Because of the many different sources of radioactivity to the Great Lakes, their patterns of release and the actions of various environmental processes, the geographical distribution of radionuclides in the Great Lakes shows considerable irregularity and non-uniformity. The presence of a specific radioactive isotope in one of the Lakes does not assure its presence in the other Lakes or connecting channels. Even within a lake, the distribution is not uniform (does not suggest a totally mixed lake) but often shows a stratification with a different activity in the nearshore region compared with the larger open lake region. Thus, some comments about geographical distribution of radionuclides are essential.
Theoretically, all radioisotopes found in atmospheric fallout, regardless of origin (cosmogenic or as the result of past weapons testing or accidents such as at Chernobyl), should appear in the waters and sediments of all of the Great Lakes and connecting channels. Only a very few long-lived isotopes, however, are detectable either in dry or wet precipitation to the Basin. Also, these isotopes may not be detected in all of the air, water, or sediment samples from all of the Great Lakes. Since rates of atmospheric deposition depend mainly on the surface area for the deposition, Lake Superior, with the greatest surface area of the Great Lakes would receive the highest load of isotopes from deposition. However, Lake Superior has no nuclear power plant facilities discharging to it, suggesting that the only isotopes expected from atmospheric deposition would be those associated with past nuclear weapons testing or originate from sources a considerable distance away from the Great Lakes and have been subject to long-range atmospheric transport.
The Task Force used the work of the Commission's International Air Quality Advisory Board to examine the typical atmospheric transport of materials from various locations within North America. Prevailing wind patterns to the Great Lakes from North American sources would suggest the possibility that the Hanford facility (Richland, Washington) and Idaho Falls facility of the Department of Energy might be suitable sources of atmospheric transport of radioactive materials to the Great Lakes. However, the monitoring reports of both facilities, which include high-altitude samples, did not show any detectable radionuclides known to be discharged in air emissions from those facilities reaching the Great Lakes within 5 days, which is sufficient time to transport more than 90% of the materials with half-lives greater than 5 days.
The Nuclear Task Force has collected and examined environmental monitoring data provided by the operators of the major nuclear facilities in the Great Lakes Basin. Virtually all of the reported radionuclides had activities or concentrations that were reported as the lower limit of detection (LLD). This does not necessarily mean that the various radionuclides were absent from the environment, nor that their environmental impacts were insignificant. It simply means that the radionuclides could not be detected by the instrumentation and procedures used.
The LLD for a measurement depends on several factors including sensitivity of the instrumentation, length of time that radioactivity from the sample is counted, and elapsed time between sample collection and counting, (a shorter elapsed time means fewer losses due to radioactive decay). Facility operators must often seek a compromise between the need for a low level of detectability, the costs of the instrumentation, and the need for a large throughput of samples in a limited amount of time. Operating parameters are often chosen for the purpose of demonstrating compliance with government regulations rather than characterizing the movement of radionuclides in the environment.
Rather than attempting to reproduce the results from all of the station reports, we have chosen three typical examples to illustrate the general significance and the limitations of these results. Table 17 shows the results for artificial radionuclides in shoreline sediments, fish, and surface waters near the Nine Mile Point Nuclear Station (Oswego County, New York) during 1994. 137 Cs was the only artificial radionuclide detected in sediments and fish, and tritium ( 3 H) was the only one detected in surface water. For the fish and water, there were no significant differences in the measured levels between the indicator locations (column 4) and the control locations (column 6). This would indicate that the measured levels were not due to emissions from the local facilities. Only the 137 Cs in sediment appeared to be elevated at the indicator locations, suggesting a contribution from the power station.
Tables 18 and 19 show environmental monitoring results from the Donald G. Cook Nuclear Plant operated by the Indiana Michigan Power Company. In Table 18, 131 I levels in air were above the limit of detection for a brief period from February 21 to March 14, 1994. These were directly attributed to station emissions. Table 19 shows that, apart from the naturally occurring 40 K (potassium-40), the only radionuclide detected in fish during 1994 was 137 Cs. These levels are typical of global fallout and do not appear to have been enhanced by station emissions.
