EEB 122: Principles of Evolution, Ecology and Behavior?

Lecture?16. Comparative Methods: Trees, Maps, and Traits

https://oyc.yale.edu/ecology-and-evolutionary-biology/eeb-122/lecture-16

Now we're going to see what happens when you either lay these trees onto maps, or you put traits onto the trees, or you do both at the same time. This is basically comparative biology?in its modern sense.?I'm running through a series of issues that can be resolved using comparative methods in phylogentic trees.

What happened in Europe after the glaciers melted.?At the peak of the last Ice Age, the glacier came down out of Scandinavia and got down?into Northern Germany and Poland. The English Channel was dry because there was so much water that had been locked up in the continental ice sheath that the level of the earth's oceans dropped about 100 meters, and there was a sub-glacial tundra that stretched?from Ireland all the way across France and through Russia?out into Siberia.

• It was called the Mammoth Step. Many of the animals that you now find in Northern Europe had retreated south?into glacial refugia. There was one in Spain, one in Italy, one in Greece and the Balkans, and in Asia Minor.

• In fact?there are some?hybrid zones?in Europe. These are places where you will frequently run into hybrids, and they are there because?populations are coming back together that had been isolated in the Ice Ages and breeding with each other.?
  

• It's possible, using mitochondrial DNA, molecular phylogenies, to reconstruct the recent history of movements of animals and trees?across the planet,?and to understand why it is that there are certain places where we see hybrid zones.

Now what about humans??I'm going to show you first what's happened in about the last 10,000 years, and we're going to see that in Europe the agriculturists spread out from the Middle East and squeezed the Celts into the northwest. In Africa we're going to see the Bantu migration out of Cameroon, and how the Hottentots were squeezed into the southwest of Africa. And in Asia we'll see that agriculturists spread from both the Middle East and from China, and squeezed Siberians into the north.?

• Cavalli Sforza, Paolo Menozzi, and Alberto Piazza?tried to come up with a method of compressing a huge amount of genetic information onto a map, and they did it by taking gene frequencies, at hundreds of genes, and then compressing them, using statistical analysis, into a few factors, and then plotting those factors onto the map.?
  

• What you can see here is basically the population differentiation of humans, in Europe, and you can see that there is kind of a wave that comes out of the Fertile Crescent and moves up to the north and to the west. And this tracks the agricultural expansion, which started about five or six thousand years ago, out of the Middle East. And you can see that the Celtic genes did get squeezed up into Ireland and England, and out into Brittany.

• If you look at Africa, what you can see is that there has been an invasion of Africa by Caucasoid Northern Africans, and by sort of an Arab Nilotic expansion?coming down this way. The Bantu expansion out of Cameroon is what is coloring part of the continent pretty red. And?this migration got through here about a thousand years ago.?This kind of study will no longer be possible after another few hundred years of jet travel and crossbreeding. Another few generations of that and this map will not be reconstructable.?

• If we look at Asia, what we see is that there were nomads and agriculturists coming out of the area around the Middle East, and around the Black Sea and the Caspian Sea, that have pushed into Central Asia, and the Chinese agriculturalists have spread into Southern Asia. There's a very interesting hotspot of human biodiversity in southeastern Asia, going from India over?to Taiwan; and actually this is where the Polynesians came out of. The Polynesians left from Taiwan about 5000 years ago, and that's confirmed both in the language reconstruction and in the mitochondrial DNA.?

So you can see that one of the themes of recent human history has basically been of the expansion of some groups at the expense of others, and that that was often a technology-driven thing, and often involved agriculture.?

It is now possible to get information on single nucleotide polymorphisms at 650,000 different sites in the human genome, and this is a paper that did that for 928 unrelated individuals from 51 populations. So these are the 928 individuals out here. These names down here are the 51 populations, and the 650,000 different positions in the human genome are on the Y axis, all compressed together.

And you can see that there is certainly a genetic signature across this; certain kinds of genomes in certain geographical areas. And if you then do the molecular phylogenetics on it, and construct the phylogenetic tree, you see that the oldest part of the modern human tree is centered in Africa. This?is the Classical view. This was a picture that could be drawn in 1995--and this is 2008; so this thirteen years later--and this tree largely confirms this picture.?So phylogenetic methods can now be used to give us quite a bit of insight into our own history, as well as into the history of other plants and animals.?

• What you see here basically is we came out of Africa, we paused in the Middle East. Then various groups moved out of the Middle East.

• One group went into Europe, thought to be about 40,000 years ago; that's these guys.

