Transuranics and Other Radionuclides in Bikini Lagoon: Concentration Data Retrieved from Aged Coral Sections (Reprinted from Limnology and Oceanography Vol. 20, No. 5 9/75, PP. 729-742)

Page 1

TRANSURANICS AND OTHER RADIONUCLIDES IN BIKINI LAGOON: CONCENTRATION DATA RETRIEVED FROhf AGED CORAL SECTIONS Reprinted from LIMNOLOGY.khTD OCEANOGRAPHY Vol. 20, No. 5, September 1975 Made in the United States of America @ Copyright, IW5, I)Y The American Society of Limnology and Oceanography, Inc.

Page 2

Transuranics and other radionuclides in Bikini Lagoon: Concentration data retrieved from aged coral sections’ Bio-Medical Division, Lawrence Livermore Laboratory, University of California, Livermme 94550 X-radiography and autoradiography of thin vertical sections were used to estimate the growth rate of a specimen of Faoites uirens from Bikini Lagoon. Discrete bands of radioactivity y were identifiable with specific nuclear test series. The coral growth rate of 8.0 mm year-’ determined by autoradiography is in good agreement with the rate of 8.1 ti 2.2 mm year-’ derived from the “seasonal” alternating light and dark bands on X-radiographs. With these bands as growth rate indicators, the coral was sectioned into yearly increments and analyzed by low-level, nondestructive gamma spectr~me~, radiochemical techniques, and mass spectrometry to reconstruct the variations in the concentration of transuranics and other radionuclides in the marine environment at Bikini since 1954. From the concentration data retained in this indicator species, the exchange rate of radionuclides between the lagoon and the open ocean is computed to be longer than exchange rates based on physical circulation data. There is no constant ratio of plutonium isotopes in the coral growth sections, suggesting that the redistributions of the several plutonium isotopes in the environment may be governed by different biogeochemical proing previous arguments that no ‘l”Pb has resulted from weapons testing. Knutson et al. ( 1972), Knutson and Budrlemeier ( 1973), and Buddemeier et al. ( 1974) demonstrated that massive coral colonies from Enewetak Atoll contain radioactive growth inclusions, detectable by autoradiography, that can be correlated in X-rayed sample sections. Results from these and other samples ( Buddemeier et al. 1974 ) show periodicity in growth, interpreted as density variations in the CaCO~ skeIeton, that can be used to estimate growth rates of massive coral colonies. Other recent applications, using naturaI son 1974; Moore and Krishnaswami 1972), have also used radiometric techniques to determine coral growth. Radiometric studies ( Dodge and Thomson 1974) have substantiated the relationship between annual growth-band composition and the changing environment in which the coral grows. Previous investigations of the growth rates of fossil and reef coral have been adequately ‘ Work was performed under the auspices of the U.S. Energy Research and Development Adreviewed ( Buddemeier et al. 1974; Dodge Our studies began in an attempt to reconstruct changes in transuranium concentrations in the marine environment of the Pacific test sites through observations of the yearly response of the coral to environmental changes, Although several radiological surveys have been made at the Pacific test sites, no chronological information is available on the behavior of specific radionuclides in these environments. The radiological records retained in the cIosed aragonite structure of coral could circumvent the need for years of real-time data and provide insight into the behavior of these radioelements in the lagoon. Knutson and Buddemeier (1973) found 90Sr in coral from Enewetak and concluded, on the basis of the observed differential changes in concentration, that the lagoon community acts as a long term source of 90Sr, probbonate materials deposited during the period of nuclear testing. It is not clear that a single mechanism can explain all environmental radiological information retained by coraI colonies; we need inforLIMNOLOGY AND OCEANOGRAPHY 729 SEPTEMBER 1975, V. 20(5)

