PLUTONIUM IN A GRASSLAND ECOSYSTEM 427 



TABLE 3 Mean Plutonium Concentrations of Soil Samples from Rocky Flats 



TABLE 4 Regression Parameters of Soil Plutonium Concentration 



(pCi/g) Adjusted for the Sample Location as a Function of Soil 



Particle Diameter at Various Depths* 



Least-squares regressions were calculated with linear, exponential, and power-function 

 models of plutonium concentration in surface soil as functions of the distance east or 

 south from the asphalt pad. The power-function model gave the best fits of the data for 

 both curves (Figs. 3 and 4). A t-test indicated that the slope of the distance-south curve 

 (Fig. 4) was steeper (P < 0.05) than that of the distance-east curve (Fig. 3). These results 

 conform to the concept of wind-distributed plutonium; the more effective, stronger 

 winds were to the east, and hence the slope of that curve was smaller. 



Several multiple linear-regression models were calculated. The model that accounted 

 for the largest fraction of the total variance (0.868) had the following form: 



In Pu = 1 1 .1 5 - 0.0535 In E - 1 .628 In S 



where Pu is the plutonium concentration (pCi/g), E is the distance east of the asphalt pad 

 centerline (m), and S is the distance south of the asphalt pad centerline (m). 



With this model plutonium concentrations oi samples in the soil depth profile were 

 adjusted to estimate the concentration expected at a common location. The adjusted 

 values were then regressed as a function of sample depth (Fig. 5). As with the distance 



