
He did a pretty good job. Basically he said, "Hey, let's see if we can agree--not everybody does--but let's see if we can agree that species are entire population level segments of lineages, from origination to extinction." So they have duration in time, and they're at the level of a population, and we could map them onto a Tree of Life.
And the criteria aren't the same as concepts. Basically what they're doing is they're marking stages in the existence of a species, and they don't actually determine whether the species is a species yet. Okay? Now, as we'll see, if you actually go through what's going on as a species starts to evolve, on the Tree of Life, and you mark off these species criteria, I think you'll see that much of the time, by the time you've stepped through all your criteria, you have a good species.
Okay. What are they? Well there's separation. There's cohesion, and the cohesion can be genetic and caused by actual or potential interbreeding. It can be cohesion of recognition; so everybody who is a potential mate can recognize each other as a potential mate. It can be a cohesion of viability and fertility. So there's post-zygotic compatibility; very important to include a lot of plant species there.
Then there can be ecological cohesion. So these things are living in the same habitat, at the same time, behaving as an ecological unit, and very probably doing so because they are actually inter-breeding. By the way, this business of remaining separate, in the same place, is a good criterion for two separate species. So if you see two things that are remaining separate, even though they are continually running into each other in the same time and place, then that's a pretty good indication that they're separate species.
So another criterion is monophyly; so they share a single most recent common ancestor. And then you can distinguish them either through fixed morphological differences or as a phenotypic cluster in phenotype space--this is a more quantitative thing than the quantum differences of distinct morphology--or they could be a genotypic cluster, in a genotype space.
Now when you think--I'm now going to lead up to a bit on the genetics of speciation, and this slide is intended as a transition into that, and it is intended to show you basically that gene trees often differ from population trees. So what's going to go on in a speciation event is that you're coming along the Tree of Life and there's going to be a split and there are going to be two new branches. And these are entire populations down here. Okay? So you should think of each branch down here as having thousands of organisms in it.
And what these lines are showing you is that there are splits in gene trees that can start occurring before the population splits, and that can continue to occur after the population splits. And what we mean by a gene tree split is actually a couple of possibilities. One is, okay, a mutation occurs here at a locus, and we now have two different alleles that are in this population, and these two different alleles then can continue to be maintained by various processes, and one of them ends up going down this branch, and then the descendents of this one end up going down this branch. That's one possibility.
Another possibility is, oh, we have a gene duplication even here; so we now have two different versions of the gene. They both are being used for something. They're functional, they don't become pseudo-genes necessarily, and the descendents of one copy--actually in that case you ought to get both copies going down both branches of the gene duplication event. So that would probably be more like this.
I suppose that it's conceivable actually that you could lose a copy and still have functional organisms, if the process of acquiring function for the second copy was still going on in this region of time here. At any rate, the point here is that if you look back to try to find the last common ancestor of different genes, or different alleles, it's quite possible that that last common ancestor for the genes will be deeper in time than the last common ancestor of the species; and that's because species have lots of genes. Okay?
So if we look at that kind of process, here is a separation point where, after this, each of these branches is going to be monophyletic. At this point some of the genes have become fixed. So by showing that this one is gone over here--that's just intended as a symbol, an analogy, to indicate that a gene has become fixed--and over here in this lineage maybe some other one has become fixed, and at that point genetically these species are distinguishable. You can tell them apart because now one lineage has one version of a gene, and the other lineage has another.
Then as these genetic differences accumulate, later in time they will be reproductively isolated from each other. If two individuals of the two species did meet, they would not be able to have viable grandchildren. And of course I've shown you that that reproductive isolation can occur at three different stages. It's likely that the sequence in which reproductive isolation is acquired is actually the reverse of the sequence that I laid it out. So it would be first post-zygotic isolation; then post-fertilization isolation; and then finally pre-zygotic isolation, with recognition that the two things are different. Okay?
And then finally up here, no hybridization at all is possible and you have complete genetic isolation. So you would think that this might be--with reproductive isolation that you wouldn't get any hybrids, but there could be a few mistakes, and so finally complete gene exclusion is usually coming after what is in practice reproductive isolation. This is a bit imperfect.