• Other groups set out into Asia and spread out through Asia.?And then out of the group that had settled basically in Eastern China and Japan and Korea, one group up here split off.

• Part of them--actually an early part of this branch--went down here, through Papua, New Guinea, into Australia, and another part went out into the New World.

When we look at something like the human expansion across the globe, it's actually difficult to get precise markers for the times when they arrived in certain places. Archeology gives us some; sometimes we can recover fossil DNA from bones. But in the case of the Hawaiian Islands, at least on the scale of the last 5,000,000 years, we have very precise geological dates.?

We know that Kohala Mountain, the oldest rock on Kohala Mountain is 430,000 years old, and that the oldest rock on Kauai is 5.1 million years old; and that's because the islands are made over a hotspot?here, and carried on a plate up in that direction. And you can actually lay down, on this plate, how long ago it was that that island was actually sitting down here. We can?then do phylogenetics start putting phylogenetic trees onto this map.

These are three different ways that spiders and some other arthropods and some insects speciated over the last 5,000,000 years in Hawaii. Interestingly, you can see that they all moved down from Kauai onto the younger islands. So they were going from older islands onto younger islands, and they just kept hopping.?If you just continue this island chain up to where it dives into Siberia, there were islands there 350,000,000 years ago that are now getting subducted under Kamchatka, and there are things that are one or two-thousand miles up to the northwest that were once high islands above water. All of these processes can actually be seen written in the genes, and this is all resulting basically from sequencing, either nuclear or mitochondrial DNA.?

On Madagascar you find a local radiation of things that look kind of like civet cats or mongooses. And the question was, did they come over from Africa separately, or did they all speciate on Madagascar??

• Anne Yoder?reconstructed the phylogeny of this group of animals well enough so she could lay this tree onto this map and determine that in fact these guys are actually all relatives of mongooses?and mongooses are close relatives of hyenas, and those things have a sister group of civets, and those things have a sister group of cats.

• Madagascar split off from Africa about 65,000,000 years ago. It's part of the Tectonic breakup that led to India splitting off, skitting across the Indian Ocean, crashing into Asia, and raising the Himalayas. How?does something like this cross a strait which is now more than 200 miles wide?

• If you go out on the Limpopo River, there are large rafts of trees and vegetation being carried down it, and in a good storm you can put a mongoose on a raft of vegetation and get it out across that strait; so it rafted across.

There are parasitoid wasps which can be either ectoparasites or endoparasites. This represents one of the phylogenetic surprises.?

• The ectoparasites lay their eggs on the outsides of caterpillars, and the eggs hatch and make little baby wasps that crawl around the outside of the caterpillar, and they eat it from the outside. They have relatives who are endoparasites. The Ichneumonids are like that.

• It had been thought that if you just look at it mechanically, it would be easier to start the evolutionary process off with the wasp just flying around and laying its egg.?But in this particular radiations,?there may have been an ectoparasitic wasp that was an ancestor. But it turns out that all of these things are endoparasites, and the ectoparasites evolved within that, and then within the ectoparasites you had a reversal again and got some endoparasites out of it.?

I would like to discuss the Anolis lizards. And people?study Anolis lizards because these are prominent, easily observed; you can get a pretty good sample size fairly quickly. The Anolis lizards have had a big radiation on islands in the Caribbean, and they have made what are called ecomorphs.?

• These ecomorphs can be grouped by appearance. So if you look at them and you see where they're living and what kinds of grasping appendages they have, basically what the phenotype looks like and how they behave, you can come up with things that live on the crowns of trees; out on twigs; down in the ground in grass and in bushes; between the trunk and the crown; only on the trunk; or down on the ground and going up on the trunk.?So there are six ecomorphs.

• After DNA sequencing came along, you could do their phylogeny. There is?independent evolution and the convergence on different islands of these different ecomorphs?among these lizards. And if you take that down and you break it down by island, what you can see is that on Cuba the trunk form, the crown trunk form was ancestral. On Hispaniola,?a crown giant form, was the ancestral form.

• These are inferences about what first got onto that island and what first got out to the other island. And then these other things all evolved from that. And what you see from that basically is that it doesn't matter which form of lizard you first throw onto an island; all of the other ones are going to evolve from it. And they?are differentiated at the level of genera.?

Across the Caribbean?essentially the same ecological community is being generated again and again, independent of which kind of species founded that group. People didn't think this would happen. So this means that you're getting the same ecomorphs from different ancestral states, and that means you've got convergence.?

