M.H. Paller and J.W. Littrell
Westinghouse Savannah River Company
Aiken, SC, 29808, USA

Eric L. Peters
Chicago State University
Chicago, IL 60628-1598

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/index.html

Available electronically at  http://www.osti.gov/bridge

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

Ecological half-lives (T_e’s) were estimated for ¹³⁷Cs in largemouth bass, sunfishes, and bullheads from two reservoirs and three streams on the Savannah River Site, a nuclear weapons material production facility in South Carolina. Ecological half-life is the time required for a given contaminant concentration to decrease by 50% as a result of physical, chemical, and/or biological processes that remove it from an ecosystem or render it biologically unavailable. T_e’s were estimated from whole-fish ¹³⁷Cs concentrations recorded during 1972-1996, following radionuclide releases that occurred primarily during the 1960s and 1970s. T_e’s ranged from 3.2 to 16.7 yr, and all were shorter than expected from the half-life for radioactive decay (T_p = 30.2 yr) alone. Fish taxa from the same locations differed in mean ¹³⁷Cs concentrations (highest in largemouth bass and lowest in sunfishes) but, in most cases, exhibited similar ¹³⁷Cs T_e’s. The shortest T_e’s occurred in the middle portions of the streams. T_e’s in lower portions of the streams were longer, as were T_e’s in one of the reservoirs. T_e’s in Par Pond, a second reservoir with a much shorter water residence time, were nearly comparable to those in the upper portions of the streams. However, in 1991, ¹³⁷Cs concentrations in Par Pond fishes began to increase following drainage and refilling of the reservoir, which apparently resuspended ¹³⁷Cs buried in the sediments. Recent samples collected during 1997-2000 demonstrated that ¹³⁷Cs levels in fish have continued to decline except in Par Pond and a stream into which Par Pond discharges, where elevated ¹³⁷Cs levels associated with the Par Pond drainage and refill still persist.

Environmental releases of cesium-137 (¹³⁷Cs) from nuclear facilities have demonstrated that hydrospheric pathways can be a major source of human exposure. Although ¹³⁷Cs residence times in natural waters are often comparatively short¹, some ¹³⁷Cs enters aquatic food chains and bioaccumulates in the edible skeletal muscle of fishes ². Fishes assimilate ¹³⁷Cs rapidly and eliminate it relatively slowly³ making consumption of fishes a potentially important source of human exposure.

¹³⁷Cs concentrations in fish would be expected to decline over time as a result of physical, chemical and biological processes that remove ¹³⁷Cs from the ecosystem or make it biologically unavailable. "Ecological half-life" (T_e) describes this decrease in contaminant concentration in an ecosystem over time. T_e estimates can be useful in determining how long (and to what degree) contaminated systems may pose an environmental or health risk, and can aid in assessing the success of remediation.⁴

At the Savannah River Site (SRS), a U.S. Department of Energy (DOE) nuclear production facility in South Carolina, there is concern about possible human health effects resulting from the consumption of fishes contaminated by ¹³⁷Cs released during the 1960s and 1970s. This concern is unlikely to abate until contamination in fishes reaches levels considered acceptable for human consumption. Environmental monitoring of fishes in SRS streams and reservoirs has produced a unique long-term data set for the estimation of T_e’s for ¹³⁷C in freshwater fishes from warm-water lakes and streams. In contrast, most published studies of ¹³⁷Cs in lentic ecosystems are for cold-water, highly acidic European lakes that received fallout from the Chernobyl accident ^{4, 5} or tropical marine ecosystems.⁶

This paper summarizes information previously presented in an article published in Health Physics ⁷. However, it includes additional information collected more recently concerning the continuance of trends noted in the first paper.

The SRS, located on the Atlantic coastal plain near Aiken (South Carolina, USA) is an 803-km² nuclear weapons materials production site established in 1951. By 1955, the SRS had five operating nuclear production reactors. Water from the Savannah River was pumped through the cooling circuits of these reactors, before being released into three tributaries of the Savannah River (Steel Creek, Four Mile Creek, and Lower Three Runs) and two reservoirs (Par Pond and Pond B). Various operational problems resulted in releases of ¹³⁷Cs into the cooling waters discharged from four of these reactors. It is estimated that releases of ¹³⁷Cs were approximately 1.220 TBq from C-Reactor into Four Mile Creek; 8.200 TBq from R-Reactor into Pond B, Par Pond, and Lower Three Runs Creek (4.70 TBq to Pond B alone); and 9.435 TBq from P-Reactor into Steel Creek and Par Pond. An additional 1.070 TBq of ¹³⁷Cs entered Steel Creek from L-Reactor. The vast majority of these ¹³⁷Cs releases occurred from 1960 through 1970, although minor releases continued until 1985. Additional information on the timing, quantity, and points of release can be found in Paller et al.⁷

