In making these dose reduction approximations, one must keep two things in mind; first, that the NVO-140 does estimates for terrestrial foods grown an island such as JANET are based on correlations between certain indicator plants and average soil concentrations in the 0-15 cm samples (Fig. 3) since foods such as pandanus and breadfruit were not found on JANET and, second, that these concentrations are averaged over the 0-15 cm depth of Figs. 4 anc Estimates of dose reductions to be expected due to removal of soil to a give depth, therefore, require an estimate of the ratio of the average concentrat of the nuclides of concern in the 0-15 em depth of the newly exposed surfac to that for the surface which is present now. This approach does not consi. the radioactivity in the soils deeper than 15 em which may be inportant, particularly for plants with roots that penetrate deeply into the soil. Ta presents these average concentrations and ratios for 905, and 1370, for eac increment from the present surface down to 105 cm as derived from Figs. 4 < These estimatcs indicate, for example, that removal of 15 cm of soil may rc the terrestrial food dose due to 905, by a factor of 3.3 and that due to 137¢6 by 3.2. However, such reduction may or may not be actually achieved There is no experience to support these reduction levels. Using the data of Table 8, one may assess the dose reductions that mi occur due to specific cleanup actions on JANET. Table 9 shows the doses that might occur due to seven different conditions. Case D represents the contributors to the 80 Rem bone dose of Table 2 using values for 0c 1376, averaged over all of JANET. Case Dl indicates that if subsistence agriculture is limited to the area shown in Fig. 2 (i.e., along the lagoa Sr and 137 : Cs levels may be reduced to such an extent that th< resulting 30 yr bone dose becomes 57 Rem. Removal of a half-thickness of mf. 90 AR shore) the