90Sr: The listed doses are calculated from the model developed by Leggett et al. [15], which is based on the structure and function of bone compartments as generally outlined in the ICRP model [11]. We refer the reader to the original articles and associated references for further information [15, 16]. (Doses listed in this paper are calculated from the Leggett model.)

Ingestion: We calculated the dose equivalent from ingestion of transuranic radionuclides (^239+240Pu and ²⁴¹Am) by ICRP methods [17, 18]. The amount of ingested plutonium or americium crossing the gut wall to the blood is assumed to be 5 * 10^-4 for Pu and Am in vegetation, and 10^-5 [19] and 5 * 10^-4 for the fraction of Pu and Am, respectively, ingested via soil. Of the fraction of Pu or Am reaching the blood, 45% is assumed to go to bone and 45% to the liver [17, 20]. The biological half-life is 50 y in bone and 20 y in liver for both elements [17, 20]. The quality factor is 20 for the alpha particles.

Inhalation: The dose equivalent from inhalation for the transuranic radionuclides is based on the intake determined from the assumptions discussed in Robison et al., [5], and the ICRP new lung model dose methodology [17, 11, 20]. The ^239+240Pu and ²⁴¹Am are considered class W particles, and the quality factor is 20 for the alpha particles. Other parameters are as described in the ICRP method previously discussed for the ingestion of transuranic radionuclides. The activity-median aerodynamic diameter (AMAD) is assumed to be 1 µm, which provides a slightly conservative dose estimate (i.e., slightly higher dose) because the observed AMAD was about 2.5 µm in the Bikini experiment [21].

Data from BNL have been summarized to determine the body weights of the Marshallese people [22, 23, 24, 25, 2]. The average adult male body weight is 72 kg for Bikini, 71 kg for Enewetak, and 69 kg for Utirik. We have used 70 kg as the average male body weight in our calculations. The average biological half-life for the long-term compartment for ¹³⁷Cs in adults is listed as 110 d in ICRP [11] and NCRP [26]. This is consistent with the median of 115 d half-life obtained by BNL for 23 males [27, 28]. We used the 110 d half-life because it is based on a much larger sample population.

These doses are based on the diet listed in Table I which includes imported foods. If a diet of only local food were used, (even though such a diet is extremely unlikely today) the doses would be about a factor of 4 greater than those listed in the following tables.

The estimated maximum annual effective dose for residence on Bikini Island beginning in 1999 is shown in Table II. The maximum annual dose is defined here as the year in which the dose contribution from all exposure pathways and radionuclides is at maximum when using the intake parameters in our diet model and the mean value of the radionuclide concentrations in the foods, water and air. The estimated annual dose for Bikini without any type of remediation is 4.0 mSv.

The 30-, 50-, and 70-y integral effective doses for current island conditions are listed in Table III. The results indicate that about 90% of the estimated dose results from the ingestion of ¹³⁷Cs. The second most significant dose is received via the external gamma pathway and is almost entirely the result of ¹³⁷Cs in the soil. The contribution of the other radionuclides to the total estimated dose over 70 y is about 3%. The contribution of each exposure pathway to the estimated dose is shown in Table IV. The terrestrial food chain is by far the most significant source of dose, and that contribution is due almost entirely to the uptake of ¹³⁷Cs into locally grown food crops with their subsequent consumption by the people.

Based on the results of our dose assessment, it is clear that any effort to reduce the dose to people returning to live at the atoll should be directed toward the ¹³⁷Cs uptake into the terrestrial foods. In other words, if we could in some way greatly reduce the contribution of this pathway, we could essentially solve the dose issue for resettlement of the island.

Natural processes also will reduce the estimated doses presented in this paper. For example, ¹³⁷Cs is transferred from soil to ground water whenever sufficient rainfall occurs to produce a recharge of the ground water lens. The ¹³⁷Cs so transferred is no longer available for uptake by most plants. This continuing loss can be defined in terms of an environmental half-life analogous to the half-life of radioactive decay. Thus, the total loss of ¹³⁷Cs from the environment is the sum of these two processes. We are now in the process of evaluating data from Enewetak, Bikini, and Rongelap Atolls to determine the actual magnitude of the environmental loss. Some remedial measures are evaluated in light of this natural loss process.

