Evolving Thoughts

Evolution, culture, philosophy and chocolate! John Wilkins' continuing struggle to come to terms with impermanence... "Humanus sum, nihil humanum a me alienum puto" - Terence

Saturday, December 24, 2005

Speciation genes

Let's recap. Following Gavrilets' suggestion, we have concluded that geography is a minor component in the conceptual taxonomy of speciation. The major component is gene flow, and geographical classification merely goes to the nature of the selection involved; whether it is intrinsic (sexual selection, lowered hybrid fitness, karyotypical incompatibility) or extrinsic (ecological). If extrinsic, then the genetic structure of the species can change in ways that inadvertently cause reproductive isolation.

I introduced the notion of reproductive reach - the genetic or developmental "distance" within which fitness is lowered insufficiently to prevent introgression between populations. Some researchers, such as Chung-I Wu and, as RPM notes in the comments to the last post, Will Provine, seek to find "speciation genes" which are modified through this inadvertent selection. This is, I believe, a category mistake, and a logical fallacy.

The category mistake is to presume that because a genetic distance causes speciation, it is therefore a gene "for" speciation. But there is no prior specification of genes that cause reproductive isolation. A genetic change may do so, or it may not. Identifying that it has done so is something that can only be done post hoc. And there appears to be no particular genes that cause speciation over large evolutionary distances - there may be an active gene complex in Drosophila which when changed causes reproductive isolation, but it doesn't therefore follow that a homolog of that complex will do the same thing in other flies, or in insects generally, or in all animals, etc. In fact, it doesn't even follow that we will find this is the complex involved in all cases of Drosophila speciation, either.

Whether or not genetic differences cause isolation depends on the entire complex of the organisms involved. The heterogeneity of the RI chart from Littlejohn shows that there are no singular reasons why interfertility is reduced in all organisms. Finding one that does is not therefore inductively generalisable over all members of the clade - this is the logical fallacy. If the complex G causes RI in species-pair S, to assert that it is therefore the active cause in species-pair S' is a fallacy of affirming the consequent. Or it is a reasonable assumption, just to the extent that S' is closely related to S and there are prior reasons for thinking that G is active in both cases (like mutability of G in that clade).

There is an overwhelming tendency amongst biologists to overgeneralise their results, and this is understandable. Science is about making generalities to cover large phenomenal domains. The trick is to generalise just so much as the data and hypothesis allow, and no further, and of course Nature isn't giving away any hints about how far is too far. But we have seen a "vertebrate bias" among zoologists, a "zoological bias" amongst geneticists and evolutionary biologists of yore (many of whom were animal biologists - Mayr was an ornithologist, Dobzhansky an entomologist, Simpson a vertebrate paleontologist. Verne Grant was the botanist, and his contribution was an afterthought). But evolution is about all living things. General conclusions should be founded on general data.

So we should reject the a priori idea that there will be "speciation genes". Speciation involves entire genomes, plus of course the developmental system in which they are expressed, plus the ecological context in which genes are "normal".

Next, I aim to summarise this in a single chart, and suggest some implications for species concepts, in particular the genetic cohesion concept of Alan Templeton. But this may have to wait until the alcohol wears off, and I get back to the office and my references.

Happy Summer solstice, everybody. For those of you in the Northern hemisphere, sorry about your lack of timing.

Late Correction: It was RPM not RBH who made the observation. My mistake.

Friday, December 23, 2005

Relating speciation

Two issues that face an evolutionary biologist rather sharply are these: the nature of species and how they come to be. When in the period of the fusion of Mendelian genetic and Darwinian evolutionary theory these matters were raised, initially by Dobzhansky, the obvious way to tackle the problem was by making use of th enew science of genetics. Scientists will not tackle a problem for which there are no tools, although they may speculate on problems. But give them a new tool, like Mendelian genetics, and they are off and running.

But this has an unwanted aspect - it can lead scientists to frame the problem solely in terms of the tools. And so it was with the matter of speciation and species. Species were, Dobzhansky informed us, largely terminal stages in the separation of gene pools. Mayr further extended this by defining species solely as populations that were permanently isolated from exchanging genes.