Page 3

730 Noshkin et al. Fig. 1. Bikini Atoll—location of sampling station and Bravo Crater. Isoactivity lines of ‘“Am mation on other possible mechanisms controlling the redistribution of specific radionuclides, principally the transuranium We prepared autoradiographs and Xradiographs from vertical slices cut from the center of a sample of Favites virens collected from Bikini Atoll. Autoradiography revealed three discrete, relatively intense, radioactive bands within the skeletal structure identifiable with the nuclear test series of 1954, 1956, and 19.58. Fourteen density bands between the outermost radioactive band (1958) and the coral surface ( 1972) confirmed the thesis that these are correlated with annual growth. Sections from each annual growth band were analyzed by low-level gamma spectrometry and radiochemical techniques to reconstruct the chronological history of specific radionuclides incorporated in the coral skeleton. Much, if not alI, of the observed artificial radioactivity in the coral at Enewetak ( Knutson and Buddemeier 1973) was due to 90Sr and its daughter ‘OY, but we identified other radionucIides in all growth sections of this Bikini coral (e.g. z3s.z39.z40,wPu ). we believe that the record ments is one of the most complete histones of changes in concentration of specific radionuclides in the marine environment of Bikini Atoll. From these data we explain the relative behavior of some specific radionuclides in Bikini Atoll. This work was part of a cooperative program to investigate the biogeochemical behavior of the transuranium elements at Bikini Atoll. Samples were collected in November 1972 from the RV Pahnnbo, operated by the Puerto Rico Nuclear Cenprovided by the crew of the RV Palumbo and specifically thank F. Lowman and W. Sectioning methods and growth results The coral used in this study was a live ber 1972 from a depth of 28 m at station B-3, Bikini Atoll ( Fig. 1), In the periods Atoll was the site of a series of nuclear device tests. Many of the events in the last three series were conducted aIong the inner northern and southern perimeter of the lagoon. The largest test on the reef ( 1954: 15 megatons) ( Telegadas 1961 ) produced Bravo Crater, also shown in Fig. 1. and ‘O’Bi in the sediment column from location B-3; water depth, 28.3 rrq collection date, 8 November *Based on counting error only.

Page 4

Transuranics in Bikini Lagoon 731 data we explain wledge the help study was a live 1. In the periods ies of nuclear dewents in the last ted along the in1 perimeter of the in the reef ( 1954: Fig, 2. Autoradiograph of a cut vertical section from the coral sample. Radioactive regions have been retouched for better contrast. The distributions of specific radionuclides in the lagoon are extremely heterogeneous ( Noshkin et al. 1974). For example, the 1972 activity levels and isoactivity lines of ‘~lAm in the surface 2.5 cm of sediment deposit (Fig. 1 ) show that the region around station B-3 was heavily contaminated with artificial radionuclides from the test series. Table 1 gives the levels of several radionucIides to a depth of 35 cm in a sediment core from station B-3. Levels of radioactivity are relatively high throughout the column. The coral from this location should reflect the changes in activity during its growth cycle in this contaminated region of the lagoon. The coral surface was washed and the specimen was dried at llO” C for 1 week. The sample was first cut in half vertically from the surface through the base. (We always cut from presumably low-radioactivity sections toward higher level sections to

Page 5

732 Noshkin et al. Fig. 3. Demonstration X-radiograph of a coral section. Annual band boundaries (not all necessarily evident on this exposure) are identified. Wedge outlines the section removed for radionuclide analavoid possible contamination. ) Vertical subsections 2 cm thick were cut from the facing coral halves. One slice, chosen for autoradiography, was placed in direct contact with lNo-screen-X-ray film for 60 days; the adjacent slice was subjected to X-radiography, with several exposures at different) Regions of the developed autoradiography film revealed inclusions of radioactivity ( Fig. 2: these bands are darkened for better contrast ). Three distinct bands ,C., near the base of the coral can be identified “, with the test series years 1954, 1956, and 1958; they are 15, 14, and 12.5 cm from the coral surface. These measurements gave average growth rates of 0.81, 0.85, and 0.87 tively, and 1972. The distance between the 1954 and 1958 bands, however, is only 2.5 cm, yielding an average grovvth rate during this period of only 0.6 cm yrl. The annual skeletal layers are not of constant-thickness; in assigning the annual boundaries to posttest growth, we relied on the density variations in the exposed X-ray film. Figure 3 shows one exposed X-ray film of the coral section. Careful examination reveals alternating light and dark bands in the skeletal matrix. Several exposures were examined before a complete selection of annual band “boundaries” could be made. Buddemeier et al. ( 1974) concluded negative ) is correlated with a growth period beginning around December or January and ending around July. In our sample, the three bands in the autoradiograph of