Part of the cleanup of Enewetak Atoll in 1977-1979, entailed removal of surface soil from several islands to reduce the concentration of Pu. Consequently, the Marshallese are aware of this option for removing radionuclides. It is also an easy option to understand and discuss. Disposal of the contaminated soil is not a trivial issue, but is beyond the scope of this paper. This method is definitely an effective way to reduce the radionuclide inventory on the island.

However, the environmental consequences of excavation are great and long-lasting. First, all vegetation must be removed. On Bikini Island, this would include some 25,000 producing coconut palms, breadfruit trees, Pandanus, papaya and other miscellaneous food crops, as well as native vegetation. Thus, scraping off the surface 40 cm would indeed remove the radionuclides, but with them the organic layer developed over centuries. Not only does the organic layers provide nutrient exchange capacity and greatly increase rainfall retention, but it also stabilizes the surface against wind and water erosion, and renders the soil friable for root development. No island-wide inventories are available, but some of the organic-rich soils contain upwards of 5,000 kg ha^-1 N and even 10,000 kg ha^-1 P. The cost of these elements in a fertilizer bag would be several thousand dollars per hectare.

Some of our studies have demonstrated how the barren scraped surface could be revegetated. Restoring and maintaining productivity, however, requires a decades-long commitment of effort and expertise that is far from assured. In any case, the long term costs of restoration are above and beyond the $100M dollars estimated to excavate only the one square mile of surface of Bikini Island [29].

Because of the severe environmental impacts of the excavation option, we have examined other remedial possibilities that might reduce ¹³⁷Cs in the terrestrial food chain.

A large-scale field irrigation experiment was designed to determine whether a significant fraction of the ¹³⁷Cs inventory in the soil could be removed by leaching. Atolls simply do not have the large stores of fresh water needed for a single, let alone multiple leachings. On the other hand, the supply of seawater is boundless, and the contained cations might be expected to dislodge any cesium held by simple exchange forces.

The adverse consequences of such treatment are the destruction of the freshwater lens and salinization of the soil itself. However, the latter turned out to be a short-lived effect, whereas restoration of the freshwater lens takes much longer.

In the experiment, we sprayed approximately 80" of sea water on three occasions at two month intervals on a one hectare area while monitoring the downward movement of both sea water and ¹³⁷Cs. (Fig. 3) The initial pulse of ¹³⁷Cs into the ground water diminished in magnitude with each application, and represented only the small fraction that was soluble or readily exchangeable. A final total application of 20 meters (depth) of seawater removed only 3-5% of the total ¹³⁷Cs inventory. Laboratory experiments with large soil columns gave similar results.

Although 95-97% of the ¹³⁷Cs remained, plant regrowth after rains leached the soil contained very little ¹³⁷Cs. Plant concentrations increased over the next 24 months, however, and eventually returned to pre-irrigation values, demonstrating renewed availability.

The large effort required for such a small reduction in ¹³⁷Cs inventory obviously eliminates the treatment as a remedial measure.

An alternative to removing ¹³⁷Cs or the soil itself would be to immobilize ¹³⁷Cs against plant root uptake. We conducted exploratory studies with two materials: finely ground dioctyhedral mica as a surrogate for silicate clays, and clinoptilolite, a commercially available zeolite. Both bind ¹³⁷Cs so tightly that it is largely unabsorbed by plant roots.

Mica: We applied mica at the rate of 6,000 kg ha^-1 uniformly over the soil surface of a productive coconut grove. The aim was to immobilize any available Cs already at the surface and thus gradually interrupt cycling of ¹³⁷Cs from plant to soil to plant again. We now know that some mica has penetrated into the 5-10 cm depth of soil, hastening its effectiveness. Sampling of coconuts indicates a reduction in ¹³⁷Cs after 8 years or so. The reduction in shallow rooted herbaceous vegetation without the large internal inventory of ¹³⁷Cs is more impressive. For the grass Eustachys Petraea, the ¹³⁷Cs concentration in November 1987 was 1.3 Bq g^-1 in the mica treated plot and 9.1 Bq g^-1 in the control plot. In November 1990, the mica treated and control plot concentrations were 2.0 Bq g^-1 and 7.4 Bq g^-1, respectively.