And this is true, for sexual species, although it is not always the case that genes are the mechanism for reproductive isolation. Behavioural and developmental processes can isolate populations while massive differences in genetic structure and alleles can be included within one species. And the nature of the term "population" is suspect itself. A population is a partition of the gene pool of a species (you don't define a species in terms of populations that have no chance of reproducing together, like asparagus and elephants), but what counts as a population, or the reproductive equivalent of a population, the deme, depends on whether they are in the same or similar enough species. Species define populations, and populations define species. Perhaps there's enough shared operational understanding in the various disciplines to make non-tautological sense of the two terms, but on the face of it "population" is merely one unanalysed term defining another.

In our diagram, several speciation modes are undistinguished, in part because what makes them different has nothing much to do with geography. Geography is only important for those speciation modes that rely upon it, like sympatric reinforcement or allopatric drift (i.e., evolution of the genetic structure of the population in isolation). But in chromosomal rearrangements, such as inversions, fusions, fissions, and duplications, geography has no real role to play as a mechanism of speciation, and is only useful as a test afterwards of whether or not speciation is complete enough to count two species where there were one.

Then there are asexual, or mostly asexual, species. Simpson proposed not to call them species because they failed to rely even upon reproductive isolation, while A. J. Cain famously invented the notion of an "agamospecies" for them. But I like Manfred Eigen's notion of a "quasispecies" (from "qua si", as if) in which there is a genetic coordinate in the space defined by the genomes of these beasties that is the default, or "wild-type" (wt), even if no strain of them actually instantiates it, around which all the others cluster. This naturally raises the question why they cluster there, and in the case of viruses, for which he developed the notion, the reason is clear. They cluster about the most efficient utiliser of the host cell resources. In effect, the wt is a fitness peak, and variants that stray too far will be of lowered fitness (unless they cross a fitness valley and start moving up the new slope).

In the case of selectively reinforced speciation, of lowered hybrid fitness, this is also the case. These populations or variants will be maintained by selection for some environmental niche or mode of living. So we can perhaps see that reinforcement of species here is effectively equivalent to quasispecies, only now, being sexual, it is not the elimination of strains that are not near the fitness peak, but of hybrids who are between the peaks.

The same thing is true of host race speciation - here the selection is for the host resource availability. So the common element here is selection, and we can perhaps collapse them into one mode, differentiating them on the basis of what the resources are, ecologically speaking.

Competitive speciation - Darwin's favourite - is likewise a case of selection, but what about karyotypic races and hybridisation? Or sexual selection? In each case, what counts here for fitness is whether there are mating opportunities. If you have a karyotypic difference that happens to match sufficiently with the ancestral karyotype, then speciation does not occur. But if the changes make it harder to mate, then eventually mating with those of similar karyotype is fitter than mating with the parental form. In hybridisation the same thing is true - the hybrid form often cannot easily mate with either parent successfully, so it forms its own species.

Sexual selection is clearly about being able to mate. The major difference here is what does the selecting. Rather than it being the extrinsic environment, it is the intrinsic one. If you cannot find a mate among one morph of the population, but you can with the other, then eventually the lowered fitness of those that try to mate both ways will tell.

And that leaves Vicariant evolution, misleadingly called "allopatric speciation" or speciation by geographical isolation and independent evolution. Originally proposed by Moritz Wagner in the 1860s, it is the form of speciation most defended by Mayr and those in his research tradition. There is selection going on, of course, but not against lowered fitness of some morph or allele or karyotype, but solely selection for local conditions, which forces genetic changes that eventually make it difficult if not near impossible (nothing physically possible is actually impossible in biology) for the now sympatric populations to interbreed. In short, so far as reproductive compatibility is concerned, the two populations just... drift. Selection maintains them at local fitness peaks, but they acquire inadvertent changes that make them reproductively isolated when they get back together, if they do.

I want to propose a notion here I shall call "reproductive reach". Any two populations have a reproductive distance that is locally defined and context-dependent. It determines how much gene flow can occur between them if they are sexual reproducers. And it specifies for any two organisms whether their progeny will be fit enough to continue a reproductive lineage. The acquisition of reproductive isolation depends on whether the conditions leave the two organisms outside the reproductive reach of each other. It need not be absolute, but if the fitness of a progeny of any organism is severely lowered (and having none at all is about as severe as it gets) then that form or variant is outside the reproductive reach of its parent or parents.