Analysis of ¹³⁷Cs in fishes of the SRS began in 1961, and many sites were sampled annually or every few years. Collection sites included Steel Creek near SRS Road A (middle reaches of Steel Creek), the mouth of Steel Creek near its confluence with the Savannah River, Four Mile Creek near Road A (middle reaches of Four Mile Creek), Par Pond, Pond B, and Lower Three Runs Creek near Patterson Mill (middle reaches of Lower Three Runs). Means, maximums, and sample sizes were documented in annual reports, which are the source of most of the data analyzed in this paper, although additonal data concerning ¹³⁷Cs in Pond B largemouth bass, were from the University of Georgia, Savannah River Ecology Laboratory. Collection methods generally included angling and the use of traps.

Fishes were separated into several categories, including largemouth bass (Micropterus salmoides), sunfishes (primarily redbreast sunfish, Lepomis auritus, bluegill, L. macrochirus, and spotted sunfish, L. punctatus), and bullheads (primarily yellow bullhead, Ameiurus natalis and flat bullhead, A. platycephalus). Combining related species, as practiced at the SRS, introduced an indeterminate source of variability in the estimates of ¹³⁷Cs concentration, because closely related species may differ in ¹³⁷Cs body burdens. Seasonal and age/size-dependent changes in ¹³⁷Cs body burdens were also potential sources of unmeasured variability in this study (although all fish were large enough to be considered potentially edible). However, these sources of variability were likely trivial compared with the large long-term decreases in ¹³⁷Cs concentrations that were the subject of this study. These issues are iscussed in greater detail in Paller et al.⁷

Individual whole fish were analyzed prior to 1992. Composite samples, each consisting of fillets from five fish, were analyzed in subsequent years. Samples collected after 1989 were counted with a shielded high purity germanium detector. Earlier samples were counted with NaI(Tl) solid scintillator or Ge(Li) semiconductor detectors. Paller et al⁷can be consulted for more details on analytical procedures.

¹³⁷Cs concentrations were measured in water as well as in fishes. Composited water samples were generally collected with continuous automatic water samplers at the same locations where fish samples were taken. Unfiltered water samples were counted in filled 500-ml containers.

Because ¹³⁷Cs concentrations were measured in whole fishes prior to 1992 and fillets afterwards, taxon-specific regression models (based on samples for which both fillets and inedible portions were analyzed) were used to predict concentrations in whole fishes from concentrations in fillets. The equation describing the loss of ¹³⁷Cs from whole fishes was dc/dt =-T_ec, where c was the ¹³⁷Cs concentration (Bq g^-1) in whole fish, and T_e was the ecological loss rate constant. Estimates of T_e (k_e) were made from the slopes of log_e-transformed ¹³⁷Cs concentrations in fishes regressed against year. The ¹³⁷Cs concentrations used in these calculations were annual means, weighted by the sample size. For composited samples, the number of fish was considered equal to the number of composited samples multiplied by the number of fish in each sample. The ecological half-life (T_e) for ¹³⁷Cs for each species group at each location was estimated as T_e = log_e 2/k_e. No attempt was made to correct for ¹³⁷Cs from global fallout, because historical samples collected from local water bodies unaffected by SRS releases indicated that this source was insignificant compared to reactor releases.

Because the equation used to compute T_e was appropriate only when ¹³⁷Cs inputs are zero, the analysis was restricted to data collected after 1971, when ¹³⁷Cs releases from SRS reactors were nonexistent or trivial. The last year included in the analysis was 1996, except in the case of Par Pond where data from 1991-1996 were excluded from analysis. Par Pond was partially drained and refilled during these years resulting in increases in increases in ¹³⁷Cs bioavailability as discussed later.

Analysis of covariance (ANCOVA) was used to determine if rates of ¹³⁷Cs decrease differed among taxa at each sample location. Log_e-transformed ¹³⁷Cs body burden was the dependent variable, time was the covariate, and taxon was the factor. Least square means were calculated for each taxon at the average value of the covariate.