Joe Felsenstein contributed?to phylogenetics. You can see that the red trait evolved in the ancestor of both B and C. So that means that B and C share a trait. However, in evolutionary terms, it only evolved once.?The reason that red increased in frequency might have been that it was adaptive and?had a fitness advantage. So microevolution was driving it at that point. But then everybody inherited it, and it is not an adaptation to the difference between whatever environments B and C now live in. So if you were to look at B and C now, and you saw that they shared some trait, you wouldn't really know why that was there?until you could get a much larger sample size.?So how do you deal with that problem?

• Joe came up with what he called the method of independent contrast. And in this context a contrast is the difference of the value, the mean value of a trait in one species and its value in another species. And if you look up just at the tips of this phylogenetic tree and you take the differences across the closest related sister pairs, at the tops of the tree, you generate these contrasts. You get X2 minus X1; X4 minus X3; and so forth.

• The important thing about the contrasts is this. The difference that evolved, after this point on the tree, is independent of the difference that evolved after this point on the tree. You've taken out whatever was there because of the common ancestor. So whatever was going on over in this part of the tree is biologically separated and now statistically separated by this method.

• It?actually is a method of getting the correct sample size?off a phylogenetic tree. And that's very important in statistics, because if you have the wrong sample size, all your statistical tests will be wrong.?

So let's take a look at some approaches that are kind of like that.?Peter Grubb?plotted?the log of seed mass of light-demanding seeds, against the log of seed mass of shade- tolerant seeds. And his question was this: Do plants living in shade produce larger seeds than plants living in sun? And he wanted to do a phylogenetically controlled comparison.?

• So what he's doing here is he's taking essentially the value within a genus?or within a family--the mean value of species within a genus or the mean value of genera within families--for related trees, some of which live in open areas and demand light for their germination, and others of which have seeds that can germinate and survive in the shade. So the open circles are comparing genera within families, and the closed circles are comparing species within genera.

• Plants living in shade do produce larger seeds than plants living in sun, but you only see it in comparison of genera within families; you don't see it in comparison of species within genera. So the ones that need light and the ones that need shade have just about exactly the same seed size, if you are looking at species within genera. But if you then give them longer to evolve, you go further out on the phylogenetic tree, you compare things between families, where that contrast is possible, then you start to see them moving off the one-to-one line, and the ones that are shade-tolerant have seeds which are falling quite a bit--not all; this is an exception--but quite a bit above the line.?

So this is a way not only to answer that kind of question, using the comparative method, but also to get an estimate on how long it takes. It takes a long time to generate that difference, because you only see it at a higher level on the tree.?

Now what about the albatrosses and their relatives??

• The Wandering Albatross, which has a wingspread of between twelve and thirteen feet, and is probably the heaviest flying bird.?And it mates?and usually will produce an offspring every other year or so, for about thirty or forty years. They mate for life; they're monogamous. And they have very precise homing behavior to their chicks. Its?wide foraging means that they're only going to be raising one child every two years or so; and there are all kinds of adaptations in the infant's physiology to deal with this irregular feeding.?

• A Storm-Petrel is a lot smaller. It forages much closer to shore. It is not so faithful to its mate. These things are related. Okay? So if you look within this family, can you ask the question about whether or not a long life really is a situation that promotes mate fidelity? The argument there is basically that if you're going to have a short life, there isn't going to be enough time to pick up the advantages that you would get from knowing a particular mate, and adapting behavior to that particular mate, and learning exactly how to be a good parent with that particular mate, rather than with some other.?

• If you look across the Procellariiformes; so this would be the Wandering Albatross up here, and this would be a Petrel down here, and you've got some other things in the middle--these are all separate species now; the dots here are all species. You are looking at independent contrasts now, and we're plotting the independent contrasts. So this is the deviation in adult life expectancy from an overall mean, for the whole group, and this is the deviation in mate fidelity from an overall mean.

• So the ones that tend to live a long time are more faithful to their mates, and the ones that tend to have a short life?are much less faithful to their mates and switch mates. And that's interesting because in fact these guys, the ones that reproduce many times, have much more opportunity. They have thirty or forty years of reproduction. They could go out and they could get divorced and pair again several times. But they don't, they stick together. And the functional reasons for that are things that are not really terribly well understood.?

I want to ask you a question about?molecular phylogenetics which?requires?you to piece together phylogenetics and genetic drift.?It's basically on human mitochondrial evolution. All human mitochondria appear to be derived from an ancestor, all of whose closest relatives,?out at the tip of the tree,?lived in Africa.?