Now this has an interesting consequence. A sexual species becomes just a special variety of asexual species. Where the fitness peak on asexual quasispecies is entirely extrinsic, the fitness peak of sexual species can be either extrinsic or intrinsic. It's about, in the end, how many progeny you can have. Quasispecies is the primitive or basal notion of specieshood. Sexuality just adds dimensions to ways of being species. Vicariant speciation relies on local fitness peaks, but put them back in sympatry and it's all about the fitness determined by the reproductive reach of the two organisms involved.

Since we can now say, with Gavrilets, that drift and selection are not mutually exclusive (because, remember, there are fitness ridges in any complex genome space along which a population can take a Brownian walk) I say that all speciation is, indeed, about selection, either ecological, sexual, or developmental. So we can dispense with geography in our taxonomy of speciation altogether, except as an index to differentiate the kind of process that modifies reproductive reach.

Does this make Darwin right and Mayr wrong? Of course not. Darwin was very vague about the modalities and reasons for speciation, and he thought that direct reinforcement of diverging selection mostly drove it. It is entirely probable, as Coyne and Orr document, that Vicariance drives most sexual speciation in animals (but not necessarily plants - or gamete broadcasters in the animal world either. If you sow your seed, so to speak, broadly, then what maintains your isolation will be selective). But geography is a secondary aspect of speciation. And it need not be the most important for all groups of organisms.

So the allopatry-sympatry axis of percentage gene flow is the spindle, as it were, around which myriads of particular processes acquire reproductive isolation by exceeding the reproductive reach of the original genomes. Geography's in there, as are chromosomal rearrangements, and host race speciation, etc. We need to revise our terminology to avoid the useless disputes of the past about what "the" mode of speciation is. What counts in the end is the rather obvious fact that the number of progeny, relative to alternatives, makes a new species.

Wednesday, December 21, 2005

Fuller, creationism and philosophy FOR science

Steve Fuller's testimony and deposition in the Dover case was ultimately self-defeating - Jones cited him 11 times with the take-home message that yes, ID is creationism. What I find amusing is that the claim Fuller made that religion has been part of science and so it ought to be allowed to be inserted into science made no headway at all, despite the overlong screed he offered up. On a philosophy of science list, Fuller post a note entitled, "Why the field is called 'philosophy of -- not for -- science'". And this marks out what is essentially the problem with sociological approaches - unless you have some reason for thinking that the subject or society under question is intrinsically interesting or worthwhile, it's no problem to encourage the interjection of foreign elements. Fuller is not "for" science, so to him it's no problem to introduce religion - after all, it's just another "worldview", isn't it?

Unlike Feyerabend, whose call in Against Method for the inclusion of creationism was, arguably, a jocular attempt to do what we Australians call "stirring the possum", Fuller believes that there is fundamentally no difference between religion and science. Or at least, he will think this until religion circumscribes the sort of sociology he is allowed to do. If he is fine with this, that's his choice. But it points up the problem - science is not just another worldview. In fact, it's not a worldview at all. There's no such thing in science as "metaphysical naturalism", just the somewhat obvious point that if something leaves no evidence or makes no difference to the phenomena, then it can't be investigated by science. Fuller thinks that sources of inspiration are somehow part of science. And some are, if they are heuristics that have gained a general acceptance in the disciplines. But religion, typically, is not, and the history of modern science is one of disengagement from religion, and politics, and fashion. These things hinder scientific learning.

If anyone wants a good and sensible overview of the relationship between religion and science since the beginning of the scientific "revolution", see John Hedley Brooke's Science and Religion: Some historical perspectives, Cambridge University Press 1991. This overcomes both the "science at war with religion" tradition, explicates the real religious underpinnings of early science, and manages to make sense of it all without retreating to an "everything is the same" or "science is Christian" position.