T_e’s reported in our earlier paper⁷ were based on data collected from 1972 to 1996. Since then, an additional four years of data have become available (1997-2000) from most locations. These data have not been included in the T_e computations (which remain the same as in the earlier paper¹⁰), but they have been graphed on the T_e regression plots to illustrate the extent to which recent data correspond to previously documented trends. These regression plots include 95% prediction intervals, which exhibit the expected variance of future individual ¹³⁷Cs concentrations.⁸

Maximum ¹³⁷Cs concentrations in water occurred between 1960 and 1970 corresponding to the period of maximal ¹³⁷Cs release from the reactors ⁷. The highest ¹³⁷Cs concentrations in water, up to » 23 Bq L^-1, occurred in Steel Creek. Maximum concentrations in the other water bodies were lower (» 1 Bq L^-1). ¹³⁷Cs concentrations in all aquatic systems declined markedly after 1970. These patterns were paralleled by those observed in fish, which also reached maximum levels between 1960 and 1970 and then declined. The highest ¹³⁷Cs concentration in fishes, 53.9 Bq g^-1, occurred in Steel Creek. Maximum concentrations in Par Pond and Four Mile Creek fishes were substantially lower (6.6 and 5.8 Bq g^-1, respectively).

The relationship between ¹³⁷Cs concentration and time was highly predictive for most taxa within most water bodies, as indicated by comparatively high regression coefficients (r²) and significance levels for the regressions (Figure 1, Table 1). T_e’s ranged from 3.2 to 16.7 yr (Table 1). The shortest T_e’s occurred in the middle reaches (near Road A) of Steel Creek (3.2-3.5 yr) and Four Mile Creek (4.7-5.0 yr, Table 1). The T_e at the mouth of Steel Creek was longer (7.0 yr), as were the half-lives in Pond B and Lower Three Runs (10.7 to 16.7 yr).

T_e's of different taxa inhabiting the same water bodies were similar (Table 1). This was verified by analysis of covariance (ANCOVA), which indicated that the slopes of the regressions of whole-body ¹³⁷Cs concentration on time did not differ significantly (P<0.05) among taxa at each location. However, mean ¹³⁷Cs concentrations differed among taxa, even though T_e's were comparable. Least-square geometric mean ¹³⁷Cs levels were 140, 240, and 300 mBq g^-1 (P=0.008) for Steel Creek sunfishes, bullheads, and largemouth bass; 360 and 610 mBq g^-1 (P<0.001) for Four Mile Creek sunfishes and largemouth bass; and 130 and 230 mBq g^-1 (P=0.04) for Par Pond sunfishes and largemouth bass, respectively.

Comparisons of recently collected data with Te regression plots that were based on 1972-1996 data (or 1972-1991 data in the case of Par Pond) indicated that past trends of declining ¹³⁷Cs levels in fish have continued in most water bodies (Figure 1). However, a different pattern occurred in Par Pond and Lower Three Runs, where ¹³⁷Cs levels increased markedly after 1991 and exceeded levels that would be expected based on previous trends (Figure 1). These increases began with and continued during the partial draining and refilling of Par Pond, which extended from 1991 to 1996. Levels were still elevated four years after refill was complete (2000), especially in the case of largemouth bass from Par Pond. Lower Three Runs, the only other water body where ¹³⁷Cs levels in fish have not declined, constitutes the tailwaters of Par Pond and is affected by water released from this reservoir.

Figure 1. Changes in ¹³⁷Cs in whole fish regressed against time for fishes at the Savannah River Site.
Circles represent data used to compute ¹³⁷Cs half-lives (T_e’s).
Triangles represent recent data collected in subsequent years.
Symbol areas are proportional to the number of fish in each sample.
Solid lines represent regression lines; dotted lines represent 95% prediction intervals.
Statistics for the regression lines appear in Table 1.

The T_e’s observed for ¹³⁷Cs in fishes were considerably shorter than the 30.2-yr radioactive half-life of ¹³⁷Cs, and were presumably the result of various physical and ecological processes that removed ¹³⁷Cs from the aquatic ecosystems under study and/or made it less bioavailable. However, there were substantial differences among ecological half-lives for ¹³⁷Cs in fishes from different water bodies. The comparatively short half-lives in the middle portions of Steel Creek and Four Mile Creek (Road A) (3.2 and 5.0 years) may have resulted from the outflow of sediment-bound and dissolved ¹³⁷Cs and/or the burial of contaminated sediments by uncontaminated sediments. ¹³⁷Cs-contaminated sediments removed from the middle reaches of Steel Creek may have been deposited downstream near the mouth of Steel Creek, perhaps accounting for the longer half-life at this location. The comparatively long half-life in Lower Three Runs Creek (10.7 years) may have resulted from the discharge of dissolved and seston-associated ¹³⁷Cs from Par Pond.