So if we look into the genome, and this has been done best with Drosophila, because it's such a good model organism--and this is, if you want to go into this kind of thing, search under C.I. Wu, that's Chung-I Wu; he's at Chicago. And basically what Chung-I said was that if we look across a genome--okay, so you can think of A, B and C just being as markers on the genome--there's gene flow between two populations at lots of places in the genome as speciation just is starting to get going. Okay?
So at this stage we have populations or races. They've got perhaps different kinds of adaptations. There isn't any reproductive isolation, and between these three loci there's lots of recombination; the three loci are actually capable of moving back and forth. So A and a can move back and forth; B and b, and so forth.
Then, in the second stage, there starts to be some block to gene exchange, and here we are at a level where we're getting a transition between a race and a species, or between a race and sub-species, with some degree of reproductive isolation, and at this stage the populations could fuse or they could diverge. But these blocks here, that you're seeing, are indicating portions of the genome where there isn't any exchange anymore.
So the idea here is that the origin of reproductive isolation, if you look at the genetic level, is a gradual process; it doesn't occur all at once. It starts in parts of the genome, and it continues, and while it is developing, there's some still some gene exchange going on, at certain loci. And you might think--and I think this is probably the case--the parts where the exchange is still going on are the parts that aren't having big impact on hybrid inviability or mate recognition. The parts that are starting to get frozen up are the parts that have to do with any of those three levels of separation.
So when reproductive isolation is complete, then all of the genes are free to diverge. So here the populations are beyond the point where they could fuse but--and they're a good species--but there's a still little bit of hybridization going on, that you can pick up, and here they're completely separated and there's no gene flow anymore.
Now Chung-I did this actually with simulans, Drosophila simulans, Drosophila mauritania, and the two sexual races of Drosophila melanagaster, which are in Africa. Okay? And he looked through the genomes of these species to see at how many places were there genes that were diverging, and what were those genes coding for? So the differences between the two sexual races of Drosophila melanagaster, which are really quite recently separated--we're talking perhaps 50 or 100,00 years here--are all in sexual behavior. That's things like how fast does the male vibrate his wings?
And Stage 3, which is about here, is between simulans and mauritania, and here we're starting to get back maybe 500,000 years, something like that; and here we're starting to pick up lots of sterility and inviability or female sterility genes, and changes in genital morphology so that lock/key mechanisms are incompatible. And then finally when you get this divergence here, which is back at one to two million years, basically you just see a lot more genes in these categories; sterility and inviability accumulate.
So how does a new species actually come into existence? Well going from that rather modern evolutionary genetics, back a bit to the arguments in the '40s, '50s and '60s over speciation, a lot of that argument is about geographic speciation. Darwin emphasized allopatric speciation. Gulick, who worked on the speciation of snails in Hawaii and Polynesia, emphasized allopatric speciation. Sympatric and parapatric speciation are things that came in, in the '60s and '70s. Allopatric means allo-patric, different places; sym-patric, same place; para-patric, next to each other.
And genetically, as you've seen from Chung-I Wu's analysis, initially there are only a few genes that are changing. So even though you have 35,000 genes in the genome, speciation might be driven by four or five, and that's what might allow it to occur fairly quickly. And then after reproductive isolation, the changes in those genes mean that all the others will diverge. So a good criterion for a species is, as I told you, they remain separate when they're living in the same place and encountering each other at the same place and the same time.
So let's take a look schematically at what these terms mean. So in allopatric speciation you have an original population. There's some geographic or geological barrier that forms. They then start to diverge in isolation--so here they just start having different colors--and then after they might come back together they won't breed again, because they've diverged so much.
So the classical large-scale example of this is the ratite birds. Those are the flightless birds of the Southern continents, and they include the Ostrich, the Cassowary, the Emu, the Kiwi, the Elephant Bird of Madagascar, the Moas of New Zealand. And arguably those are all birds that originated from a common ancestor that was on Gondwana, and when Gondwana broke up through plate tectonics, these just rode around on the plates. They never flew and they never swam. Okay? So they were actually carried on pieces of rock into different places.
There are, by the way, many other things that can form barriers. Mountain ranges can go up, and when they do, river basins can change, drainage basins can change. Things can become reproductively isolated when they fly out to an oceanic island, and so forth.
