Answering Evolutionist Attempts to Dismiss "Haldane's Dilemma"
[Author's note: see update at end of article]
In 1993 Walter ReMines book "The Biotic Message"1 hit the street, bringing with it several devastating
arguments against evolution that are still clamoring through the halls
and smoke rooms of the evolutionary faithful. One of these arguments is
based on a paper by J. B. S Haldane in 19572
that showed the reproductive capacity of vertebrates was way too low to
pay the costs needed to account for large-scale evolution. This problem
is referred to as Haldanes
Dilemma (go here
for an online discussion of the problem by Walter ReMine).
Refuting Robert Williams
So far I have only encountered one attack against Haldanes Dilemma that offers any kind of sophistication, one posted on the internet by Robert Williams. It regularly shows up early in search engines when searching on Haldanes Dilemma, so evolutionists often cite it or copy from it.
There are many, many problems with Robert Williams
article. When I first read it, I became very suspicious that he had never
read ReMine's book since ReMine deals with most of Williams arguments
in his book. I contacted Mr. ReMine, and he confirmed that Williams eventually
admitted on the newsgroup sci.bio.evolution to not having read
the book. On several occasions I attempted to contact Williams about this,
but he did not reply. It is very unfortunate that Williams refuses to
do the right thing and properly review ReMines book before posting
ReMine neglects the
fact that humans did not evolve from chimpanzees, rather humans and chimps
evolved from a common ancestor. Therefor we have actually had two different
branches each evolving independently, thus allowing for twice as many
gene substitutions (3300 vs. 1700) as ReMine has allowed, even if all
of the above is true.
Remine makes no such assumption in his book.
This can't possibly
be the case because many of the differences are known to occur at the
3rd triplet of gene codons and thus usually do not change the amino acid
coded and can't affect fitness. Furthermore, since 95% of the genome is
not transcribed (although that does not mean it is all non-functional
), most point mutations will not affect fitness. This reduces the number
of selected substitutions by 5 x 2/3 % or from 4.8 x 107 substitutions
to 1.6 x 106. Please remember that changes in the genome due to drift
and other "random" processes do not add to the cost of substitution.
I should add that Haldane's Dilemma has been viewed by scientists
as possible evidence for the importance of Neutral Evolution as proposed
by Kimura in 1967.
In fact, neutral mutations incur a greater cost, since they will have a greater propensity to drift back
and forth in frequency since they have no selective value. Every time
the frequency goes down, it negates any previous payment made by reproductive
excess to get it to that frequency; when it drifts back up, a new payment
via excess reproduction is needed, hence net cost is increased. According
to ReMine, Haldane showed that cost is minimized only when fixation moves
ReMine ignores the possibility
of gene hitchhiking - the concept that even though some mutations are
neutral, they will be carried to fixation because they are physically
close to a gene that is beneficial. ReMine does not ignore this possibility, he discusses it in the book
Williams pretended to read4. ReMine
also cites Haldane as addressing this possibility and that Haldane also
dismissed it as very negligible5.
ReMine does not ignore this possibility, he discusses it in the book Williams pretended to read4. ReMine also cites Haldane as addressing this possibility and that Haldane also dismissed it as very negligible5.
For linkage to pay the cost of two for the price of one, the following must occur:
a) The neutral mutation must occur about the same time as the beneficial mutation it is linked to. If it occurs say 50% into the fixation cycle of the beneficial mutation, it cant just magically appear on all the other chromosomes in the population. It has to begin its own payment cycle when it first appears. All those without the mutation, which would be the entire population plus all descendants without the mutation, must eventually be removed.
b) the two would have to remain very tightly coupled through at least half the fixation process to give the neutral mutation an even chance to reach fixation6.
c) gene hitchhiking is very rarely found in sexually reproducing populations7.
I hope it is now quite apparent
why linkage effects have negligible impact on cost evaluations.
hope it is now quite apparent
why linkage effects have negligible impact on cost evaluations.
