Estimating Prehistoric Production

    The annual sustainable population was regressed, using stepwise multiple regression analysis with Systat 9, against the following variables for 1957 through 1988: reconstructed precipitation from previous August to current July (AUGJULPRECIP), reconstructed precipitation from previous October to current May (OCTMAYPRECIP), reconstructed current June PDSI (JUNEPDSI), June PDSI from previous year (PDSI_YR_1), previous year’s October to May precipitation (OCTMAY_YR_1), and previous year’s August to July precipitation (AUGJUL_YR_1) (Laboratory of Tree-Ring Research 1994).

Linear Regression

Dep Var: POP N: 31 Multiple R: 0.530 Squared multiple R: 0.281
Adjusted squared multiple R: 0.201 Standard error of estimate: 545.840

Linear Regression Equation

The resulting regression equation is:

POP=(OCTMAY_YR_1*734,680.940)+(OCT_MAY_PREC*142.335)+ (PDI_yr_1*-479,418.573)-3,883,625.951

where POP= sustainable population with two years storage, OCTMAY_YR_1= the e stimated precipitation from October to May of the previous year, OCT_MAY_PREC= the estimated precipitation from previous October to current May, and PDSI_YR_1= the June PDSI from the previous year.