Late note: Fuller, in respeonse to the suggestion that he was perhaps the best witness for the evolution side, said this in the HOPOS list -

R. J Murphy:
Have been reading through Judge Jones decision in the Dover PA case for references to the testimony of HOPOS list favorite Steven William Fuller. There are about a dozen such references. Most are variations on, "Moreover and as previously stated, there is hardly better evidence of ID's relationship with creationism than an explicit statement by defense expert Fuller that ID is a form of creationism." So either Steven was secretly playing for our team, and list members owe him a debt of gratitude, or he was just a royally shitty defense witness.

Judge Jones criticizes Fuller's attitude towards ID at one point as follows: "Science cannot be defined differently for Dover students than it is defined in the scientific community as an affirmative action program, as advocated by Professor Fuller, for a view that has been unable to gain a foothold within the scientific establishment."


Steve Fuller replied:
You seem to have left out a third possibility, the one that strikes me as most obvious: namely, I said what I actually believed about ID, which was supportive but not unequivocally so, and the judge selected the parts of my testimony that supported the judgement that he had reached on the balance of the evidence and testimony.


The defence rests.

Dover judge says ID is creationism

Judge Jones has ruled in favour of the plaintiffs - that is, in favour of the parents who objected to teaching ID or disclaiming evolution in textbooks - without qualification. It is creationism in disguise, supported by a lie, and should not be taught in science class. It will take me a while to digest this, so I recommend you go to Panda's Thumb, Dispatches from the Culture Wars, or Pharyngula for comment.

The [Lack of ] Discovery Institute is already spinning:
Apparently Judge Jones has forgotten what Justice Jackson said in the flag salute case:
"If there is any fixed star in our constitutional constellation, it is that no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion." West Virginia Bd. of Ed. v. Barnette, 319 U.S. 624, 642, 63 S.Ct. 1178, 87 L.Ed. 1628 (1943)
Although there was some fascinating testimony by ID scientists such as Michael Behe and Scott Minnich, and many interesting aspects to the case, Discovery Institute and many other supporters of Intelligent Design saw this as a poor test case for ID and predict that this will be just the first court case on the Intelligent Design, not the last.
This is, of course, a political fight. To them the law is just a tool to impose their own political influence. A loss here has no substantive implication, like, for example, ID is a load of religious dogma, but is only a temporary setback. So disparage the judge, call him an "activist" judge (which we know is code for "fails to get with the program" in right-speak), and keep the faithful hopeful. After all, gotta keep those donations rolling in.

Tuesday, December 20, 2005

Season's Greetings

As part of the Atheist Attack Upon Christmas, I bring you this. So now we know roughly what time of year the K-T Bolide came...

Monday, December 19, 2005

A word about this blog

My Mirecki posts came in for a bit of flack on several sites. I expected this. Of course they took suggestions about inference and hypothesis to be substantive claims. I did not say that I thought the police were involved. In fact I think it unlikely, but more likely than the much less probable claim that Mirecki made it all up.

But reasoning isn't a major virtue in the worldview of some, so let it pass. One criticism I got a bit was how I know that it's hard to be an atheist in Kansas, sitting over here in Australia. That's a good question - I know it because atheists from Kansas who I know in correspondence say it is, and based on similar experience here in Australia about how those who fail to adhere to the prevailing religious line are treated, including myself to a limited extent (I'm not an atheist but an agnostic. Try explaining that difference in suburbia, where I have lived for the past 30 years).

I have seen kids isolated, beaten and bullied by students and teachers alike. I have seen this with my own kids in an Anglican school (worst mistake I ever made was to believe the "policies" of a Christian school about tolerance). I have every good reason to think it is harder in Kansas. I'm not making it up - I am basing it on evidence.

Finally, I am challenged to explain how I know that Mirecki was forced down from the chairmanship of his department, and there I can only say, try working in a university environment. Anyone who thinks it was all sweet reason and manners has never seen academic viciousness. I am presently the collateral damage in a case of it.

On this blog I post stuff I have passing through my head. Some of it is well thought through, some isn't. If I am wrong about Mirecki - and I'm going to need a lot more than the unsupported assertions of the conservative side of things (which is funny, because I'm what an Australian would call a conservative, which has nothing much in common with what American's call conservative these days) - I'll retract and revise. A reasonable person does this.