Page 6

wever, is only 2.5 md dark bands in plete selection of “ could be made. Iy. In our sampIe, Transuranics in Bikini Lagoon 733 Table 2. Coral section data. Growth Estimated bun At in years section year of growthf from Nov 72 from growth year (collection date) the Bikini coral are either associated with the low-density bands of the X-ray negative or close to the transition zone between light and dark bands. By correlation of the bands shown by both the autoradiograph and X-ray exposures, the years 1956 through 1958 were identified. The 1954 growth, however, could not be resolved from the 1955 growth on the X-ray negative. From the 1958 band to the coral surface ( 1972 growth), 14 alternating light and dark areas were identified ( Fig. 3). TabIe 2 gives the thickness of each annual section along with the estimated year of The varying dimensions of the yearly increments show that growth rates vary from year to year and also that the dimensions follow no predictable trend with time. The mean annual growth from the X-radiograph is 8.1 * 2.2 mm yr-l, a value in agreement with the autoradiography results. No activity from the 1946 test was detected because it predated the earliest growth of this particular coral. A wedge, containing the area with the more nearly linear growth record, was removed from the center of the slice (see wedge outlim, Fig. 3). A bandsaw was used to remove each defined annual growth band; at least 1 mm of material was lost at each cut, so that the coral record between each two sections was lost. Each section was then ground and homogenized in a ball mill and a known weight was transferred to a vial for radionuclide analysis by gamma spectrometry. A low-backtector system was used for some of the ples were then processed for Pu, 90Sr, and tectors and alpha spectrometry. Selected samples were also analyzed by mass specsentially those described by Wong (1971) and Noshkin and Gatrousis ( 1974). Stable strontium was determined by atomic absorption. Our discussion here will concentrate primarily on the transuranium elements; however, Tables 3-5 give the results for all the radionuclides analyzed, rected for decay back to the estimated year Two sources contribute to the “1Am leveIs found in the coral. Part is from the radioactive decay of the parent radionuclide “1Pu, while the remainder is unsupported “1Am. Based on the quantities of 241Am and 241Pu, the time between separation, and the age of the coral section, levels of both supported and unsupported 2dlAm are computed and given in Table 5. Geruwal-The coral bands identified with the 1954, 1956, and 1958 test series, as would be expected from the autoradiograph, contain the highest concenb-ations of radionuclides. A spectrum from 12 g of the 1954 growth section is shown in Fig. 4 with each prominent gamma ray identified. Severai gamma-emitting radionuclides with half-Iives less than 1-2 years have been identified in earlier surveys ( Welander 1969 ); these were not detected in any coral sections with our spectrometer system. Other radionuclides requiring radiochemi-

Page 7

and pl utoniurn activity ratios. Estimated year of pCi/g dry weight - decay corrected to year of Srowth + % error *Determined by aipha spectrometry. +Determined by gamma spectrometry. Table 4. Other gamma-emitting radionuclides detected, decay corrected to year of coral growth. Estimated year of pCi/g dry weight f % error *A blank space indicates the data were not collated or the concentration %<s.s below detection

Page 8

Transuranics in Bikini Lagoon 735 Period of nuclear test . series at Bikini Atol 1 Estimated year of coral growth analyzed Fig. 5. Concentration of ‘9Pu and the PhLtonium isotope activity ratios in each annual seccal separation from the coral (such as 55Fe) and known to be present in atoll samples ( Welander 1969) were not determined. The concentration of 239Pu and the plutonium isotope activity ratios in each annual band, decay-corrected to year of coral growth, are shown in Fig. 5. In surface water samples collected from the north equatorial current or east of Bikini, the ZM+WOpuconcen~ation averaged average in 10 filtered lagoon samples tion in the lagoon averaged 570 fCi literl compared to 71 in the surface waters outr of coral growth. Fig. 4. &unma-ray spectrum of the 1954-1955 growth section. Unidentified photopeaks are from background and naturally occurring radionuclides.

Page 9

decay corrected to year of growth. Un.suvported and supported ‘J’Am concentrations. supported by Pu decay. side and east of the lagoon. Radioelernents tecte’d by gamma spectrometry in any water sam~les collected outside the atoll but. were prominent in biota and sediment samples ( Fig. 1, Table 1 ) from the lagoon. We conclude that the atoll is the principal source of radionuc]ides to the Iagoon enOur investigation differs from others attempting to use corals to determine the trace element composition of adjacent waters. We are not comDarinE concentrations whole specimens from different areas. We expect coral samples from different lagoon locations to have siznificantlv different ab-. solute radionuclide concentrations, and preliminary data have verified this assumption. Having only a single sample from one location, however. we need on~J assume for. each element or’ radionuclide that each yearly increment of coral growth concentrates the same available fraction from seawater per unit weight of coral; then, Fig. 6. Box model diagmm of Bikini Lagoon describing the general flow, and sources of radionuclides into and out of the lagoon. growth band is proportional to the concentration in the adjacent environment. On this basis, we find that all detected radiothe lagoon by two to three orders of magnitude since the test years. The rate at

Page 10

n of Bikini Lagoon Id sources of radio- “) in the annual Ial to the concenave decreased in time is not con- : was most rapid but real changes rith the peak in Ierived from only le atoll may lead ter correction for it recycling from and surface runoff, or some combination of these or other biogeochemical processes, are responsible for replenishing activity levels of some radionuclides in the lagoon at rates that compensate for the rates of reNo model can adequately account for these unanticipated changes in concentration unless the specific mechanisms responsible for the changes are understood. However, even though a degree of uncertainty exists, we can use a simplified model of the Iagoon environment based on the diagram in Fig. 6 and the coral data to describe the rate at which the radionuclides are exchanged between the lagoon and open