PLUTONIUM IN A GRASSLAND ECOSYSTEM 435 



tion vs. plutonium concentration in underlying soil was also statistically significant 

 (P < 0.01) but was less conclusive than the litter vs. soil curve and is not shown here. 



The CR's of the vegetation were higher than those produced in greenhouse studies. 

 Typically, uptake of plutonium under laboratory conditions has been on the order of 

 10"^ to lO"'* of the soil concentrations (Newbould, 1963; Wilson and Cline, 1966; 

 Romney, Mork, and Larson, 1970; Schulz, Tompkins, and Babcock, 1976). The Rocky 

 Flats CR of 3.4 X lO"'^ suggests either increased root uptake by grassland species or 

 another method of contamination, such as aerial deposition of resuspended soil particles. 

 The high surface-to-volume ratio of grasses and the hairy nature of the leaves of many 

 members of the sunflower family would be amenable to a high rate of impaction and 

 attachment of small soil particles. Given wind-redistributed plutonium at Rocky Flats, 

 surficial attachment of contaminated soil particles to plants is the likely mechanism of 

 contaminating the vegetation. 



Plutonium Isotopic Ratios 



Ratios of plutonium isotopes or ratios of ^^^Pu and "^' Am have been reported from 

 several sites (Emeiy et al., 1976; Gilbert et al., 1975; Hakonson and Jolinson, 1974; 

 Markham, 1976). In the hope that the examination of the isotopic ratios of ^^^Pu and 

 ^^^Pu in the grassland would give some insight into the relative ecological availability of 

 these two nuclides, isotopic ratios were calculated for samples analyzed by alpha 

 spectrometry [isotopic ratio (IR) = ^^^Pu pCi/g of sample -^ ^ ^ ^ Pu pCi/g of same 

 sample] . Ratios were not calculated for samples where either isotope was below the 

 detection limit. Ratios were tabulated according to sample type and tested for goodness 

 of fit to a normal distribution. The distribution of IRin the various soil depths was either 

 lognormal or marginally normal. Small-mammal tissues appeared to be lognormal with 

 respect to the IR. 



As suggested by Doctor and Gilbert (1977), the concentration of ^^^Pu was plotted 

 vs. ■^ ^ ^ Pu for each of the seventeen sample types. Seventy percent of these groups exhibit 

 a zero intercept, based on a t-test. These results implied that the ratios were constant 

 within the tested sample groups and that ^^^Pu/'^^^Pu would be an unbiased estimator. 

 However, because of the likelihood that the IR is lognormally distributed within most of 

 the sample groups, median IR's are reported here (Table 8). The R4 method of 

 calculation discussed by Doctor and Gilbert (1977) was used to calculate these values. 



At first glance the median isotopic ratio in soil appeared to decrease as depth 

 increased. However, the overlapping 95% confidence intervals for the listed medians 

 suggested that the ratio is relatively constant. As expected, neither linear, exponential, 

 nor power-function regressions of the raw IR data vs. soil depth were statistically 

 significant (P > 0.05). 



Despite the limited number of htter and vegetation samples analyzed for both ^^ Pu 

 and ^"^"Pu, the median IR's of these two compartments were very similar to IR's of the 

 soil. These results tended to indicate that the litter and vegetation were closely linked to 

 the soil. 



The IR's in the animal compartments raised some very interesting questions (Table 8). 

 Only two sample types (GI tract and muscle) exhibited 95% confidence intervals that 

 overlapped with soil IR confidence intervals. Therefore, it appeared that the small- 

 mammal and arthropod compartments had lower IR's than soil. A lower IR would imply 



