Hi George, David, Todd, Allan, Paul, Tim
In my discussion with George Murphy, I (again) brought up the topic of
macro- vs. microevolution (12 Jun 2001 21:38:34 +0200). There were
several comments, some questioning my doubts about microevolution being
sufficient to account for macroevolution. Let me try and answer the
following posts together:
George Murphy <gmurphy> (14 Jun 2001 07:28:07 -0400)
David Campbell <bivalve> (14 Jun 2001 14:39:35, 15 Jun 2001 15:17:56
-0400)
Todd Greene <tgreene> (14 Jun 2001 23:39:22 -0400)
Allan Harvey <SteamDoc> (14 Jun 2001 23:42:07 EDT)
Paul Nelson <pnelson2> (15 Jun 2001 09:32:09 -0500)
Tim Ikeda <tikeda> (15 Jun 2001 15:14:25 -0400)
In case some of the points raised will not be adequately dealt with,
please let me know, and I shall try to say something to these in
particular.
My definition of macroevolution, together with an argument for the
inability of microevolution to account for it, may be found in my
article, "How has life and its diversity been produced?" PSCF 44
(2/1992), 80-94. For your convenience I quote:
"... For a realistic evaluation of the adequacy of proposed mechanisms,
a clear distinction has to be made between _microevolution_ and
_macroevolution_. I define a macroevolutionary step or development as
any transition producing a fundamentally novel structure and function,
based upon a sequence of deoxyribonucleic acid (DNA) which is not
derivable from a previous one by means of a series of individually
selected mutational steps, but only through a random-walk process
involving a series of _nonselected_ intermediates. This definition may
not be conventional, but it points out a crucial distinction. The
assumption that any macroevolutionary event consists of a series of
microevolutionary ones is usually treated as axiomatic. If it holds,
any distinction between the two modes of evolution is basically
irrelevant. An argument that it does _not_ hold [from P. Rüst, "The
unbelievable belief that almost any DNA sequence will specify life",
Conference on "Sources of Information Content in DNA", Tacoma, WA
(1988)] will be summarized below... [p.82]
"Apart from point mutations, there are other mechanisms producing
variants, but they usually do not create any new _functional
information_. A definition of functional (constructive, or semantic)
biological information will be given below. Deletions and most
insertions destroy such information, sequence shufflings by genetic
recombination, transposition, duplication and other mechanisms move
preexisting information. These other genome modifications may, of
course, have profound functional consequences, often on a regulatory
level, but possible constructive effects they might have on their target
genes or larger contexts are likely to occur very much less frequently
than constructive effects of point mutations.
"One has to distinguish between new features produced by shuffling or
recombining preexisting functionalities, on the one hand, and new
functional features which _never existed_ before, but arose in sequences
having no function, or a different one, on the other hand. Although it
might in some cases be difficult to distinguish between these two kinds
of novelty, it is clear that many fundamentally new features must have
been produced in the biosphere as a whole. Unfortunately, the term
"evolutionary novelty" is sometimes indiscriminately applied to both of
these possibilities. The first kind is certainly relevant for the
origin of biological information. A recent investigation led to the
(still disputed) estimate of 1000 to 7000 basically different protein
exon or domain subunit families... [p.83]
"Natural selection of a new function presupposes a _minimal
functionality_: where nothing is selectable, nothing can be selected.
This minimal functionality, therefore, must be produced by random
processes. The probability of its spontaneous emergence depends on its
semantic information content, or the size of the minimal specification
required to define it, but not on the possible pathways leading to it.
It is, however, difficult to estimate the size of such minimal
specifications.
"One approach might be to consider the invariant configuration of a set
of known sequences performing a given function in different organisms.
Certain sequence positions are observed to be occupied by the same amino
acids in all known versions of a protein of a given specificity. It is
then assumed that functionality requires these specific occupations. An
anologous argument applies to positions permitting a certain restricted
variability. For good measure, all amino acids chemically similar to
the ones actually observed at a given position might be added to the set
of permissible ones [H.P. Yockey]. The totality of these restricted
occupations found for a given protein type constitutes its invariant
configuration. This is a lower-bound estimate for minimal
functionality, since positional interdependencies and species-specific
requirements are ignored. It may be compared with an upper-bound
estimate of the longest feasible non-selected path.
