PLUTONIUM IN A GRASSLAND ECOSYSTEM 433 



tissue-sample means ranged from 0.288 pCi/g for bone to 13.6 pCi/g for kidney, and the 

 mean of whole arthropods was 5.48 pCi/g. 



The patterns, or rather, lack of patterns, in the small-mammal data were puzzling. The 

 tissues were arbitrarily classed either external or internal, depending on whether or not 

 the tissue had a direct contact with the animal's environment. External tissues included 

 GI tract, hide, and lung; internal tissues included bone, kidney, Uver, and muscle. By 

 virtue of the supposed low biological availability of plutonium and the proximity of the 

 external tissues to the contaminated soil, external tissues were expected to have larger 

 plutonium concentrations than internal tissues. Inexplicably, this was not the case. The 

 three highest plutonium concentrations were found in internal tissues, i.e., kidney, 

 muscle, and liver; hide and lung comprised two of the three lowest means. Additionally, 

 the amount of variation in samples within a given tissue was quite high. The minimum 

 tissue variation was in hide samples (CV = 1.84), and the maximum was in muscle 

 (CV = 5.85). 



Only two explanations for the high degree of variability are at hand. First, the 

 possibility of cross contamination always exists no matter how carefully one removes 

 tissues during dissection. Second, the extremely small sample mass of a few samples (a 

 dry kidney may be as small as 0.05 g) may have had a tendency to magnify the relative 

 plutonium concentrations. However, a plot of plutonium concentration in small mammals 

 vs. sample mass indicated that about as many samples had large mass and small plutonium 

 concentrations as had small mass and large concentrations. Beyond this, however, the 

 tendency for small-mass samples to skew the distribution has not been investigated. 



The nonparametric Kruskal-Wallis technique (Siegel, 1956) was used to test whether 

 or not the seven tissue means were from the same population. The resulting chi-square 

 value of about 44 indicated that the difference between the tissue groups was highly 

 significant (P < 0.001). Although no test was performed, it was intuitively obvious that 

 the mean plutonium concentration of the bone samples (0.29 pCi/g, n = 28) was lower 

 than that of other tissues. 



Plutonium Concentration Ratios 



The concentration ratio (CR) is a potential indicator of plutonium redistribution by 

 wind, water, or plant uptake. Concentration ratio is defined as the concentration in 

 activity per unit mass or volume divided by the concentration of the same nuclide in the 

 same units in another material. In this section the CR will have 0- to 3-cm-deep soil as the 

 material in the denominator [e.g., CR of Utter = (pCi Pu/g litter)^ (pCi Pu/g 0- to 

 3-cm-deep soil)] . 



The CR's of litter, vegetation, arthropods, and small mammals are listed in Table 7. 

 Litter had the largest CR followed in descending order by vegetation, small mammals, and 

 arthropods. Regressions of litter and vegetation CR's vs. distances east and south of the 

 asphalt pad did not achieve significant correlation coefficients (P > 0.05). 



The plutonium concentrations in litter and in vegetation were plotted vs. soil 

 plutonium concentrations from the same locations. Only the litter curve is shown here 

 (Fig. 7). The litter regression was interesting because of its high correlation (r = 0.975) 

 and near-unit slope (1.001). Although the number of samples here was limited, the data 

 comprising Fig. 7 suggested that litter may be an excellent estimator of soil plutonium 

 concentration in the grassland. The regression of plutonium concentration in vegeta- 



