"Icons of Evolution"

From: pruest@pop.dplanet.ch
Date: Tue Jun 05 2001 - 13:12:22 EDT

  • Next message: William T. Yates: "Re: "Icons of Evolution""

    I sent this already to Tim Ikeda, but forgot to add the list as a
    recipient. So here it is.


    It seems to me that occasionally we are talking past each other. Whereas
    you are concentrating on the macroscopic end of the scale (morphology),
    I am primarily looking at the microscopic end (molecular biology). Even
    concepts like "compound system" or "component" tend to be taken to
    designate quite different entities, unless concrete examples are named.
    Although a macroscopic body is built up hierarchically through a series
    of smaller component levels (something like: atom - active site -
    molecule - molecular complex - organelle - cell - tissue - organ - body
    part - body) and the evolutionary principles hold somewhat continuously
    throughout this hierarchy, there are quite substantial differences
    between the two ends. At the lower end, it is usually possible to
    understand function in detail, but at the upper end this is usually not
    possible and much remains speculative. Selection usually operates at the
    level of the individual organism (and its entire developmental
    program!), never directly at a lower level. As far as selective forces
    are concerned, it is therefore difficult to be sure of what happens: the
    relationship between mutations and selection is very indirect.

    > [...]
    > Peter:
    > >>[big snip...]
    > >>>3) The homology of vertebrate limbs: the main problem with all
    > >>>similarities between functional features is that they may need to be
    > >>>similar in order to function: thus, they cannot be evidence for common
    > >>>descent.
    > Tim:
    > >>I would agree with this if a singular component was being used for
    > >>comparison. For instance, there are examples of enzyme active sites where
    > >>strong arguments can be made for convergence rather than homology.
    > >>However, when the item being considered for homology is itself composed
    > >>of multiple components or features, as vertebrate limbs are, then I think
    > >>similarities of the component parts can be used to make a legitimate case
    > >>for common descent. I think that with more components available to produce
    > >>a "function" there would tend to be more ways to produce that functionality,
    > >>particularly in macroscopic systems such as limbs. That would tend to
    > >>undercut the argument for "similar functions requiring similar forms"
    > Peter:
    > >You are right. But there is one more problem: similar cladistic trees
    > >may be a consequence of the features concerned being interdependent.
    > Tim:
    > Co-evolution.
    > >The function of limbs is certainly a consequence of the functions of many
    > >component parts. Now, if two components are functionally interdependent,
    > >their evolution is interdependent, and we would expect the two cladistic
    > >trees formed for the two features in a group of taxa to be similar, no
    > >matter how this came about.
    > Possibly, but...
    > What is the likelihood of two unrelated species using two unrelated
    > components to fulfill a similar function in a manner that renders
    > the components difficult to distinguish? I'd bet it can happen, but
    > relatively often, compared to the number of cases where the phylogeny
    > is clear-cut? Even in protein co-evolution there are often other parts
    > of the peptide chains which reveal information about ancestry.

    Yes, the main key to reliable information about the true phylogeny hides
    in these "other parts", as far as they are selectively neutral. How do
    we recognize a selectively neutral part of a molecule? By assessing its
    variability among the homologs. How do species-specific requirements
    show up? Also by parts variable among the homologs in different species.
    So how do we decide whether a variable part reflects neutrality or
    species-specific requirements? Possibly by looking at polymorphisms
    within a species. And how do we decide all alleles are exactly
    equivalent under all conditions ever encountered? How can we detect the
    phylogenetically truly neutral and therefore informative nucleotide
    pairs in the genome? It probably rarely happens that we find a component
    difficult to distinguish between species, because we always have the
    species-specific part. But it's just this part that is difficult to sort
    out and interpret.

    So we need another key, which appears to be the large-scale comparison
    among many different features of many different species. And we hope we
    have sufficient data to get a significant answer. I grant that there are
    regions where the phylogeny is pretty clear-cut, but among all parts of
    the "tree of life", confidence levels and probabilities of correctness
    vary greatly. Homology is often not so straightforward to recognize. As
    I wrote to George Murphy, the analysis of the huge database of the human
    genome demonstrates how difficult it is to be sure of a homology on the
    molecular level (W.H. Li, Z. Gu, H. Wang, A. Nekrutenko, Nature 409 (15
    Feb 2001), 847), not to speak of the much more intricate problem of
    homology on the macroscopic level.