The T_e’s reported for Par Pond fishes represent the period from 1972 to 1990, when changes in ¹³⁷Cs levels resulted from the processes of physical and biological removal that operated in this ecosystem before disturbance. These T_e’s (4.8 and 5.0 years) were only slightly longer than for Steel Creek or Four Mile Creek fishes (Table 1). In 1992, however, ¹³⁷Cs levels in Par Pond largemouth bass and sunfish increased, probably because of resuspension of sediments and/or changes in water chemistry associated with the draining and subsequent refill of Par Pond. During and following the refill, approximately one-half of the lake bottom was exposed, permitting sediments to erode into the lake basin. Since the pumping of water from the Savannah River into Par Pond was also stopped at this time, most of the water entering Par Pond during 1991-1996 was rainfall and runoff from the local watershed. This water was low in dissolved potassium compared with Savannah River water. The resulting decrease in dissolved potassium levels may have further contributed to the increased ¹³⁷Cs concentrations observed in fishes after 1992.⁹ These same factors are probably also responsible for the increases in ¹³⁷Cs recently observed in fishes from Lower Three Runs, which received water discharged from Par Pond.

The longest T_e’s for ¹³⁷Cs in fishes were observed in Pond B. In addition to the relatively long T_e’s for largemouth bass and sunfishes observed in this study (16.7 and 13.4 years, respectively), a 50 yr T_e for ¹³⁷Cs in bullheads from Pond B has been reported by other researchers.¹⁰ Factors that may contribute to relatively long T_e’s in Pond B include a relatively low rate of water turnover, low potassium concentrations, and large amounts of rooted aquatic macrophytes (that may translocate sediment-bound cations into the water column.¹¹ The exceptionally long T_e reported for Pond B bullheads¹⁰ suggests a unique exposure scenario for this species (due to habitat or feeding preferences) or changes in ¹³⁷Cs bioavailability in Pond B. Further research will be required to determine causes with certainty.

The results of this study are comparable with those of other studies concerning ¹³⁷Cs T_e’s in fish. An earlier study of Pond B fishes indicated that ¹³⁷Cs T_e’s ranged from approximately 5 to 19 years.¹² T_e’s reported herein were also similar to the 9 to 12 yr range reported for fishes from lagoons in Bikini and Enewetak atolls.⁶ In contrast, the T_e’s observed at SRS were often longer than those observed for fishes in European water bodies after contamination by Chernobyl fallout. T_e’s for several years after deposition ranged from 1.0–2.0 yr for fishes at lower trophic levels (perch and roach) to about 1.0–2.9 yr for brown trout and Arctic charr to 2-4.9 yr for pike.^3,4,13 T_e’s for a variety of fishes (nine species) from the Chernobyl NPP cooling pond were generally under two years.¹⁴ Reasons for these relatively short T_e’s. are uncertain, however, removal rates for ¹³⁷Cs in Steel Creek appeared to be more rapid in the years initially following release (see Paller et al.¹⁰), offering a possible parallel to the rapid T_e’s observed near Chernobyl. In terrestrial ecosystems, some ¹³⁷Cs is bound loosely to terrestrial soils and may be removed rapidly by flushing.¹ Similar phenomena may occur in aquatic ecosystems, resulting in a rapid initial rate of decrease in ¹³⁷Cs levels in fishes.

Data from streams and reservoirs on the SRS indicated that ¹³⁷Cs T_e’s in fishes were substantially shorter (with the exception of bullheads from Par Pond and Pond B) than expected from the rate of radioactive decay alone. Particularly high rates of removal were observed in areas characterized by high rates of water turnover, rather than high sedimentation rates (e.g., middle portions of Steel Creek and Four Mile Creek). Should current T_e trends continue, fishes from most or all contaminated ecosystems on the SRS will likely reach ¹³⁷Cs concentrations acceptable for consumption in not more than 50 yr. This consideration is important, because potential consumption of radioactive fishes is considered a principle impediment to unregulated human use of the SRS.

Data from Par Pond indicate that anthropogenic disturbances may result in relatively persistent increases in ¹³⁷Cs levels in fishes. By inference, it is also possible that invasive cleanup efforts could increase radiological risk to the public by resuspending buried radionuclides, possibly increasing their bioavailability and incorporation into human food webs. In contrast, data from other SRS water bodies suggest that natural succession and concomitant sedimentation processes in relatively undisturbed waters will often lead to burial and sequestration of bioavailable radionuclides such as ¹³⁷Cs until they decay.

