More Recent Comments

Wednesday, December 08, 2010

The Cause of Speciation

 
Jerry Coyne is an expert on speciation. That's why it's always informative to read his latest thoughts on the problem. In spite of what many people might believe, the main cause of reproductive isolation—the actual speciation event according to the biological species concept—is due to accident, not adaptation. It's just one more example of the importance of random genetic drift in evolution.

Here's how Coyne puts it in his latest posting [“Reinforcement” and the origin of species].
Genetic barriers aren’t thought to arise for the purpose of keeping species distinct. Rather, they are usually thought to be evolutionary accidents: geographically isolated populations diverge genetically under natural selection or other evolutionary forces like genetic drift, and that divergence leads to the evolution of genetic barriers (mate discrimination, the sterility of hybrids, ecological differences, etc.) as byproducts of evolutionary change. For example, populations could adapt to different environments (one dry, one wet, for example), leading to them becoming genetically different. When these populations meet each other again, this genetic divergence could result in hybrids that don’t develop properly because the parental genomes are sufficiently diverged that they can’t cooperate in building a single individual.
I wish more people would assimilate this message. It seems to be the overwhelming consensus among the experts in speciation but the average scientist still has an adaptationist view of speciation (and of evolution in general).

Speaking of adaptationists, Coyne also likes the idea that some examples of reproductive isolation can be reinforced by natural selection. You can read about those cases on his blog.


25 comments :

El PaleoFreak said...

Three ideas:

-If speciation is often a byproduct of adaptation, then speciation is an *accident*, but the *cause* of this accident is adaptation.

-If Coyne thinks that most of speciation is accident -except in some cases that could be directly driven by natural selection, then I suppose Coyne can't be labelled as "adaptationist", but more as "pluralist"...

-Has the "purpose of keeping the species distinct" ever been proposed as a cause or mechanism for speciation? Sounds strange to me.

Anonymous said...

In the U.S., it is interesting how the "ascendency" of the adaptationist stance is based, at least in some degree, on politcal and/or cultural bases. Many otherwise intelligent and agnostic/atheist non-scientists in the U.S. subscribe to all the tenets of modern biology, but they attach themselves to the adaptationist stance in large part b/c, in my opinion, the Gould-Lewontin wing of biological explication is, in their minds, tied up with leftist political philosophy.

I once saw Michael Shermer, a noted sceptic and atheist, give a talk to a group of Libertarians (I think it was the CATO Institute), and he basically said, "Well there was this big war in biology between the adaptationists (Dawkins, E.O. Wilson, Dennett, et al...) and those who wanted to impose their political views on Darwin's theories (Gould, Lewontin,etc..). And it went back and forth for a while, but the upshot was that the adaptationists won."

I'm not sure anything has changed much since then in American political/cultural discourse. But this seems to be a kind of benevolent form of accomodationism: a way for smart but culturally conservative people to embrace modern biology and keep goose-stepping to victorian mores.

Larry Moran said...

@ El PaleoFreak,

Coyne is a "reformed adaptationist". He's struggling hard to become a true pluralist but he ain't quite there yet.

He may have to give up the cowboy boots. :-)

Unknown said...

Well, if you physically separate two populations long enough, and prevent any gene flow, speciation will happen eventually due to this reproductive isolation. This is so with bonobos and chimpanzees who became separated by the Congo river.

However, are Coyne and Moran really suggesting that adaptation has not been the central force in the evolution of canids, for example? Is the coyote not a dog adapted to the desert and the Arctic wolf one adapted to the frozen tundra?

Random drift only makes sense when you are dealing with small populations: I take it Larry is aware of Kimura's diffusion approximation?

Corneel said...

El PaleoFreak said:
Has the "purpose of keeping the species distinct" ever been proposed as a cause or mechanism for speciation? Sounds strange to me.

This mechanism is often invoked to explain sympatric speciation. People who use such arguments usually assume that the incipient species have adapted to different niches, and that reproductive isolation limits the influx of maladaptive alleles from the other species.

Larry Moran said...

Reza says,

Random drift only makes sense when you are dealing with small populations: I take it Larry is aware of Kimura's diffusion approximation?

