Re: The evolution of "evolution"

From: John Wilkins (
Date: Fri 28 Oct 2005 - 12:28:27 GMT

  • Next message: Robin Faichney: "Re: Me against the meme"

    As I understand it, and I am neither a botanist nor a geneticist, but I have read a fair bit of the speciation literature, plant speciation is massively due to allopolyploidy - in ferns and fern allies
    (gymnosperms), it has been estimated as much as 97% of species are formed either from auto- or allopolyploidy, and angiosperms around 75% (but Coyne and Orr reject this figure, offered by people like Verne Grant and Warren H Wagner, in favour of their preferred allopatric model based on the reproductive isolation concept of species misleadingly called the "biological" species concept).

    Sergey Gavrilet's book on speciation does point out that there is a definitional problem with the allopatric model - it ranges from 100% sympatric to 100% allopatric, but mostly demes and gene pools are in some kind of parapatry. However, the standard view is that reinforcing selection against hybrids is post-speciation rather than a cause of it. Moreover there are numerous cases, particularly in flowering plants, of introgression between good species. So much so that the Botanical conventions allow for a term
    "compilospecies" (plundering species) for species that continuously introduce genetic material from nearby species without losing their integrity as taxa.

    As to species concepts - there are, WRT speciation, really only two concepts here that matter - one presumes reproductive isolation for sexual species (the phylogenetic concept; actually there are three, but one of them is more a diagnostic concept than an ontological one), and the other treats reproductive isolation for sexuals as the definition of specieshood. For asexuals (well, for primarily asexual taxa, like protists that are not gene exchangers) the relevant concept is *all* about selection.

    There's a lack of hard evidence about relative rates. So far most of the discussions have been about the conditions under which the realtive kinds occur, and the problem is that we cannot say without investigation - that is, not a priori or on theoretical grounds - how often those conditions occur. Coyne and Orr tend to assume that the sexual cases are common, and so assert that allopatric speciation is common. Gavrilets, Schilthuizen, Berlocher and others offer cingular cases or theoretical arguments as counters. But we simply do not know; and it is a theoretically hard domain to find this out.

    For myself, I think that most sexual speciation that is not polyploidy (and in this I include partial polyploidy like karyotypic difference, as in the "stasipatric" speciation MJD White proposed back in the 70s) is allopatric or at the allo end of parapatric, and mostly due to contingent evolution. But this might end up the equivalent of saying that only a small proportion of *all* speciation is allopatric. In any case, little of the speciation process will turn out to be due to ecological selection. A lot will turn out to be due to accidental (that is, selectively uncorrelated), and some as yet undetermined proportion will be due to sexual selection (runaway superstimulus selection).

    Hope this helps, rather than hinders.

    On 28/10/2005, at 10:07 PM, Chris Taylor wrote:

    > Hmm. I get what you're saying but this comes down to definitions of
    > species (and by extension, speciation) really. The inability to
    > interbreed at all where that occurs is often a product of drift/
    > rearrangements/founder, and of course directional _selection_, but
    > it is the 'no role in most cases' thing I take issue with; consider
    > the plants that form (almost continuous) sequences of whatever-you-
    > want-to-call-them; allochronic flowering or divergence of flower
    > structure is absolutely selected for and fairly rarely is there
    > complete isolation. Also consider parapatric in animals rather than
    > sympatric -- at the boundary there is definitely reinforcement
    > under selection by call/smell/shape of bits or whatever to prevent
    > gene flow (driven obviously by the drift/rearrangements/dir.sel. in
    > the bulk of the population), without which no true speciation could
    > be said to have occurred.
    > Full-on allopatric is fine, but even then when the species come
    > back into contact reinforcement shows its head to avoid fruitless
    > matings (under selection), and reinforcement is pure selection.
    > _Drift_ in flowering time (for example) could have the same effect
    > though of course.
    > Anyway it was not the substance I took issue with, rather the
    > degree (which I have to admit is hard to characterise).
    > Hair-splitters Inc. :\
    > Cheers, Chris.
    > John Wilkins wrote:
    >> Most speciation is thought to occur through geographical
    >> isolation and subsequent evolution, mostly by drift, founder
    >> effect sampling of the original gene pool and selection for local
    >> adaptation. But the selection here is not speciating selection
    >> most of the time. Speciation is a by-product of evolution of the
    >> isolate population that results in reproductive isolation when
    >> back in sympatry.
    >> The type of speciation in which selection plays a role *in
    >> causing speciation* is sympatric speciation. In this case
    >> variants within a local population adapt to divergent fitness
    >> peaks, and so results in divergent selection, leading to lowered
    >> fitness of hybrids. But most of the time this is caused more by
    >> sexual selection than ecological adaptation. And it requires
    >> quite rare circumstances.
    >> Darwin thought that most speciation was caused by divergent
    >> selection but it seems not, at least in sexual organisms, to be a
    >> major factor. Selection causes adaptation, but adaptation doesn't
    >> drive most speciation events.
    >> On 28/10/2005, at 7:28 PM, Chris Taylor wrote:
    >>>> But of course most speciation now is in fact thought to occur
    >>>> through random variation and random fixation rather than by
    >>>> selection as Darwin thought. There's good reason to think that
    >>>> some speciation is due to selection, but not much. I worry
    >>>> that we think only that Darwinian evolution is about
    >>>> selection (natural or sexual), when in fact another really
    >>>> deep aspect of his view is common descent, and this is not
    >>>> tied now to selection.
    >>> ???
    >>> Selection has _no role_ in the generation of species the
    >>> majority of the time? Are you just purely talking about
    >>> permanent absolute allopatry / completely discrete allochrony or
    >>> whatever equivalent you care to pick?
    >>> Elephants and fleas will never successfully mate (having
    >>> diverged somewhat); but where this matters (i.e. in recent
    >>> speciation events, where those species ranges [or whatever]
    >>> overlap) selection is key in ensuring that hybrids are (1)
    >>> demonstrably crap and that (2) parents who find a way to avoid
    >>> sinking their genes into such crappy hybrids propagate more of
    >>> those genes forwards to subsequent generations..?
    >>> Random variation and fixation is _not good_ at producing
    >>> adaptation without selection. Have I misunderstood you?
    >>> Cheers, Chris.
    >>> ~~~~~~~~~~~~~~~~~~~~~~~~
    >>> ~~~~~~~~~~~~~~~~~~~~~~~~
    >>> ===============================================================
    >>> This was distributed via the memetics list associated with the
    >>> Journal of Memetics - Evolutionary Models of Information
    >>> Transmission
    >>> For information about the journal and the list (e.g. unsubscribing)
    >>> see:
    > --
    > ~~~~~~~~~~~~~~~~~~~~~~~~
    > ~~~~~~~~~~~~~~~~~~~~~~~~
    > ===============================================================
    > This was distributed via the memetics list associated with the
    > Journal of Memetics - Evolutionary Models of Information Transmission
    > For information about the journal and the list (e.g. unsubscribing)
    > see:

    John S. Wilkins, Postdoctoral Research Fellow, Biohumanities Project
    University of Queensland - Blog:
    "Darwin's theory has no more to do with philosophy than any other
    hypothesis in natural science." Tractatus 4.1122
    This was distributed via the memetics list associated with the
    Journal of Memetics - Evolutionary Models of Information Transmission
    For information about the journal and the list (e.g. unsubscribing)

    This archive was generated by hypermail 2.1.5 : Fri 28 Oct 2005 - 12:47:32 GMT