What interests me so much more than the asseverations and tribal loyalties of this side or that is the nature of knowledge about the world. Hume set that agenda and that is what I am most concerned to explore. We do not know the world by stating our beliefs forcefully and refusing to tolerate dissent. We know it by testing our claims against evidence.

And I agree with one comment on a blog that I shall not dignify with a link (you can find it by Googling if you need to) - my tautology FAQ on talkorigins.org is terrible. I was an undergraduate when I wrote that, struggling to deal with the literature. One day, when I get the time, if ever I do, I will rewrite it completely. But anyone who wants to do so can do so.

Modes of speciation 2: flow versus place

[I'm sorry this has been delayed. I moved house...]

Gavrilets notes that one can use any of the isolating processes to classify speciation, but that the traditional way is to frame it in terms of migration between diverging populations (which is to say, it's what Mayr said in 1942). He lists the varieties I mentioned last time: allopatric, sympatric and parapatric, and aligns them along a single axis of migration rate:

Of interest here is that some of the standard forms of speciation don't appear. For example, speciation by chromosomal rearrangement ("stasipatry", defined by Michael White in the 1970s) is missing. It is clearly in sympatry, but there is no "migration" between the new population and the old (except in cases like the Sphalax ehrenberghi mole rat in Israel, which has major chromosomal races, some of which are mutually infertile, but there is introgression between them). Polyploidy - or the duplication of entire chromosomal complements, common in plants and arguably in animals - is also missing.

Gavrilets deals with them by calling them "scenarios" and "patterns" of speciation. He lists the following cases
  • Vicariant - divergent selection and stochastic factors like drift after division of a population by extrinsic factors such as geographical changes;
  • Peripatric - a small subpopulation, mostly isolated, at the extreme of the parent range. The idea is that it will have both a non-standard sampling of alleles, and also be subjected to divergent selection pressures in extreme environments (for that species);
  • Centrifugal - central populations that carry a sample of many alleles which become isolated through, say, "island" formation (such as the mountain "islands" in the Amazon);
  • punctuated equilibrium - the appearance of relatively rapid speciation and subsequent stasis as the population reaches equilibria of alleles. This is a "pattern" rather than a "process" as such, and can be caused by any of the other scenarios;
  • Chromosomal speciation;
  • Hybridisation - including allopolyploidy. In this case the genetic complement of two species is paired up by a loss of secondary division, giving a symmetrical set of chromosomal strands;
  • Reinforcement - once hybrids are of lowered fitness for whatever reason, selection will tend to reinforce separation of the gene pools (for example, a hybrid rock and grass dwelling lizard might be less able to survive in either environment as well as the "pure" lines);
  • Competitive - this is Darwin's scenario. Members of a species that are in strong selection for a limited resource may result in specialised forms that are thus in less competition with the ancestral forms that make use of some other resource;
  • Clinal - he calls it "speciation along environmental gradients", where limited migration and selection leads to aggregation of forms at the terminal ends of the cline;
  • Host shift - this is the famous case of the Rhagoletis fruit flies I mentioned, that Stuart Berlocher and colleagues have studied. Host fidelity replaces geographic isolation;
  • Runaway sexual selection - this is obvious.
Now, this is a somewhat heterogeneous list, so I wondered how to assign them to the scale above. For example, many of these are in full sympatry - chromosomal rearrangements and polyploidy, host shift, competitive, and so on. Is there a way to separate these modes conceptually? I therefore drew up this chart:


Replacing percentage of gene flow for migration rate and geography as the other axis, we now get a clear picture - sympatry-allopatry is not the only important difference between these "scenarios" or modes of speciation. So also is the relation of the populations to each other. I have placed the sexual selection example between peripheral and isolated because I doubt it would occur species-wide, and is therefore less likely to be a central population case, but it could occur in revenant populations.

We are left with two groups of modes - reinforcement and host race shift (in the case of almost-sympatric parapatry: again it need not be so specific), and competitive, hybrid and karyotypic. Each of these is distinguished by a particular mode or mechanism of speciation.

In my next post I will explore some of the implications of this for the speciation debate, and also for the definitions of "species" used by biologists.