"The result is that reaching a given invariant by a mutational random
walk within 300 million years is already too improbable for three
specific amino acids. This estimate, presupposing 3.05 codons per amino
acid, 2.16 mutations per specific amino acid change (geometric average),
and a mutation rate of 10^-8 per nucleotide replicated, is based on very
optimistic assumptions: 10^16 moles C per year metabolized in the
earth's biosphere (today's total biomass production) consisting entirely
of bacteria (5x10^6 nucleotide pairs and 10^-14 moles C per bacterium),
and all of this DNA continuously participating in this particular random
walk. Yet _known invariants_ comprise not 3, but about _30 amino acids_
for basic enzyme functions, such as cytochrome c or ribonuclease, or at
least 5 amino acids for additional adaptations differing between groups
of organisms. These requirements are even below the real lower bounds
for functionality, as they reflect unique occupations only. At present,
it is unknown whether any smaller invariants might provide some minimal
functions. The restrictions on functional structures, such as enzymes,
are such that all mutations we observe today are detrimental or at best
neutral. To suppose otherwise for earlier organisms is speculative..."
[pp.84-85] (End of quote. For 3 specific amino acids, 40 billion years
would be required, on the average!)
All this, of course, deals with one of the lowest levels of biological
structures, that of specific functional protein domains. The reason I
chose this level is that it is understood in sufficient molecular detail
to permit the estimation of rough probabilities for the occurrence of
particular evolutionary paths (as long as there is no selection). All
hierarchically higher structural levels, like complex proteins, protein
complexes, organelles, cells, tissues, organs, physiology, limbs,
organisms, are as yet beyond such possibilities. I realize that the
different uses of the term "macroevolution" found in the literature
usually deal with the higher levels. I have not found any suitable term
specifically denoting the distinction I wanted to make, so I adapted the
term "macroevolution" for this (is this a coaptation? ;-)). Does anyone
have a better suggestion?
As the two main factors contributing to evolution are the generation of
variation followed by natural selection, the initial event in each
evolutionary step is a change in the genome, i.e. an event on the level
of molecular biology (apart from the random physical or biochemical
event causing it). All other changes, on higher levels, resulting from
this initial event, as well as selection and fixation of a change, are
much more complex and more difficult to analyze, as far as mechanism
(rather than mere phenomenology) is concerned. An attempt at evaluating
the probability and plausibility of an evolutionary step happening must
therefore begin on the level of molecular biology. Evaluations of
similarities on the morphological or molecular level start at the
opposite end, that of observed phenotype differences between different
species. Comparing the relevant plausibilities of different hypothetical
paths (or evolutionary models) is always based on a series of
assumptions, including those pertaining to the unknown selection
coefficients for the intermediates. The basic mechanism cannot be traced
in this way, and I have never found any data allowing probability
estimates for specific mutational paths on these higher levels.
I think the assumptions made in my model probability calculation are
sufficiently over-optimistic for evolution to allow the conclusion that
we are in trouble if we hope for the accidental emergence of all minimal
functionalities required for life. What would be needed now are
experiments demonstrating ways in which the initial random-path
evolution of a molecular function can indeed occur. This problem has to
be solved for peptide evolution, even if there was an intermediate RNA
world displaying more easily reached minimal functionalities, because
the take-over of a ribozyme function by a protein can only occur once
the new protein sequence has the required minimal functionality, which
must be generated by a random-walk mutational process. And this problem
is the same for many of the perhaps thousands of fundamentally different
protein domain "folds" or super-families (for the origin of life, the
combination of all minimally required functions has to be taken into
consideration). I think it is clear that my definition of macroevolution
doesn't produce a specification for classifying any given evolutionary
transition on the morphological level, as there probably is no case in
which we know what exactly happened on the DNA sequence level.
Now who should have the burden of adducing more evidence relevant for
the issue of whether the microevolutionary mechanism is or is not
adequate to explain macroevolution - the "unificationists" or the
"dichotomists"?
To again set the record straight: I am _not_ postulating that different
species (or higher taxa) must have been created ex nihilo (I believe in
common descent, including humans) or that divine interventions should
replace macroevolutionary steps. For _theological_ reasons I _do_
believe God used evolution and other natural processes throughout, with
the probable exceptions of the origins of life, of the psychological
domain (not the bodies of these animals, Genesis 1:21) and of the
spiritual domain (not the psychosomatic dimension of these humans,
Genesis 1:27), cf. A. Held & P. Rüst, "Genesis Reconsidered", PSCF 51
(4/1999), 231-243. But an adequate mechanism for macroevolution on the
molecular level just has not yet been defined, let alone made plausible
experimentally. So, I think it would be unfair to charge me with
defining microevolution as "evolution I believe in" and macroevolution
as "evolution I do not believe in". For the time being, I propose "God's
hidden options" as a worldview (not scientific) solution to the riddle
of macroevolution: P. Rüst, "Creative Providence in Biology", PSCF,
accepted for publication.
Peter Ruest
-- -------------------------------------------------------------- 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 : Mon Jun 18 2001 - 11:11:11 EDT