A Difference between the Historical Data and the Fertility Curves:
Leaving aside the computer model, the data I have been able to present falls into two categories. 

In one category there is a comparison of mating pool size, or some parameter of that, suggesting ultimately matching DNA graphed against fertility.  The curve starts with low birth rate where there is inbreeding, rising birth rate as mating pool size increases, followed by a fall again, leveling off below replacement.  This we see from multiple sources and it matches the computer model.  Bear in mind two facts: the curve becomes concave upward as it descends and where it is possible to estimate a time course it seems to work out over about 150 years.

The other category of data is the historical evidence in which societies tend to collapse as they reach an age of 300 years.  And as that date approaches, the descent of the survival chances of the society follows a curve that is convex upward.

Superficially, there seems to be a conflict, even though the computer model confirms that the time course of the fertility graph is as one should expect and that the predictable behavior of the risk of social collapse follows a stereotyped time pattern. 

The different time courses can be reconciled if one assumes that fertility cycles and that in human societies, collapse occurs after two and only two cycles.  The Long House Valley experience does suggest two cycles.  The first dip may have challenged the community, but did not destroy it.  Looking at the historical data, it appears that the first crisis never has a statistically significant chance of causing a collapse.

The convex curve versus the concave curve presents somewhat more of a challenge.  Of courses as described previously, it may be simply an artifact of the way in which different regimes gain control.  But looking back at the Long House Valley experience, it does appear that the fertility decline is indeed convex upward, just like the historical data, on the second descent.  This strongly suggests that the genetic structure of the population has changed. 

The only thing I can think of is that what is going on is recombination.  As I pointed out earlier, recombination can cause detuning of genes that accumulates over generations.  So on its maiden cycle, the newly expansive population grows, falls but recovers.  On the second cycle enough recombination has accumulated so that recovery is not possible.

What remains to be done is to go back to the computer program and revise it so as to model recombination along with the other factors.  It should then be possible to have a descending fertility curve that is convex upward.

But that is only a guess.  The proof will be in the program.  This of course means adding another dimension to an already complex model.  It also means getting access to more computing power.  In the years since I first made the model, more power has become available.  So if I can, I shall attempt to re-tweak the program and present the new results in a few months.  (I find this work is time consuming, but the real problem is making up ones mind to get started.  (Actually by today, 7/2/9 I am working on it, and I’ve changed my mind.  The hardest part of the work is the work.))

There have been 1,743 visitors so far.

Home page.