    > >Just as a similarity between the same feature in two taxa may be a
    > >consequence of either common descent or common requirements, so a
    > >similarity between the cladistic trees constructed for two component
    > >features in the same group of taxa may be a consequence of either a
    > >descent acording to the cladistic tree found or of the two components
    > >being interdependent. This case has also been described in PSCF 44
    > >(2/1992), 80-94.
    > I'd bet this particular concern was discussed by taxonomists before 1992.
    > >Of course, if the compound function under consideration
    > >has been formed in different ways in different taxa, there will be at
    > >least some different interdependencies between the component functions
    > >or features, and there might be a stronger argument for evolution.
    > >
    > >>>Genuine evidence for homology may require functionless features (cf. my
    > >>>"How has life and its diversity been produced", PSCF 44 (2/1992), 80-94).
    > >> [...]
    > >>
    > >>That's definitely one strong class of evidence for homology. And I think
    > >>it's certainly a legitimate concern in molecular biology for studies about
    > >>the origin of single enzyme domains or protein families. But for multi-
    > >>enzyme or higher level, multi-component systems that are shared across
    > >>different species, it can be harder to make the case for convergence due
    > >>to functional necessity, particularly if the ultimate functions are not
    > >>quite the same across species (i.e. Consider the many end-functions of
    > >>vertebrate limbs in various species: terrestrial locomotion (quad- and
    > >>bipedal movement), aquatic propulsion, grasping, ripping, mating, cleaning,
    > >>burrowing, communication, & etc.)
    > >
    > >My argument given above includes the species-specific modifications of a
    > >given structure or function. If a given requirement is similar in two
    > >species, we expect the corresponding structures in the two species to be
    > >similar.
    > Grossly, perhaps. But not at the lower, structural levels. Both dolphins
    > and many fish have dorsal fins which they use for identical purposes.
    > However, these fins are structurally very different. All things being
    > equal (or undecidable, perhaps), I can see no a priori reason to suppose
    > that similar functional requirements will necessitate the same solutions
    > or the same routes. Only if the requirements severely limited the possible
    > solutions, or if historical factors (e.g. common ancestry) or horizontal
    > transfer opened the path to similar solutions would I expect otherwise.

    (My last sentence above is only the introduction to what follows below.)
    Of course, comparing fish and dolphins, the final dorsal fin shape is
    not all we have to consider. The different embryonic development will
    also have consequences for how the fin is going to be built at the
    structural level.

    > >A natural corollary of this is that the differences between
    > >these requirements in the two species will lead us to expect
    > >corresponding differences between the two structures, no matter how
    > >this state of affairs came into being.
    > If dolphins and humans were separately created, why would they appear
    > to share many skeletal features, particularly in their fins/arms which
    > have very different functions? I think historical contingency is being
    > seriously overlooked as a major source of biological similarity. Sure, the
    > gross, external morphology of human and dolphin limbs are very different,
    > but the component structural elements reveal their common ancestry.

    I am not talking about dolphins and humans being separately created
    (apart from the "image of God"). I agree with you that historical
    contingency has to be considered as a possible source of biological
    similarity, but even in the case of the fin/arm comparison, I am not so
    sure this is the decisive aspect. The radius-ulna system, e.g., is
    presumably important for twisting motions in both dolphin and human.

    > >We would expect limb buds to be very similar between frogs, ducks, dogs,
    > >and dolphins, but the further they advance in their embryonic development,
    > >the more they will differ from each other in order to function properly.
    > Possibly, but I'm not convinced that there wouldn't be exceptions.
    > The timing and location of developmental events may vary.

    Yes, of course.

    > >The same would apply to any other multi-component system.
    > I would think that an increase in the available degrees of freedom in
    > a system would correlate with an increase in the number of possible,
    > non-identical solutions. Thus I'd be surprised if many multi-
    > component systems with separate evolutionary histories would converge
    > on forms that would be mistaken for having common ancestry.

    The correlation is certainly correct in a general way. How this
    translates into individual design decisions is another question (being
    nasty, I use ID language, but without insisting on a non-evolutionary
    solution). How far are morphological structures dependent on the
    requirements of embryonic development? Both the dolphin and the human
    have to grow out of a single cell, and their environments may be rather
    similar in many respects until near birth.


    Dr Peter Ruest                  Biochemistry
    Wagerten                        Creation and evolution
    CH-3148 Lanzenhaeusern          Tel.:   ++41 31 731 1055
    Switzerland                     E-mail: <pruest@dplanet.ch
     - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            In biology - there's no free lunch -
                    and no information without an adequate source.
            In Christ - there is free and limitless grace -
                    for those of a contrite heart.
    Dr Peter Ruest			Biochemistry
    Wagerten			Creation and evolution
    CH-3148 Lanzenhaeusern		Tel.:	++41 31 731 1055
    Switzerland			E-mail:	<pruest@dplanet.ch
     - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    	In biology - there's no free lunch -
    		and no information without an adequate source.
    	In Christ - there is free and limitless grace -
    		for those of a contrite heart.

    This archive was generated by hypermail 2b29 : Tue Jun 05 2001 - 13:12:19 EDT