In Tables 17-19, the original units of picocuries (pCi) have been retained to illustrate how lower limits of detection are reported (1 pCi = 0.037 Bq).
Table 20 shows levels of radionuclides in Lake Ontario fish in the vicinity of the Pickering Nuclear Generating Station during one year (1988). The level of naturally occurring potassium-40 in fish and other biota is regulated by a homeostatic mechanism and is unaffected by inputs from human activities. The 14 C and 137 Cs detected in the fish are probably due to residual fallout from the earlier testing of nuclear weapons. Levels of 250-350 Bq of 14 C per kg carbon were typical of background values in 1988. One to 2 Bq/kg of 137 Cs are quite typical of fish taken from lakes across Canada (Elliott et al. 1981).
Discharges and runoff from the West Valley waste storage site enter Buttermilk Creek, which discharges into Cattaraugus Creek, which then discharges into Lake Erie. Cattaraugus Creek has received continuous radiological monitoring since 1968. Small amounts of radioactivity are detected in the creek, and water samples collected below the discharge area of the facility contain measurable levels of radioactivity. Data through 1985 were previously reported by the Commission (1987 b ). More recent data through 1993 were graciously supplied by Mr. William Condon of the New York State Department of Health's Bureau of Environmental Protection (N.Y. Department of Health 1983-1993). Table 21 summarizes data on the monitoring of Cattaraugus Creek.
A simple material balance for a radionuclide in a lake has three source terms (atmospheric fallout, tributary inflow, and direct discharge of effluents to a lake), two storage terms (water column and sediments), and an export term (lake outflow). A sixth term, also an export term, the revolatilization of gaseous radionuclides from a lake surface, applies to a very few radionuclides, mainly those of the noble gases ( e.g. , 41 Ar, 133 Xe, 85 Kr, 222 Rn), 14 C in radioactive methane and carbon dioxide, and tritium in water. While tritium moves through the environment via the hydrological cycle, volatilization from the Lakes does not appear in any of the inventory calculations available for this radionuclide.
By considering the Great Lakes as a system of five lakes and several connecting channels, material balances are possible at two levels of scale: an individual lake, and the "whole" system of lakes and connecting channels. The hydrological parameters of the Great Lakes are shown in Table 22. The material balances share many terms, and the system has several simplifying features. The absence of nuclear power plants on Lake Superior or its tributaries means the input of radioactivity to this Lake comes mainly from atmospheric deposition. The outflows from Lake Superior become inflows to Lakes Michigan and Huron, and the outflows from these lakes become inflows to connecting channels to Lake Erie, whose outflow is Lake Ontario's inflow, and Lake Ontario's outflow goes to the St. Lawrence River.
In a relatively uniform (completely mixed) lake, the product of the lake volume and its average radionuclide concentration or activity estimates the inventory term for lake water storage of the radionuclide. Many inventory calculations separate the inventories of water column and sediments but fail to report the sediments. See the section, Sediments , below for further discussion of sediment inventories.
The inventory of radionuclides in the water column is often difficult, because radionuclides are difficult to detect unless very large samples of water (300 L) are taken. The major exceptions are tritium and isotopes of strontium and rubidium, which because of water solubility of most of their inorganic compounds, remain in the water column preferentially over the sediments.
The long history of monitoring of 90 Sr and 137 Cs (each with half-lives of about 30 years) is the basis for calculation of the deposition inventories of many other radionuclides. The calculation depends on using geochemical ratios with latitude band adjustments and begins with choosing a given year and noting the depositional flux for 90 Sr or 137 Cs from the worldwide fallout data for latitude bands of 30-40N and 40-50N (which cover the area of the Great Lakes). The longest fallout record is for 90 Sr data with 137 Cs fluxes estimated using a ratio of the activities of 137 Cs/ 90 Sr of 1.6. This ratio describes a relatively constant distribution of the two radionuclides in many environmental media and compartments.