Larry is aware of lots of things including the fact that random genetic drift is an important force in evolution regardless of population size.

I take it you are unaware of basic population genetics which shows that the rate of fixation of nearly neutral alleles is independent of population size and only depends on the mutation rate?

You may be confused by knowledge that the probability of fixation of one particular allele by random genetic drift can be affected by population size. A result that also applies to fixation by natural selection.

Just out of curiosity, do you think random genetic drift has no impact on current human evolution? That would seem to follow from your question since the current size of the human population (>6 billion) isn't "small."

Larry Moran said...

Reza says,

However, are Coyne and Moran really suggesting that adaptation has not been the central force in the evolution of canids, for example? Is the coyote not a dog adapted to the desert and the Arctic wolf one adapted to the frozen tundra?

No pluralist would ever say that adaptation is unimportant.

I'd even be willing to consider the possibility that humans in Scandinavia are adapted to cold climates and humans in Nigeria are adapted to warm ones. Does this mean they are different species?

Jud said...

Larry writes: I'd even be willing to consider the possibility that humans in Scandinavia are adapted to cold climates and humans in Nigeria are adapted to warm ones. Does this mean they are different species?


Which for me raises a couple of interesting questions (interesting to me at least, since I don't know the answers). (1) What, besides geographic separation, is needed to create sufficient genetic distinction to result in speciation? (2) Does it seem a bit arbitrary that in sexually reproducing life, the designation of species has everything to do with germ cells, and virtually nothing to do with what can be quite gross differences in appearance?

Something else I find interesting is looking at Coyne's example and thinking of it as not only an example of speciation due to geographic separation (separation in space), but also over time. Since Coyne's two geographically separated populations can no longer interbreed with each other, at least one and perhaps both of these populations could no longer interbreed with representatives of the original population prior to separation.

Unknown said...

Larry,

Kimura's diffusion equation informs us that truly neutral mutations(i.e. alleles where s=0) have a probability of fixing that is 1/2N (for a diploid population).

Obviously, a high mutation rate does make it easier, but population size is still the dominant obstacle to fixation.

If you look at speciation in fruit flies, much of this is attributed to either founder effects or population bottlenecks - these can have a random dimension to them.

But once you have your new settler population, it will adapt to any new environmental conditions (as with Darwin's finches).

Yes, intraspecies adaptation in humans is impressive - we haven't speciated (i.e. become reproductively isolate) because of gene flow and constant migration.

Larry Moran said...

Reza says,

Kimura's diffusion equation informs us that truly neutral mutations(i.e. alleles where s=0) have a probability of fixing that is 1/2N (for a diploid population).

Obviously, a high mutation rate does make it easier, but population size is still the dominant obstacle to fixation.


Did you hear that "whooshing" sound? It was the sound of a point going over your head.

It's true that the probability of fixation of an individual allele depends on the population size. It's not true that fixation of alleles by random genetic drift only occurs in small populations.

Think about it.

But once you have your new settler population, it will adapt to any new environmental conditions (as with Darwin's finches).

It's quite true that separated populations can adapt to different environments. Nobody disagrees. What's your point?

Remember we're talking about true biological speciation events where the two populations become species because they are incapable of producing viable offspring.

That's something entirely different. I tried to alert you to this by giving you an example of two human populations that may have adapted to different environments but are not species.

The question is whether evolution of biological reproductive isolation is more likely due to the effects of natural selection or random genetic drift.

So far you've done nothing to address the question except to say, "Random drift only makes sense when you are dealing with small populations:..."

Is that your only contribution to the discussion?

Paul McBride said...

Reza says:"Kimura's diffusion equation informs us that truly neutral mutations(i.e. alleles where s=0) have a probability of fixing that is 1/2N (for a diploid population).

Obviously, a high mutation rate does make it easier, but population size is still the dominant obstacle to fixation."


The point Larry is making - and it comes straight from Kimura - is that the probability of a single neutral allele going to fixation decreases with population size, but because the number of mutations increases in a large population (not the mutation rate, but the raw number) there should be no effect on fixation.

Thus the mutation rate, independent of population size, should control the fixation rate for neutral alleles.

