Question re:Evolution and Menstruation

Ram Samudrala ram at mbisgi.umd.edu
Sun Jan 5 14:15:52 EST 1997


John R. Campbell (campbejr at phu989.mms.sbphrd.com) wrote:
>On 3 Jan 1997 05:12:18 GMT, Ram Samudrala <ram at mbisgi.umd.edu> wrote:
>>B. Moosang (bmoosang at uoguelph.ca) wrote:

>>>OK, just trying to settle an argument between two non-science people - 
>>>how long would something like menstruation take to evolve?  
>>>[...]

>	Yes, but for most other mammalia (IIRC) this is a *yearly*
>	process, reducing the reproductive rate (though, come to
>	think of it, rabbits may destroy this general line of thought).

>	The argument, as I understand it, was how would a monthly
>	cycle have evolved in humans?  While a monthly fertility cycle
>	would have advantages in expanding the birth rate, how would
>	this have provided enough of a selection pressure to eliminate
>	the yearly fertility cycle...

That's not what I gather from the question above, but suppose a
chimpanzee had an yearly rate which the first humans also had, and
there were some mutations that led to one human having the monthly
rate, it doesn't take too much selective pressure and too many
generations for it to spread throughout the population (exact numbers
can be worked out given all the mathematical models that exist of
modelling selection).  

>>Humans have been around a bit longer than that.  Assuming a mechanism
>>for transporting the egg to the uterus, for the lining of the uterus,
>>and for getting rid of debris, didn't exist, it boggles my mind to
>>think how long such a function would take to evolve in the first place
>>(going from 0 frequency to >0 frequency), let alone be naturally
>>selected for.

>	He's thinking that the monthly cycle was selected for in the last
>	15K years. 

Again, he uses the word "evolve", which could mean evolution of the
function itself or selection within the population.  I answered
both---the former is unlikely within 15,000 years, but the latter
isn't impossible I'd intiutively guess, but again, he'd have to plug
the respective numbers into the equations that model selection.

The thing is that the function has to evolve (ARISE) in the
first place BEFORE it can be selected for.  

--Ram

me at ram.org  ||  http://www.ram.org  ||  http://www.twisted-helices.com/th
                                The dollar will never fall as low as what
                                           some people will do to get it.
                                                       ---Alfred E Neuman



More information about the Mol-evol mailing list