— the Atoll have shown that there i¢ essentially no resuspended material at 10' to 50' in height, where most of these food samples are collected (Shinn et al., 1980). Thus, uptaky via jear absorption is highly unlikely. The structure of the cora} soi] provides a very large sink in which the 90sr can be immobilized. Cora} 03) is composed primarily of CaCO3 in which 90sr is able to replace Ca in the crystal structure. As a result, a large fraction of the Sr is unavailable for dissolution in the soil solution or for absorption by plants. Moreover. plants in general tend to accumulate Ca in the leafy portion of the plants tdja greater degree than they do in the edible fruit. The plants at the Atoll also tend to have higher 90Sr concentrations in the leaf portion of the plants than in the fruits. As a result of these unique features of coral soil, the relative availability of '37cs and 90s 4, the reverse of the availability in aluminum Silicate clay soils common jn the U.S., Canada, and Europe and reported in most of the literature (Ng et al., 1982; Cline and Rickard, 1972). In silicate soils, especially when the organic content is low, the !37cs is found in the clay matrix and is relatively unavailable to plants (Barber, 1964: Kaddah, 1968; kihn et al., 1984: Menzel, 1965: Fredriksson, 1963: Fredriksson et al., 1958, 1966; D'Souza at al., 1972, 1980; Tamura and Jacobs, 1960). In contrast, the !37Cs js mobile and relatively available in coral soils. Correspondingly, %0Sr is relatively available in silicate-type soils but is bound in the carbonate matrix and js relatively unavailable to plants in coral soils. This information has important ramifications, if radionuclide concentrations in soil are to be used to estimate the concentration of the radionuclide in edible foods for radiological dose assessment. If C.R.s based on silicate soils were used for such a purpose, the estimated 137Cs concentration in plants at coral] islands would be in serious error and would underestimate the potential effective dose to people residing on an island, and the concentration of %sr would be greatly overestimated and its contribution to the bone marrow dose would be exaggerated. The transfer of Pu and Am from the coral soil to plants is very low, similar to that observed in other soi] systems. The C.R. for Pu and Am ranges between 10-4 and 10-5. | 67