• Human mitochondria show that we came out of Africa. We?know it?from SNP polymorphisms and?nuclear genes as well.?It said there was one woman, living in Africa, about 220,000 years ago, from whom all other mitochondria in all humans on the planet are descended, and so they gave her the name Mitochondrial Eve.

• How old are the polymorphisms in our MHC genes? And these are things that have been selected, probably through frequency dependent selection, in co-evolution with diseases.?Say we have two MHC genes that have resulted from a gene duplication,?and at each of those genes there is a polymorphism, so that we have different alleles at that locus?for each of those two genes. Who are those alleles most closely related to?

• It turns out that Allele 1 in humans is most closely related to Allele 1 in chimps, and Allele 2 in humans is most closely related to Allele 2 in chimps. In other words, the closest relatives of the alleles are not in this species but?in another species. The only possible way that could have occurred is if the polymorphism originated before the speciation event. So that you had ancestral species here, and this polymorphism originated, and you got this one coming down from the ancestral species into the chimp, and into the human; and this one coming down from the ancestral species into the chimp and into the human.

On the one hand we have the claim all the mitochondria came out of one person. On the other hand we have evidence that there are trans-specific polymorphisms. How those two observations are consistent with each other.?

• All of the mitochondria came from one female because it's asexually inherited. The trans-specific polymorphism tells us how many females there were in the African population 220,000 years ago. When a population is really small, genetic drift will occur. If a population were as small as one female, it would've been impossible to maintain this trans-specific polymorphism.

• People have done simulations to find out what is the average size of a population that would, over a period of 5,000,000 years, maintain the amount of trans-specific polymorphism that we see in our MHC complex; in other words, the amount of genes that we share with chimpanzees, where our alleles are more closely related to the chimp's alleles than they are to the other human alleles?

• The answer is the minimum size is about 10,000. In other words, we have good genetic information that tells us that the smallest the human population ever has been, over the last several hundred-thousand years, actually over the last 5,000,000 years, since we shared ancestors with chimps, is about 10,000.

• It's possible that in fact she had a particularly advantageous mitochondrion, and that it then got selected and it fixed.?And that could've been done in a larger population. However, it could also have been done with drift, and it happened so long ago that we can't really tell whether it was selection or drift that gave that one woman the advantage.

• The same thing has been done for the Y chromosome. The Y chromosome is also asexually inherited, and the estimate on the Y chromosome is roughly also about 200,000 years ago, also in East Africa. And the fact that the mitochondrion and the Y chromosome both converged to a common ancestor, at about the same age, might suggest that drift is more likely than selection to explain it.?

If you go back far enough in time, it will always converge on a single common ancestor. But there was a controversy?on dating that point of convergence. It was?all about how long has it been since we split with chimps? Because that turns out to be the baseline that gives us an estimate of how rapidly evolution is going on, molecular evolution is going on, in the human clade. Well when you apply that criterion to what the confidence limits on the estimate are, hey, it was anywhere from last year to about a million years ago.?

We should never be surprised if particular mitochondria or particular chromosomes converge at some point in time--it looks like they were just in one individual. That's just because of the way that branching processes work, and that will be going on all the time, and it's happened over and over and over again.?

This process, convergence back to a single ancestor from the mitochondria, it's not entirely independent of population size--it'll take longer in a big population than it would in a small population--but it doesn't give us an estimate of how big the population had to be. Whereas the trans-specific polymorphism could only have been maintained, even with strong selection, in a population that was larger than about 10,000 individuals; and that can be done with computer simulations.?

Okay, so to summarize this part of our exploration of macroevolution.?

• These molecular methods allow us to reconstruct geographic movements, as well as phylogenies. And we saw that in the hedgehogs going north from Spain and the Balkans, and we saw it in the humans moving out of Africa, and we saw it in lots of things.

• We see that our own migrations have left genetic traces on all the continents. And there's an imprecise map that's suggestive between the genetic geography and the linguistic geography. Greek genes go with Greek family names, from the boot of Italy, up to about Rome, and then stop; that kind of thing. So even in the last two or three-thousand years, you can see that family names and genes have been inherited in similar ways.

• We can use these methods to determine which trait states were ancestral and which are derived. And that was particularly interesting in the case of the parasitic wasp, whether it was an ectoparasite or an endoparasite, because it changed received opinion about fundamental biology.

• And this method that Joe Felsenstein worked out for independent contrast is something that will control for common ancestry, and it can reveal the correlated changes in two or more traits that have taken place since branches in the tree. So it can be used as a fairly powerful method to explore hypotheses in behavioral ecology, evolutionary ecology, and ethology.?