Clinoptilolite: The zeolite study entailed addition of clinoptilolite at the rate of 20, 40, and 80 mt ha ^-1 to soil in small plots. The upper 30 cm of soil and clinoptilolite were mixed in a cement mixer achieving a far more intimate contact and distribution than would be possible in field scale application. Since the clinoptilolite itself contains appreciable K with some availability to plants, the study included both a conventional control and a control with added K.

We grew six successive crops on these plots, including corn and sweet potatoes. Combining mass production of the fifth and sixth crops, seedling and ratoon crops of sorghum from the same root system, illustrates treatment effects:

Using the control plus a moderate application of K as a standard, this conventional control (without added K) yielded only 61% as much biomass, but it contained 170% more ¹³⁷Cs. On the other hand, biomass yields at the lowest clinoptilolite were about equal, but contained only 71% as much ¹³⁷Cs, or only 41% as much as the conventional control. The highest rate of clinoptilolite produced about twice the biomass as the control, but contained only 12% as much ¹³⁷Cs (or 6% of the control concentration). Thus, application of either mica or clinoptilolite, (and related clays with zeolites) are capable of immobilizing ¹³⁷Cs in soil. Nevertheless, the studies involved large quantities of materials plus either intensive mixing with bare soil as rather long reaction times. Moreover, immobilization counter acts natural leaching, and largely eliminates removal indicated by an environmental half-life.

A frequently proposed concept for reducing the ¹³⁷Cs inventory in soil is to repeatedly grow and dispose of successive crops of some plants that have a high uptake of the element. Consideration of plant growth factors on atolls demonstrates the limited application of the concept.

The source of fresh water is rainfall that occurs mostly from June through November, which limits natural vegetation growth to that six month interval. The average annual mass of vegetation that can be produced is about 1 kg m^-2 or perhaps even less with continued cropping. The maximum uptake of ¹³⁷Cs in vegetation expressed as a concentration ratio (Bq g^-1 in plants (wet weight)/Bq g^-1 in soil (dry weight) is about 3. A square meter of soil, 40 cm deep, weights about 440 kg. Thus, the loss of ¹³⁷Cs is:

Also, whole coconut trees have been analyzed to determine the ¹³⁷Cs concentration in each component, and litter fall (fronds and coconuts) has been quantified to determine the annual loss of ¹³⁷Cs in the coconut grove if all the litter were collected and disposed of during the year. This loss rate is l = 0.0063 y^-1 , which is similar to the loss calculated above if annual plants covered the entire island.

Over 90 y radioactive decay (l = 0.023 y^-1) reduces the inventory of ¹³⁷Cs to 12% of its initial value. The decay constant for both processes is l = 0.023+0.0068 @ 0.030 y^-1 which over 90 y reduces the ¹³⁷Cs inventory to 6.7 % of its initial value. Thus, cropping the entire island for 90 y and disposing of the vegetation in some manner-an enormous task- would achieve only an additional 6% loss of ¹³⁷Cs. Practically, this would never happen. Moreover, our field experiments demonstrate that repeated cropping depletes the pool of available potassium so that growth greatly diminishes. This is a self-defeating cycle. Adding K to increase productivity reduces the ¹³⁷Cs concentration in plants (see previous and next sections), and consequently, the total removal of ¹³⁷Cs by cropping becomes miniscule.

The mineral matrix in soils in deep ocean atolls consist almost entirely of sands and gravels of calcite and aragonite containing small amounts of substituted Mg and Sr. Silicate clays are undetectable and the sole source of cation exchange capacity is organic matter. Total potassium content is the order of 300 mg kg^-1 [30]. Exchangeable K in the upper 25 cm of soil averages only 30-40 mg kg^-1, with the higher concentrations in the upper 5 cm of soil. The major source of K available to plants derives from oceanic spray rather than mineral weathering.

