A Valley of Anasazi:
Years ago I read an article that I thought was interesting.  It gave data from Long House Valley, that had been occupied by the Anasazi Indians for hundreds of years.  The data graphed rainfall and population in the valley.

Jared M. Diamond, “Life with the Artificial Anasazi,” NATURE, vol 419 no 6907, October 10, 2002 p 567.

The tan line shows the population each year.  The way they arrived at the number was to find everywhere there had been a house and to comb through the charcoal from each house and do carbon dating.  This permitted them to know exactly which years each house was occupied.  I am not sure how they dealt with the question of whether some of the wood had come from old trees, but apparently they were sure of their ground.  They could count up the number of families every year and calculate how many people lived in the valley that year.  Again, I am not sure how they knew every household was the same size and remained the same size every year, but again they did not seem to have a problem with it, so I am happy to accept it.

One can only gaze in awe at this display of American resources.  Several years ago the government of Syria had a dam built on the Euphrates, flooding a wide area. This was the Fertile Crescent, and any number of rich archeological sites significant in the emergence of what has become the world civilization were destroyed.  The Syrians invited in archeologists and invited them to learn what they could.  They chose a single tell, a place where city had been built on the ruins of city for thousands of years.  They were able to do only a cursory examination, hardly more than a single cut through the tell, in the time they had available.  Had it been in the American southwest, it would have been exhaustively analyzed. 

So the American scholars wound up with this graph of the population in the valley.

The blue line is a measure of rainfall.  Jarred Diamond had data on tree rings, I think it was, that indicated temperature and rainfall, essentially how well things were growing, and used that to calculate the maximum carrying capacity of the valley.  The time course of the two graphs is similar.  The implication was taken to be that the environment limits human populations.

A closer look however shows a problem.  From about 900 to 1000 AD, the actual population is maybe twice the theoretical maximum.  One is forced to suspect that the “theoretical maximum” was arrived at by comparing rainfall with households and assuming that the population peak in about 1250 and the rainfall peak at about the same time matched.  Then it was seen that the rest of the two curves matched pretty well.

But to play devil’s advocate, let me suggest that the 900 to 1000 data indicate that the blue line needs to be brought up to about twice what we are shown.  That would mean that rainfall and temperature were never in fact the limiting factors and we will proceed on that assumption. 

The question is whether gene pool size may be the critical factor.  If so, the data are not clean.  There are at least four places where the number of household rises vertically.  Either there was an unaccountable stepwise increase in birth rates, or possibly a lot of young people decided to go out and start their own homes in a single year or, far more likely, there were at least four separate years of substantial immigration.  One suspects the immigrants each time represented one or a very few extended families, rather large kindreds, but all Anasazi, just like those already there. 

On the face of it, then, we are not looking at the experience of a single family undergoing natural growth but at the experience of immigrating families joining the locals. 

We would not anticipate any difficulty on excessive gene pool size before about 1000.  The total population looks to be under 300, which ought to be ideal.  But then the population rises exponentially and then crashes with a minimum at about 1150, when it rises again, only to crash finally in about 1300.  In other words it undergoes cycles of growth and collapse just like the computer model of a population limited to 20,000 we showed on June 24, 2008. 

But there is a problem. Our historical data suggest that a cycle should take about 300 years.  Here we are looking at a 150 year cycle, much more like the UN pattern we presented on August 25, 2008.

Why are the two cycles different?  This is something we will have to look at in the future.

In the end the population drops to zero.  A final emigration is postulated to account for that fact that the population vanishes completely at a time when the climate is better than it has ever been except between 1150 and 1300.  It is better than it has been in all but 150 years out of the 500 the valley was occupied.

The effect is so striking that Diamond is at pains to explain just why they all left.  He does not explain why many did not all leave at the same time. 

We are still left with the correlation between tree ring thickness and population size.  Maybe when people stop taking firewood the trees grow slower.  So maybe the correlation is driven by population size.  Unfortunately this whole experience is a single data point, telling us nothing for sure in isolation.

It looks to me like they all got old and died.  I am sure that people were coming and going all the time, but at the end it looks like natural attrition rather than mass migration.  The immigrants appear to have arrived in groups; the population size goes vertical.  They should have moved out together, but the population never falls vertically.   The last new home was built in about 1270 and the last people, who may have set up housekeeping when they were about 20, gave up 30 years later.  50 is not very old, but for getting along in the strenuous business of subsistence farming it is not so very young if not supported by vigorous adolescents. 

Altogether, this proves nothing.  But the data set is quite consistent with the idea that when the valley population exceeded a certain size this induced a lethal fertility decline such as is evident in so many other places, whether the effect is a single 300 year cycle or two 150 year cycles.  In the end the result is the same. 

It may be that for once we can see the actual numbers as a population dies out. 

There have been 1,619 visitors so far.

Home page.