Re: Bipedalism in different substrates

From: Jason Eshleman (jae_at_vici.ucdavis.edu)
Date: 07/04/04 

Date: Sun, 4 Jul 2004 21:45:54 +0000 (UTC)

Bob Keeter <rkeeter@earthlink.net> wrote:

>"Jason Eshleman" <jae@vidi.ucdavis.edu> wrote in message
>news:cc9dtl$2hd$1@woodrow.ucdavis.edu...
>> In article <r8afe0hsardbc48h3mopnk274bg09543g5@4ax.com>,
>>
>> >On Sat, 3 Jul 2004 17:49:00 +0000 (UTC), jae@vidi.ucdavis.edu (Jason
>> >Eshleman) wrote:
>> >
>Snippage. . . . . .
>
>> >You and Bob are saying much the same thing (although in more
>> >scientific language in your case). He asserts that there is no
>> >selective effect from migratory river crossings because the outcome is
>> >random. You say that there must be heritable differences.
>
>Glad to see that I at least get by with my "less scientific" lexicon! 8-)
>
>> Just for fun, what happens to the rate of evolution when selection is
>> removed?
>
>Without some form of "selection" I would say that evolution, as a
>constructive
>process, would cease. At least IMHO, evolution occurs when a species
>adapts to "better fit" the environment. If there is no selection based on
>some
>form of "the fittest", there would be no move in any defined direction, sort
>of
>a molecular genetics equivalent to Brownian motion.

This was a leading question to get someone to bite. I'll get the hook out
of your mouth soon enough. I was fishing for someone to fall for the
evolution = natural selection trap. Evolution, in a biological sense, is
the change in allele frequency in a population. Selection is *one* of the
mechanisms of this change, but a mechanism that serves to slow change
considerably, but in doing so, directs along particular successful
avenues. It does this by reducing variation.

(I hope Larry M. is aware that he really did teach me something many
years ago.)

Without selection, evolution speeds up *considerably.* Variation results
from mutation and recombination without which evolution would soon cease.
Selection is a brake, a mechanism to reduce variation, to weed out
mutations that don't work, to weed out recombinations of alleles that
don't work. Eventually, without variation increasing from the influx of
new genetic material, all loci would eventually reach fixation of an
allele and evolution would stop. In its absence the rate of mutation and
the rate of evolution are equal as every new mutation results in a change
in the relative frequencies of alleles. However, the *direction* of
evolution without selection moves about equally random.

The process of adaptation is significantly slower as it has to wait for
the variation to occur. Selection does not produce variation, it only
works on variation that already exists.

>On the other hand, that Brownian motion might not be without results.
>Regressive
>tendencies might be enhanced if not "weeded out" by selective processes. An
>easy example might be found in the domestication of the corn plant. The
>artificial
>selective processes of human farmers have turned corn into a grain that
>probably
>would be extinct in a matter of a very few years without human intervention.
>IOW
>the corn "got soft" without the survival inputs of a wild environment. Some
>might
>even say that the reduced cranial capacity of modern HS vs early HS might
>just
>indicate a reduction in the need for brain power to insure survival, but
>that is another
>subject I think. . . . .
>
>> I am saying that there may not be a selective effect. I am more certain
>> that the need to get across rivers from time to time isn't a sufficient
>> selector to make a terrestrial creature develop the suite of
>> characteristics I've seen various wet-apers say indicate a "more aquatic"
>> existance. I am saying that the null hypothesis is not that selection is
>> *always* going to play a factor in all events though.
>
>Ah, you are not talking about just a species left in a very benigh
>environment
>for a long time and loosing its edge, you are talking about something other
>than
>"natural selection" and the survival of the best swimmers or whatever! Hmmm

You do catch my "drift" then.

>> >I am saying that the individuals with the highest probability of
>> >crossing safely are those who enter the water early or slightly away
>> >from the main herd, not mid-pack; are good swimmers; are good at
>> >choosing a safe exit route. There may be any number of other factors,
>> >of course.
>>
>> And of this number of other factors, many of them some of them are not
>> going to be under genetic regulation and others may not be predictable in
>> a manner that selection can favor any optimization. If the number of
>> random occurances is high enough, selection won't be the dominant factor
>> in evolution in these instances.
>
>hmmm. . . . difficult to turn your statement into a probabalistic equation I
>think.

Not difficult to model actually. It's upper division population genetics.
It's too time consuming for me to run through it all here, but John
Gillespie's "Population Genetics: a concise guide" (available through
Amazon I believe and not too pricey does a good job of walking through the
steps if you're interested.

The difficulty comes in trying to predict the outcome. A known selective
force gives a predictable outcome. It's the recognition that drift, those
random events blind to the genotype of the organism, though powerful, do
not move in a predictable direction (or a constant direction). Drift,
like selection, is a variation destroying process. This force can be more
powerful that selection in many cases and if so, drift will in many
instances remove selectively favorable adaptations at random. There is a
probability that a favorable allele will be eliminated by drift before it
reaches fixation.