Population bottlenecks are potentially important because the speed of fixation (for any given allele that does fix) is increased, and because such phenomena have other effects on non-neutral alleles.

Unknown said...

Larry,

If speciation is mostly cladogenetic - i.e. a few splitting off from the many and becoming isolate - then random drift *will* be a major factor in the evolution of the new lineage and could lead to the biological change you have in mind.

One of the most important differences, as far as speciation is concerned, are chromosomal re-arrangements. Red foxes cannot breed with dogs because of their chromosome number variation: this is due to fusions and the like. I would suppose that this was down to drift rather than selection.

And on the subject of dogs: none of the breeds around today would have been allowed by natural selection: many dogs are just deformed descendants of wolf-like ancestors.The variations we have selected are cute but do not contribute to reproductive fitness. But all dogs are the same species of >canis lupus familaris and can interbreed with each other successfully.

This is probably because the dog and its various breeds have not been around too long.

El PaleoFreak said...

Corneel said:
"People who use such arguments usually assume that the incipient species have adapted to different niches, and that reproductive isolation limits the influx of maladaptive alleles from the other species"

Of course, but that can hardly being described as "purpose of keeping species apart".

Psi Wavefunction said...

Larry, to be fair to Reza (and I *really* don't want to, trust me), many intro pop gen books do make the outrageous claim that drift applies only to 'small', ie "endangered", populations. I have one that also proceeds to claim humans have a "large" eff pop size, which is zoocentric to the point of being hilarious.


Sure, the probability that a given new allele gets fixed by drift in itself can be quite low, but it is often ignored that there's a shit ton of loci upon which drift acts, and thus the total force of drift becomes quite significant. Particularly when you factor in that VERY few mutations have any gain in fitness, and most are nearly neutral at best. In that context, suddenly drift becomes even more important. Nor does drift stop acting once selection gets its hands on something or vice versa. Nor does adaptation preclude the involvement of non-adaptive processes along with it. Nor do any vertebrates have any particularly "large" effective population sizes, compared to the vast majority of life on earth that happens to be microbial (and quite well off success-wise, however you define that). But those points are apparently not cool enough to be emphasised in textbooks, so all we get is "no interesting organism has a population size small enough for drift to do anything". And few of them are interested in the slightest to hear anything else.

It's getting to the point where I just don't bother talking evolution to people who's evolutionary views I disagree with anymore. It's just not worth it. Not that it helps the problem or anything, but there's only so much time and patience in a day...


Reza, RE drift and real populations:
http://www.indiana.edu/~lynchlab/LynchPublications.htm
All freely accessible. You have no more excuses for your obnoxious ignorance. Happy reading!

Anonymous said...

Just passing by.

In my view, there's some confusing language in this post, Larry. When you write about the role of adaptation in driving divergence, you need to differentiate that usage from the points you make about the "adaptation" of old -- as in "adaptationism"; e.g., arguments about scientific agendas/worldviews from the 80's.

There's little or no debate that adaptation drives the incompatibilities that underlie divergence (as by-products or whatever). See "ecological speciation".

Larry Moran said...

anonymous says,

Just passing by.

In my view, there's some confusing language in this post, Larry. When you write about the role of adaptation in driving divergence, ...


Why don't you stay awhile? You may learn that I was not talking about "divergence." I was talking about the genetic changes that lead to biological reproductive isolation.

There's little or no debate that adaptation drives the incompatibilities that underlie divergence (as by-products or whatever).

Of course there's debate about the causes of differences between species. It's not all due to adaptation. Haven't you ever thought about the famous Asian and African rhinoceros that Richard Lewontin talks about so often?

And what about the various species of zebra? Are all those differences in stipes and coloration adaptations?

El PaleoFreak said...

"And what about the various species of zebra? Are all those differences in stipes and coloration adaptations?"

Those are the differences we use to identify the species, but perhaps they weren't "key" characters when the genetic barriers arose. There are other differences in zebra species, social behavior differences for instance. We don't know yet how and why zebra speciation happened.

Corneel said...

El PaleoFreak said:
Of course, but that can hardly being described as "purpose of keeping species apart".

Sure it can, because in this scenario reproductive isolation serves an adaptive purpose and natural selection will work to "keep the species apart".