We have conducted many large-scale field-scale experiments with coconut and other plants evaluating the effect of K additions, their times of application, and the longevity of the effect [31]. Fig. 4 illustrates some of the results obtained when 1260 and 2520 kg ha^-1 of K were added to a portion of the coconut grove. In a separate experiment the addition of 440 kg ha^-1 reduced the ¹³⁷Cs in coconuts to about 20 % of pretreatment levels. Near maximum effectiveness is achieved with the addition of the 1260 and 2520 kg ha^-1 which reduce the ¹³⁷Cs in coconuts to about 5% of pretreatment levels. These results are reflected in the "Scope +K Option" column in Table I. Such reduction endures because K content increases and ¹³⁷Cs content decreases in the large stem. Ten years after application of 2500 kg ha^-1 of K, the ¹³⁷Cs concentration in coconut meat is only about 2 Bq g^-1 versus an original concentration of about 8 Bq g^-1 and low point of less than 1 Bq g^-1.

The large and prompt reduction of ¹³⁷Cs in coconuts and all other food crops tested, the ease of application, and the relatively low cost makes K addition the preferred treatment option although this was known 12 years ago. Added K also increases growth and productivity of plants on the island.

Consequently, we have developed a "combined option" that we have presented to the communities consisting of K treatment over most of the island with limited excavation of the surface soil in the hosing and village area. Such excavation significantly reduces the external gamma dose and the exposure to ^239+240Pu and ²⁴¹Am in those places where adults and children spend most of their time.

The "combined option" reduces the estimated maximum annual effective dose from 4.0 mSv y-1 (Table I) to about 0.4 mSv y^-1. The 30-, 50-, and 70-y integral effective doses (Table V) are 9.8 mSv, 14 mSv, and 16 mSv, respectively. A comparison between pre-treatment and post-treatment effective doses is given in Table VI. The "combined option" leads to estimated dose about a factor of 10 below the estimates for current island conditions.

The background effective dose in the Marshall Islands from both external and internal exposure is about 2.4 mSv y^-1 [8]. The estimated total maximum annual effective dose for a returning population is then 2.4 + 0.41 = 2.8 mSv y^-1. For comparison, the average annual effective background dose for Europe is about 2.4 mSv [32] and for the U.S. about 300 mSv [33]. Consequently, the "combined option" will lead to total doses on Bikini Island that are similar to background doses around the world. People in some areas of the world live with background doses of 1000 mSv or more [33].

The remedial options discussed with the communities over the last several years are limited to the excavation option and "combined option." The excavation option was the primary focus of the communities for years and, with all its inherent adverse environmental consequences, is still much discussed. More recently, much more attention has been given to the "combined option" which for most of the islands does not remove the ¹³⁷Cs (or other radionuclides), but simply prevents the uptake into foods until radiological decay reduces the nuclides to insignificant levels. Both the Bikini and Rongelap communities are now giving serious thought to this option.

Ultimately, it is their decision on which rehabilitation plan they will use. But in addition to doing the science, it is our job to discuss with them the plus and minuses of the options so they fully understand the ramifications of each in order to make an informed decision. This has led to many meetings with the communities over the last several years.

This work is performed under the auspices of the U.S. Department of Energy at Lawrence Livermore National Laboratory under contract W-7405-Eng-48.