Now parapatric speciation means next door, right next door. So some individuals of a population get up and they move and they just go right next door. This is thought to have been actually how the Galapagos finches might have speciated on a single island. So some of them might have simply gone up the mountain, while others stayed down on the coast. That would be parapatric speciation. It is also thought to have been what has driven the speciation of frogs in the Eastern United States, and is still in the process of going on.
So, for example, in the Leopard Frog, on the East Coast of the United States, you can breed individuals from Connecticut and New York and they do just fine. Now that frog has a range that goes from Canada to Georgia, and if you try to breed individuals from Quebec with individuals from Georgia, they won't. But all along the way you can make the crosses, which means that it is possible in principle for a gene from Quebec to end up in Georgia. So according to the biological species definition, it's a species, but in fact it's in the process of splitting up. Okay? And by the way, this pattern is something that got re-established after the glaciers melted. So the frogs moved north, and this pattern has gotten set up in the last 12,000 years, about. Oh, I should go back and do these guys.
Okay, that's parapatric and this is sympatric. Sympatric is particularly interesting because the controversy over this was so violent and it involved some of Ernst Mayr's Ph.D. students rebelling against Mayr; so that was Guy Bush. And basically the idea behind sympatric speciation is that it's possible for a population, where all of the organisms are living in the same general place and encountering each other at the same general times, to split according to ecological processes that are going on in that little area.
The initial example was a fruit fly that switched from living on apples to living on cherries, where the apple trees and the cherry trees were all in the same orchard, And because the offspring were imprinted on the smell of the fruit in which they grew up, when they grew up, they tended to go to that same kind of fruit, and because mating took place on the surface of the fruit, they became reproductively isolated and they started to diverge. The name of this one is Rhagoletis, and its sympatric speciation has actually been documented and followed, and it's occurring in Michigan. You'll see later, we have an example from fish in African lakes where it looks like this is going on.
Now pushing experimental evolution from questions of microevolution into questions of macroevolution is not easy, because by definition macroevolution describes phenomena that are taking place over longer periods of time. Nevertheless, there have been attempts to cause speciation to happen in the laboratory. They have mostly been with short-lived fruit flies, and I'd just like to summarize some of that experimental evidence so that you can see how close people have come to actually being able to create new species in the lab.
If you do divergent selection in allopatry--so you split a population in two and you don't let any gene flow occur between them, and then you select strongly on one trait in one direction in one population, and in the other direction in the other--you can actually cause reproductive isolation to evolve; you can do that in a couple of hundred generations. If you select them in the same direction in allopatry, that doesn't seem to work.
If you destroy the hybrids, when they are in sympatry, and you divergently select in sympatry, that usually works. So you can start the speciation process in the lab, in sympatry, by destroying any hybrids between the two things that you're trying to make diverge, and selecting those two things in different directions, morphologically. But if you divergently select without destroying hybrids, that almost never works. Okay? So there is an experimental biology of speciation, and there is a literature on that. This is something that's particularly been done in Spain, by the way.
Now recently there has been some recent evolutionary theory on whether or not sympatric speciation is possible, and it does appear that it's possible, and it can occur rapidly if you have divergent selection to different habitats, and you couple that with preferential mating with your own ecological type. So this kind of thinking has mostly been applied to fish, and it's been applied to sticklebacks living in lakes in British Columbia, and it has been applied to cichlids living in lakes in Africa. Let's take a look at some cichlids in Africa.
So here are two cichlids that are recently derived from the same common ancestor and often classified into the same species. They are Tilapia. They're living in a lake in Cameroon. This is male size versus female size, and so you can see that there are basically two clusters with a few intermediates. There are some little males and females, and there are some big males and females, and you tend to find the small ones in deep water and the big ones in shallow water.
And if you look at the frequency with which you find them in deep and shallow, you can see that you only get the small ones in deep water, and there are a few of the big ones in deep water, but most of the big ones are in shallow water. You'll notice--by the way, it's darker in deeper water, and you'll notice that the ones that are the small ones living in deeper water already are evolving bigger eyes, and there's starting to be some reproductive isolation between these.
So this is a circumstance in which there is a different ecological habitat; there are different foraging and sensory specializations that are needed in the deep habitat than in the shallow habitat. These guys are probably feeding on snails on the bottom, and these are probably feeding on plankton in mid-water. There is good reason for these two not to cross, because the intermediate phenotypes aren't nearly as good; okay, they aren't ecologically as good, they don't perform as well in eating snails or in foraging for zooplankton.
