Re: Hamilton's Rule In The Mirror

Guy Hoelzer <hoelzer@xxxxxxx> wrote in
news:d4ojon$2bik$1@xxxxxxxxxxxxxxxxxxx:

> in article d4oc9n$2909$1@xxxxxxxxxxxxxxxxxxx, Tim Tyler at
> tim@xxxxxxxxxxx wrote on 4/27/05 8:49 AM:
>
>> Reproductive strategies that aim at producing many immediate
>> offspring are not necessarily effective at producing many long-term
>> descendants - since offspring *quality* as well as offspring
>> *quantity* matters to the evolutionary process.
>
> I don't intend to highjack this thread in a new direction but, just
> out of curiosity, do you (or anyone who cares to answer) have any
> thoughts about how natural selection might optimize the time scale of
> its effects (i.e., what is the modal time scale(s) at which fitness
> effects are integrated across generations?)? I know that JE's answer
> is one generation, and I know his argument. I am looking for
> alternative arguments. I imagine that some such arguments could be
> based on the constraints of hereditary mechanisms, and others might be
> more general (less idiosyncratically constrained).
>
> Guy Hoelzer
>
>
>

I have argued in a previous thread for the use of a "discount rate"
(borrowed from economics) in determining how much a future fitness gain
might offset a current fitness loss. Discount rate is simply the time
value of money - as a rough measure it is the interest rate you could
earn by a relatively conservative investment less the long term inflation
rate. (JM, who IIRC has a degree in economics, can perhaps correct me on
this). One of the things about a discount rate is that it sets a time
horizon beyond which future benefits have very little present worth. For
instance, for a 3% annual rate, any benefits beyond 60 years have only a
5% effect on present worth, while for an 8% rate the time horizon for a
5% effect reduces to 30 years.


Obviously this does not imply that an organism is anticipating the
future, only that assuming the current fitness loss is not great, a trait
that creates a future fitness gain can still be selected for. It is also
clear that for many, perhaps most, traits we don't need to bother with
the calculation, since they have both an immediate and a longer term
fitness gain. For example, if a mutation makes me able to run faster with
no other side effects, the discount rate won't matter.

For human economic questions, we have a lot of proposed discount rates.
If the concept is useful in evolution, an interesting question is what
rate to use, and whether to apply it on a simple time scale or on a
generational basis. Since seventeen year locusts have evolved, and since
most insects have a maximum one year generation time, we can guess that
the discount rate can easily be as low as 15% per year or generation
(remembering that a higher discount rate gives a shorter future time
horizon for affecting current benefits).

Yours,

Bill Morse


.