The actual strength of selection is independent of population size.
Positively selected alleles will move towards fixation with a probability
equal to the strength of the selective advantage. Ergo, a mutation that
makes an organism 10% more likely to survive and pass on offspring, has a
10% chance of reaching fixation in the population. This means that
there's a 90% chance it will disappear, though advantageous, by random
cause before it can become fixed.

The strength of drift is dependent on the size of population. In small
populations, it is even possible to fix deleterious mutations if the
selection against them is not strong.

>I do know that at least one of those papers I referred to focuses on the
>multi-environmental (pan-environmental) inputs to human evolution. IOW
>a situation where the wide ranging hominds became "optimized" not so much
>for any one of the paleo-african environments but rather optimized across
>the
>entire range of inputs.
>
>If I caught your drift, the situation would seem to be a case where the
>advantages
>and disadvantages of a given adaptation would need to be evaluated across
>the entire range of possible homind environments. Some might offer some
>degree
>of advantage across the range (for example, intelligence), some might be
>very
>scenario specific (for example, profuse sweating in a hot/dry atmosphere),
>but a
>"push" in a hot but more humid environment. (the "push" only because the
>water
>that would be wasted in the efforts to cool via evaporation could be very
>easily
>replenished in the wetter environment).

We have to look at a number of factors. Relative advantages in changing,
unpredictable environments is one (and one that I think is very important
to our successful generalist adaptations), but even in steadier
environments, selection isn't the only factor at work. The
hyper-pan-adaptationist view, the common misunderstanding of evolutionary
theory that too many people fall for, is insufficient. It's easy to fall
for it because the basic Darwinian selective principles are easy to grasp.
The relative strength of random processes are not so easy to get a hold
on, though they are every bit as important to evolution. That's
essentially my argument.

>> >All these are heritable: entering early is a risk-taking
>> >characteristic; moving away from the main herd is caution; good
>> >swimming is obvious; choosing a safe exit route is intelligence. Yes,
>> >many die through random chance, but over time, repeated several
>> >several times a year, there will be selection.
>>
>> These *may* be heritable characteristics. Note the word *may* as it is an
>> important one. We don't know the heritability of risk-taking in zebras.
>> You are making the assumption that the characteristics in question are
>> heritable. This is a big assumption, a departure from the null. Do be
>> aware of this.
>>
>> But if they are heritable, there is still much at random going on.
>
>Imagine that every fourth zebra in the water were grabbed by a croc with
>no limiting factors like bloated crocs or timing. That would give you a
>25% attrition rate, but what characteristic would increase the chances of
>not being in that 25% and therefore surviving to pass on genes? I get to
>take "aquaphobia"; you have to supply a different answer! 8-)

Right. Imagine that there is variation in the swimming ability of zebras.
But the odds of the best swimmers dying for other reasons are better than
the odds that their swimming ability saves them from death. This sort of
thing does happen. That's what drift is.

This is essentially why I don't think that small differences in the
swimming speed of a human mean squat when we start talking about aquatic
predators. The odds that swimming speed saves you from a creature who is
at worst three times faster than you, at best 2.9 times faster (note: I'm
making these numbers up, but the point is that the difference relative to
the speed of an aquatic predator is likely negligible). When we see
selection for fast escapes from predators in nature, we don't
generally see creatures who are less than half again as fast as the
predator.

Relevant Pages

  • Re: Bipedalism in different substrates
    ... evolution occurs when a species ... It does this by reducing variation. ... > Selection is a brake, a mechanism to reduce variation, to weed out ...
    (sci.anthropology.paleo)
  • Re: Addressing Scientific Reductionism
    ... theory of evolution the greatest thing since sliced whitebread. ... Its emphasis is on how variation in organisms is generated, ... rather than how natural selection works. ...
    (sci.bio.evolution)
  • Re: Mayrs Way
    ... occurs if there are no forces at all is called random genetic drift. ... Louis Dollo in 1890 that states that evolution is not ... important for selection because mostly temporarily isolated demes provide ... Dr Moran's argument represents a mathematicians rehash of the discredited ...
    (sci.bio.evolution)
  • Re: Bipedalism in different substrates
    ... It does this by reducing variation. ... >> from mutation and recombination without which evolution would soon cease. ... >> evolution without selection moves about equally random. ...
    (sci.anthropology.paleo)
  • Re: Digital Genetics and Evolution Theory
    ... aspects of inheritance and their implications for evolution theory. ... variation of characteristics between different individuals. ... creates differences for selection to select. ... It seems to be a somewhat circular situation: organisms are ...
    (sci.bio.evolution)