But I understand what you mean. From experience I can tell you that many students (and some researchers too) have the unspoken conviction that species "want" to split all the time. It is usually a shock when I remind them that any allele causing reproductive isolation is in principle deleterious, because it limits the reproductive output of an individual.

Unknown said...

Psi Wave:

If you bothered to actually read what I wrote, you would understand that I do not envisage small populations being necessarily "endangered" ones, but those that have split off from the superset.

So, if I and a few other humans set off to colonize Mars, our population would be much more subject to random drift than the people of China with its large breeding population size.

There may be a danger of mutational meltdown due to the fixation of slightly deleterious alleles, but I would like to think that our descendants would evolve to be intellectually and physically superior than those back on Earth.

Unknown said...

Larry, ok I'm staying. Love you're doing here, what I'm seeing of it.

Didn't say it's all due to adaptation. Said it adaptation was inarguably important in causing the "genetic changes that lead to biological reproductive isolation". Drift does what it does too. Are you writing about relative importance. Welcome to the party!

No one who is seriously debating that natural selection (read: adaptation) causes the incompatibilities that ultimately result in speciation. Why confuse things, especially if this isn't your cup of tea? See Schluter & Conte 2009 PNAS 106: 9955-9962 etc etc, or just Darwin will do. Yea, there's a lot of holes to fill. But that's a good thing.

Larry Moran said...

Patrick says,

No one who is seriously debating that natural selection (read: adaptation) causes the incompatibilities that ultimately result in speciation.

If I read this literally then it's wrong. If by "speciation" you mean the actual event of biological reproductive isolation then there's lot's of debate. In fact, most of us are convinced that adaptation is hardly ever the cause of that kind of speciation event.

Perhaps you meant to say something else?

David Winter said...

the actual event of biological reproductive isolation

Event? Speciation is not an event, it's a process. Once they are isolated, populations are capable of diverging, and, as Coyne says, local adaptation often speeds that divergence.

If those species come back into contact and fail to hybridise then selection didn't act to keep those two species apart, but it might have been the cause of the particular changes that had that effect. Much more interesting are the cases in which hybrids develop just fine but face an ecological penalty, in which case selection might directly act to "finish off" a speciation process.

The Other Jim said...

Maybe we need to ground the discussion. Here is one example...

http://www.cell.com/abstract/S0092-8674%2808%2901385-8

Summary (aka abstract)
Hybrids between species are usually unviable or sterile. One possible mechanism causing reproductive isolation is incompatibility between genes from different species. These “speciation” genes are interacting components that cannot function properly when mixed with alleles from other species. To test whether such genes exist in two closely related yeast species, we constructed hybrid lines in which one or two chromosomes were derived from Saccharomyces bayanus, and the rest were from Saccharomyces cerevisiae. We found that the hybrid line with Chromosome 13 substitution was completely sterile and identified Aep2, a mitochondrial protein encoded on Chromosome 13, to cause the sporulation defect as S. bayanus AEP2 is incompatible with S. cerevisiae mitochondria. This is caused by the inability of S. bayanus Aep2 protein to regulate the translation of the S. cerevisiae OLI1 mRNA. We speculate that AEP2 and OLI1 have evolved during the adaptation of S. bayanus to nonfermentable carbon sources, thereby driving speciation.

Anonymous said...

All sexual populations are perennially in a state of internal fragmentation. This is because sex does not instantly connect the entire group and keep them all on the same page, as per the viscosity-free pop-genetic ideal (from which such canards as 'drift only affects small populations' arise - mathematicians understand sampling error and convergence on the mean, which depend upon sample size). In real populations there is a lag, since the population only coheres, genetically, by being hooked up couple by couple in a geometric area. Drift can work just fine at a local level. Anything that causes a reduction in flow, even an incomplete one, can result in divergence, regardless whether the alleles are being locally concentrated by drift or selection. To regard speciation as an adaptation (rather than the result of adaptation, such as two ecotypes arising) would require that individuals who prefer mates on their side of the divide leave more offspring than those that hybridise.

Dharmit said...

So does speciation occur due to adaptive transformation or does adaptive transformation occur more rapidly due to speciation?