We thank the many people who collected fishes and analyzed them for radionuclide levels over the years at the SRS. We also thank I.L. Brisbin, Jr., J. Peles, J.E. Pinder, III, M.H. Smith, and D. Sugg of the University of Georgia Savannah River Ecology Laboratory for sharing their detailed data on Pond B bass. The information contained in this report was developed during the course of work conducted for the U.S. Department of Energy under Contract No. DE-AC09-96SR18500.

1. P.H. Santschi, S. Bollhalder, S. Zingg, A. Lück, K. Farrenkothen, "The self-cleaning capacity of surface waters after radioactive fallout. Evidence from European waters after Chernobyl, 1986-1988," Environmental Science and Technology 24: 519-527, (1990).
2. S.E. Kolehmainen, "The balances of Cs-137, stable cesium, and potassium of bluegill (Lepomis macrochirus) and other fish in White Oak Lake", Health Physics 23: 301-315, (1972).
3. T. Forseth, O. Ugedal, B. Jonsson, A. Langeland, O. Njastad, "Radiocaesium turnover in arctic charr (Salvelinus alpinus) and brown trout (Salmo trutta) in a Norwegian lake", Journal of Applied Ecology 28:1053-1067, (1991).
4. J. Brittain, U. Bergström, L. Håkanson, R Heling, L. Monte, V. Suolanen, "Estimation of ecological half-lives of caesium-137 in lakes contaminated by Chernobyl fallout", Proceedings of the International Symposium on Environmental Impact of Radioactive Releases, Vienna, Austria, May 8-12, 1995, pp 291-298, International Atomic Energy Agency, IAEA-SM-339/58, Vienna, Austria, (1996).
5. L. Håkanson, J. Brittain, L. Monte, R. Heling, U. Bergström, V. Suolanen, " Modeling of radiocesium in lakes–the VAMP model", Journal of Environmental Radioactivity 33:255-308, (1996).
6. V.E. Noshkin, W.L. Robison, K.M. Wong, J.L. Brunk, R.J. Eagle, H.E. Jones, "Past and present levels of some radionuclides in fish from Bikini and Enewetak Atolls", Health Physics 73:49-65, (1997).
7. M.H. Paller, J.W. Littrell, E.L. Peters, "Ecological half-lives of ¹³⁷Cs in fishes from the Savannah River Site", Health Physics 77:392-402, (1999).
8. R.R. Sokal, F.J. Rohlf, Biometry, second edition, p. 477, W.H. Freeman and Company, San Francisco, (1981).
9. F.W. Whicker, T.G. Hinton, D.J. Niquette, "Effects of a partial drawdown on the dynamics of ¹³⁷Cs in an abandoned reactor cooling reservoir", Studies in Environmenal science: Freshwater and Estuarine Radioecology, Hilton, Desmet, Comans, Blust, deBetencourt, eds., pp.193-202, Elsevier, Amsterdam (1997).
10. C.D. McCreedy, C.H. Jagoe, L.T. Glickman, I.L., Jr. Brisbin, "Bioaccumulation of cesium-137 in yellow bullhead catfish (Ameiurus natalis) inhabiting an abandoned nuclear reactor reservoir"’ Environmental Toxicology and Chemistry 16: 328-335, (1997).
11. F.W. Whicker, J. E. III Pinder, J.W. Bowling, J. J. Alberts, I. L. Jr. Brisbin, "Distribution of long-lived radionuclides in an abandoned reactor cooling reservoir", Ecological Monographs 60:471-496, (1990).
12. H.J. Mohler, F.W. Whicker, T.G. Hinton, "Temporal trends of ¹³⁷Cs in an abandoned reactor cooling reservoir", Journal of Environmental Radioactivity 37:251-268, (1997).
13. J.E. Brittain, A. Storruste, E. Larsen, "Radiocesium in brown trout (Salmo trutta) from a subalpine lake ecosystem after the Chernobyl reactor accident", Journal of Environmental Radioactivity 14:181-191, (1991).
14. I.I. Kryshev, I.N. Ryabov, T.G. Sazykina, "Using a bank of predatory fish samples for bioindication of radioactive contamination of aquatic food chains in the area affected by the Chernobyl accident", The Science of the Total Environment 139/140:279-285; 1993.