1. W.L. ROBISON, W.A. PHILLIPS, and C.S. COLSHER, "Dose Assessment at Bikini Atoll," UCRL-51879, Pt. 5, Lawrence Livermore National Laboratory (1977).
2. R.P. MILTENBURGER, N.A. GREENHOUSE, and E.T. LESSARD, "Whole-Body Counting Results from 1974 to 1979 for Bikini Island Residents," Health Physics 39:177-183.
3. W.J. TIPTON, and R.A. MEIBAUM, "An Aerial Radiological and Photographic Survey of Eleven Atolls and Two Islands within the Northern Marshall Islands, EGG-1183-1758, EG&G, Las Vegas, NV (1981).
4. P.H. GUDIKSEN, T.R. CRITES, and W.L. ROBISON, "External Dose Estimated for Future Bikini Atoll Inhabitants," UCRL-51879, Rev. 1, Lawrence Livermore National Laboratory (1976).
5. W.L. ROBISON, K.T. BOGEN, and C.L. CONRADO, "An Updated Dose Assessment for Resettlement Options at Bikini Atoll-A U.S. Nuclear Test Site," Health Physics 73(1): 100-114 (1997).
6. W.L. ROBISON, W.A. PHILLIPS, M.E. MOUNT, B.R. CLEGG, and C.L. CONRADO, "Reassessment of the Potential Radiological Doses for Residents Resettling Enewetak Atoll, UCRL-53066, Lawrence Livermore National Laboratory (1980).
7. W.L. ROBISON, C.L. CONRADO, and W.A. PHILLIPS, "Enjebi Island Dose Assessment," UCRL-53805, Lawrence Livermore National Laboratory (1987).
8. W.L. ROBISON, C.L. CONRADO, and K.T. BOGEN, "An Updated Dose Assessment for Rongelap Island," UCRL-LR-107036, Lawrence Livermore National Laboratory (1994).
9. W.L. ROBISON and C. SUN, "The Use of Comparative ¹³⁷Cs Body Burden Estimates from Environmental Data/Models and Whole Body Counting to Evaluate Diet Models for the Ingestion Pathway," Health Physics 73(1): 152-166 (1997).
10. K.L. SHINGLETON, J.L. CATE, M.G. TRENT, and W.L. ROBISON, "Bikini Atoll Ionizing Radiation Survey May 1985-May 1986," UCRL-53798, Lawrence Livermore National Laboratory (1987).
11. INTERNATIONAL COMMISSION ON RADIOLOGICAL PROTECTION (ICRP), "Age-Dependent Doses to Members of the Public from Intake of Radionuclides," ICRP Publication 56, Part 1., Pergamon Press, New York, NY (1990).
12. INTERNATIONAL COMMISSION ON RADIOLOGICAL PROTECTION (ICRP), "Annual Limits of Intake of Radionuclides by Workers Based on the 1990 Recommendations," ICRP Publication 61, Pergamon Press, New York, NY (1991).
13. R.W. LEGGETT, "Predicting the Retention of Cesium in Individuals," Health Physics 50: 747-759 (1986).
14. W.L. ROBISON and W.A. PHILLIPS, "Estimates of the Radiological Dose from Ingestion of 137Cs and 90Sr to Infants, Children, and Adults in the Marshall Islands," UCRL-53917, Lawrence Livermore National Laboratory (1989).
15. R.W. LEGGETT, K.F. ECKERMAN, and L.R. WILLIAMS, "Strontium-90 in Bone: A Case Study in Age-Dependent Dosimetric Modeling," Health Physics 43: 307-322 (1982).
16. M.R. CHRISTY, W. LEGGETT, D.E. DUNNING JR, and K.F. ECKERMAN, "Age-Dependent Dose-Conversion Factors for Selected Bonej-Seeking Radionuclides," NUREG/CR-3535, ORNL/TM-8929, Nuclear Regulatory Commission (1984).
17. INTERNATIONAL COMMISSION ON RADIOLOGICAL PROTECTION (ICRP), "The Metabolism of Plutonium and Related Compounds," Publication 48, Pergamon Press, Oxford, UK (1986).
18. INTERNATIONAL COMMISSION ON RADIOLOGICAL PROTECTION (ICRP), "Limits for Intakes of Radionuclides by Workers: An addendum," Publication 30, Part 4, Pergamon Press, New York, NY (1988).
