Topics, Tools and Techniques in Paleoclimate Research

Dendrochronology
Speaker: Jeff Dean, LTRR/ANTH
March 24th 2004

Summary provided by Jessica Rowland

Jeff Dean spoke about the relationship between humans and their environment, with a special focus on the Anasazi inhabitance and abandonment of Long House Valley, AZ. Jeff’s talk focused around five key points:

1) reconstructing the past environment in terms of both high frequency events (precipitation estimates from tree rings, PDSI) and low frequency events (alluvial chronostratigraphy, pollen and packrat midden data)

2) reconstructing past human behavior through archaeological excavations

3) reconciling spatial and temporal sensitivities of different reconstruction methods

4) coming up with testable hypotheses

5) implementing research through agent-based modeling.

Long House Valley is an ideal study area because it is physically bounded by the Black and Skeleton Mesas. Because the valley dissects a monocline, different soils and vegetation are found on opposite ends of the valley, leading to various potential farming zones. By using paleoenvironmental data within agent-based modeling, Dean and coworkers attempted to explain in a model the disappearance of the Anasazi from Long House Valley. The archaeological record shows that climate greatly affected movement of the Anasazi. The period from 950 – 1150 AD was a paleoenvironmental optimum, marked by widely distributed settlements along the alluvial plain. It was also a time of low social complexity. After this period, most southern sites were abandoned, and clustered sites in the north became more common. During the Great Drought of 1250 – 1300 AD, arroyo cutting drove the water table down and forced an even greater northward movement of people in the valley (to where the best soils were). This convergence of people brought about great social complexity, but by 1300 AD all Anasazi had abruptly left the valley.

The model had many parameters in addition to the paleoenvironmental data:

  1. base interval = 25 years (one human generation)

  2. agent unit = one household (5 people)

  3. need to have 2 years corn stored, or agents must move

  4. household must be within 1.6 km of farmlands

  5. farmlands can be used by only one family

  6. the soil type determines production (crop yield in kg corn/ha/yr)

  7. each household consumes 800kg corn/yr

The model used two separate clocks. One tracked the amount of food in household storage from April-April, and the other tracked the beginning/fission/extinction of a household in regards to marriage, childbirth and death. The archaeologically known number of households were placed randomly throughout the valley, and as time progressed, the agents in this model made decisions based on the landscape conditions.

The model conforms quite well with archaeological data from 900 – 1300 AD. It predicts farming plots in areas with the best soils, and shows interdecadal population variability. It also simulates household dispersal during favorable environmental times, and household congregation during more difficult periods. However, according to the model, the Anasazi could have stayed (although in reduced numbers) in Long House Valley after the Great Drought. The model suggests that about 2/3 to 3/4 of the Anasazi’s movement was based solely on paleoenvironmental/climatic factors. Thus, social, political and ideological factors must have ultimately contributed to the abandonment of the valley.


Back to TTT04 syllabus.