So that's sort of the classical ecological speciation hypothesis. Dolph Schluter at British Columbia has written a whole book on this. There's a big literature on it, and a lot of people are interested in the idea that there could be some sympatric speciation going on.
Now we saw that these species criteria work fairly well for a certain range of organisms, but then there's also the genotypic cluster thing, and that really solves a problem with asexual and cryptic species. So I'd like to just introduce you to that issue.
One context where it's useful actually is with bacteria. So here are a whole bunch of bacteria that all live in the gut, enteric bacteria, gut flora. And these guys are all pretty proficient at horizontal gene exchange. There's about 30% of their genome that they exchange back and forth, pretty freely, so that there's a lot of recombination going on across most of this tree here. But the housekeeping genes of these bacteria don't exchange horizontally so well.
There's a core group of genes, and they are the genes that are responsible for energy metabolism and the construction of ribosomes and cell walls and stuff like that, that actually form a core that doesn't participate in horizontal gene exchange. And if you just concentrate on that core of genes, you get a perfectly good phylogenetic tree. It forms a gene cluster. If you include the 30% that are being horizontally exchanged, the whole thing gets very fuzzy. So that gene cluster definition is good in bacteria, if you concentrate on the core genes, which are housekeeping genes.
There is a radiation of salamanders, Batrachoseps, in California--well studied by Dave Wake's lab at Berkeley, by the way--where there are a very large number of species and local forms, and many of these can only be recognized if you look at gene sequences. So cryptic species are common in some groups of salamanders. Cryptic species are also common in the algae that coral farm. So reef-building corals have endosymbiotic commensile algae that they farm, and the strain of algae and the strain of coral have a lot of cryptic diversity. You look at it and you can't tell the difference from the outside. There's a lot of diversity like that, out there in the world.
But this is really one of the most interesting cases of a cryptic species, Tetrahymena. So this is a ciliate, and there are ten morphologically, completely indistinguishable forms that had a last common ancestor about a hundred million years ago. So these things are a good example of stasis. They still look exactly the same. Okay?
However they have diverged a tremendous amount in their DNA sequences, and interestingly even though they look exactly the same, the proteins that build those structures have diverged. It's like there has been really strong stabilizing selection to keep those things looking the same, even though the genes, and the proteins that build them, have diverged. So that's deeply cryptic, and there's things like that out there. So there are cryptic species.
So first, evolutionary biologists have come up with a bunch of different species concepts. It's like the Hindu tale of the blind man and the elephant, and one of them touches the trunk, calls it a rope; one of them touches the leg, calls it a tree. They're all coming from slightly different perspectives; they're emphasizing different aspects of the process.
What is helpful in thinking about a species tree is that a species is a segment of a tree from start to finish, and if you're thinking about gene genealogies, a species is a population that is monophyletic and it is distinct from other similar populations that are monophyletic.
The criteria you might want to think about, if you want to sit down and logically distinguish two populations and call them different species, are separation, cohesion, monophyly and diagnosibility--can you really tell differences? And I've just shown you--sometimes you can only see the differences in the DNA sequences. And these criteria do mark stages in the divergence of lineages.
The genes that are involved in speciation start off few in number. You might only need to be influencing some aspect of hybrid inviability or some aspect of mate choice and behavior; perhaps just changing the frequency of wing fluttering or something like that. But then, over time, the number of genes involved increase, and other kinds of reproductive isolation mechanisms evolve. And then there are these very cryptic species that can only be distinguished as clusters in genetic space.
Now if you get interested in this stuff, here's some--I've stuck this in, so you'll have this for reference--the January 20, 2009 issue of Heredity has ten articles on the genetics of speciation. Jerry Coyne and Allen Orr have a good fairly recent book on speciation. You can also, if you want to, search for Trevor Price, Dolph Schluter, Sergey Gavrilets, Ulf Dieckmann, Michael Doebeli; and you'll pick up books and papers on recent speciation ideas. And next time we're going to take this basic element of the Tree of Life, which is the separation of the branches, and start looking at phylogenetic systematics.
[end of transcript]