19. J.D. HARRISON, G.P.L. NAYLOR, and J.W. STATHER, "Gastrointestrial Absorption of Plutonium and Americium in Rats and Guinea Pigs after Ingestion of Dusts from the Former Nuclear Weapons Test Site at Marolinga: Implication for Human Exposure," NRPB-M196, Radiation Protection Board, Chilton, Didcot, Oxon (1989).
20. INTERNATIONAL COMMISSION ON RADIOLOGICAL PROTECTION (ICRP), "Age-Dependent Doses to Members of the Public from Intake of Radionuclides: Part 2 Ingestion Dose Coefficients," Publication 67, Pergamon Press, New York, NY (1993).
21. J.H. SHINN, D.N. HOMAN, and W.L. ROBISON, "Resuspension Studies in the Marshall Islands," Health Physics 73(1): 248-257 (1997).
22. R.A. CONARD , J.S. ROBERTSON, L.M. MEYER, W.W. SUTOW, W. WOLLINS, A. LOWERY, H.C. BRSCHEL JR, J.M. BARTON, M. GOLDMAN, H. HECHTER, M. EICHER, R.K. CARVER, P.W. POTTER, "Medical Survey of Rongelap People, March 1958, Four Years after Exposure to Fallout," BNL-534, Brookhaven National Laboratory (1959).
23. R.A. CONARD, H.E. MACDONALD, A. LOWERY, L.M. MEYER, W.W. SUTOW, B.S. BLUMBERG, J.W. HOLLINGSWORTH, H.W. LYON, W.H. LEWIS JR, A.A. JAFFEE, M. EICHER, D. POTTER, LANWI II, E. RIKLON, J. IAMAN, J., HELKEW, "Medical Survey of Rongelap People Five and Six Years after Exposure to Fallout," BNL-780, Brookhaven National Laboratory (1960).
24. R.A. CONARD et al., "Medical Survey of Rongelap People Eight Years after Exposure to Fallout," BNL-780, Brookhaven National Laboratory (1963).
25. R.A. CONARD, K. KNUDSEN, and B.M. DOBYNS, "A Twenty-Year Review of Medical Findings in the Marshallese Population Accidentally Exposed to Radioactive Fallout," BNL-50424, Brookhaven National Laboratory (1975).
26. NATIONAL COUNCIL ON RADIATION PROTECTION AND MEASUREMENTS (NCRP), "Cesium-137 from the Environment to Man: Metabolism and Dose," NCRP-52, National Council on Radiation Protection and Measurements, Washington, DC (1977).
27. R. MILTENBERGER, E. LESSARD, and N.A. GREENHOUSE, "Cobalt-60 and Cesium-137 Long Term Biological Removal Rate Constants for the Marshallese Population," Health Physics 40(5): 615-623 (1981).
28. R. MILTENBERGER and E. LESSARD, "Body Burden and Dose Assessment for Bikini Island Residents 1969-1980," Private Communication, Brookhaven National Laboratory (1987).
29. BIKINI ATOLL REHABILITATION COMMITTEE (BARC), "Report No. 4, Status March 31, 1986," Submitted March 31, 1966 to the U.S. Congress, House and Senate Committees on Interior Appropriations, pursuant to House Report 99-450, Department of Interior, Account No. TT-1580X08 (1986).
30. F.R. FOSBERG and D. CARROLL, "Terrestrial Sediments and Soils of the Northern Marshall Islands," Atoll Res. Bull 113 (1965).
31. W.L. ROBISON and E.L. STONE, "The Effect of Potassium on the Uptake of ¹³⁷Cs in Food Crops Grown on Coral Soils: Coconut at Bikini Atoll," Health Physics 62(6): 496-511 (1992).
32. UNITED NATIONS SCIENTIFIC COMMITTEE ON THE EFFECTS OF ATOMIC RADIATION (UNSCEAR), "Sources Effects, and Risks of Ionizing Radiation," United Nations sales publication E.77.IX.1, United Nations, New York, NY, Report to the General Assembly with annexes (1993).
33. NATIONAL COUNCIL ON RADIATION PROTECTION AND MEASUREMENTS (NCRP), "Exposure of the Population in the United States and Canada from Natural Background Radiation," NCRP Report No. 94, NCRP, Bethesda, MD (1987).

BACK