206 TRANSURANIC ELEMENTS IN THE ENVIRONMENT 



Ho: 



K^i ,2 



L^i,p-U L 



using the means of the observed IR's for each site. This test is, however, dependent on 

 which p - 1 of the possible p(p — 1) IR's is selected. This is not true of Eq. 10, where the 

 T^ value obtained is independent of the ordering of the compartments and the 

 acceptance of H'o implies that all the possible p(p — 1) IR's in the two locations have not 

 been demonstrated to be unequal. 



From these two examples we have seen that multivariate statistical techniques can be 

 used to address such questions as the assumption of a muUiphcative relationship between 

 variables and the simultaneous test of equality of IR's, which are not easily dealt with by 

 the use of ratios or traditional univariate techniques. 



10' 



,4 _ 



Q. 



O 103 - 



I- 

 < 



I- 



Ol 



u 



o 

 o 



Q. 



00 



n 



CN 



10 



2 _ 



10 



RESUSPENDIBLES 



LIVE 

 VEGETATION 



COMPARTMENTS 



LITTER 



Fig. 13 Plutonium-238 standing crop in three ecosystem compartments for two sites. 