This is a completely bogus argument for several reasons. First, the must common mutations are point mutations (base pair substitutions)8. Second, even when multiple mutations occur, the harmful ones will incur an immediate reproductive cost, and any remaining neutral or beneficial ones must still pay their own cost if they are to reach fixation!!! Also, it appears Williams again forgot that humans reproduce sexually, not asexually. If multiple mutations occur, they will be divided among the offspring, and only so many of these will reproduce on to the next generation. Hence only a handful will remain, only to face the same shredding machine the next generation. Because sex continually scrambles genes every generation, population geneticist Ronald Fisher (1930) estimated that a beneficial mutation will have at best only a 1 in 50 chance of ever reaching fixation in a population 9.
Haldane assumed that the cost of substitution had to be paid on top of the "natural" death rate! In other words, it didn't matter that 90% of a mammal's offspring died without reproducing - any death that resulted from the substitution of one gene for another had to be additional death that the animal would not "normally" have suffered. This is known as hard selection and we can now easily see why Haldane only allowed an excess fertility of 10% to go towards the cost of substitution. However, most Biologists today consider all or some selection to occur as soft selection. In this scenario, the cost of substitution is "paid" in the natural death rate of the animal. That is, a disproportionate number of the individuals that die without reproducing in any generation are the ones that have lower fitness due to their genes. The Biologist Bruce Wallace has been the champion of soft selection, and you can learn more about this topic in his book "Fifty Years of Genetic Load - An Odyssey".
Let me bring in another Williams to refute Williams! Highly regarded evolutionist George C. Williams wrote the following regarding Wallace and soft selection:
As we can see, Robert Williams last
effort to soften the blow of Haldanes Dilemma is disputed by an
evolutionist of considerably more standing.
Despite various attempts by evolutionists over the last 40 years to soften the impact of Haldanes Dilemma, it still remains an enormous problem for their theory. It is worth noting that Haldane's analysis even used very favorable assumptions for the evolutionary theory, such as assuming the mutations are dominant (recessive mutations pay an exponentially higher cost). Regardless, the numbers do not bode well for the evolutionists, and is very likely why the problem stays buried in back-room discussions and does not see the light of day in evolutionary textbooks.
Current molecular data is making matters even worse for the evolutionist faithful, because it makes the problem easier to see for the layman. I document this in my article Monkey-Man Hypothesis Thwarted by Mutation Rates. This article stands on its own and does not rely on the validity of Haldanes calculations. Using a conservative estimate of mutation rates based on current studies, it shows that the ape/human line would have required at least 40 offspring per mating pair just to maintain equilibrium! This forcefully argues that the Monkey-Man shared ancestor hypothesis is simply implausible.
Update: Several months after I wrote this, Robert Williams to his credit removed most of the arguments I addressed above from his web page! (he keeps a copy of the original here). His first line of defense in his latest installment is his claim that 1667 beneficial substitutions may be enough to account for human evolution from our alleged simian ancestor! As far as I'm concerned this is a complete capitulation of the issue! Remember that this is not just a problem for human evolution, but for mammalian evolution in general.
Robert also still defends gene hitchhiking as a cost reducer, and gives an example of it occurring in nature. I have not had a chance to confirm his example, but it doesn't really matter. It is still a rare phenomenon, as Futuyma points out in his Evolutionary Biology testbook7. A blind squirrel, well, you know the story.
Finally, Robert mentions that Haldane did address the issue of "multiple simultaneous substitutions". Haldane did indeed, but Robert's citation from Haldane's paper is completely inaccurate. In the paragraph Robert referred to, Haldane is not addressing the impact on cost of "multiple simultaneous substitutions". Where Haldane does address this is the 4th paragraph on page 522, where he explains "[for three mutants]...since the cost of selection is proportional to the negative logarithm of the initial frequency, the mean cost...would be the same as that of selection for the three mutants in series..."
1. Walter ReMine, The Biotic Message, 1993, St Paul Science