For other radionuclides, the activity ratio of a radionuclide to either the strontium or cesium in fallout is used. Data from the atmospheric testing of nuclear weapons during the period of 1954-1963 provided usable ratios. Wahlgren et al. (1980) used 0.176 for the activity ratio 239, 240 Pu/ 90 Sr to estimate the atmospheric deposition of plutonium to the Great Lakes. Separate data sets exist for 210 Pb, which has been intensively studied as an atmospheric tracer. This isotope has both natural and artificial sources and, therefore, was not studied by the Task Force.
Joshi (1991) has noted that, except for 3 H, 90 Sr, 137 Cs, and 239,240 Pu, data for very few radionuclides have been reported for open lakes. Some open lake data are available for uranium, radium, and thorium in Lakes Huron, Michigan, and Ontario, but most inventory calculations require geochemical correlation and estimation methods. Both Joshi and the GLWQB (IJC 1983, 1987 b ) have noted that radionuclide data reported for drinking water intakes often do not differ statistically from open lake cruise data, and that averaging the data on a radionuclide from drinking water intakes at different locations in a lake may provide a reasonable estimate of the open lake or even whole-lake average radionuclide concentrations.
Although the Task Force has examined a significant quantity of environmental data collected by nuclear facility operators, the analysis of open water data was constrained by the limited number of lakewide monitoring surveys conducted in the past, making environmental and biological assessments difficult to perform. The Surveillance Work Group of the Commission's Water Quality Board (IJC 1987 b ) has previously recommended that radionuclide monitoring be conducted in the open waters of the Great Lakes every 5 years in a manner similar to the surveys conducted by the National Water Research Institute, Environment Canada, between 1973 and 1983. The last open water surveillance program was initiated in 1990 by Environment Canada, Conservation and Protection, but was limited to Lake Ontario. Its scope was to ensure that nuclear facilities and other sources of radioactive contamination were controlled in a manner that met the broad objectives on the Lake Ontario Toxics Management Plan and the Great Lakes Water Quality Agreement (IJC 1987 a ).
Table 23 gives the average concentrations in open lake water of the three major radionuclides ( 3 H, 90 Sr, and 137 Cs), together with their estimated inventories in the water column. Also shown are the estimated inputs to 1983 from fallout and to 1993 from reactors. Virtually all of the fallout deposition was complete by 1983. No attempt has been made here to carry out a mass balance. It must be recognized that large quantities of radionuclides have been lost to the water column through outflow and sedimentation.
Radionuclide levels measured during the 1990 Lake Ontario survey are similar to those from the 1983 survey. 3 H in open waters ranged from about 9.1 to 10.8 Bq/L (average: 10.1 Bq/L); concentrations at sites near the inflow to Lake Ontario averaged about 6.8 Bq/L, and a level of 9.2 Bq/L was measured at the outflow (Environment Canada, unpublished data). More recent sampling of 3 H levels in Lake Ontario between 1991-1993 showed average concentrations between 9 and 11 Bq/L, with a projected yearly increase due to routine CANDU operations of about 0.12 Bq/L (Chant et al. 1993). Other radionuclides measured in Lake Ontario open waters in 1990 include uranium (0.270 µg/L), 226 Ra (0.002 Bq/L), and 210 Pb (0.12 Bq/L). Levels of 90 Sr in fish ranged from <0.01-0.059 Bq/g in flesh and 0.018-0.059 Bq/g in bone. 137 Cesium in fish flesh ranged from <0.01 to 0.064 Bq/g; all levels measured in bone were <0.01 Bq/g.
The inventory for plutonium is a model for further calculations as it is based on actual field data with only the atmospheric deposition term estimated from 90 Sr. Table 24 presents the inventory of plutonium from the work of Wahlgren et al. (1980). The Task Force considers that inventory one of the most effective uses of field data, extrapolation, and modeling methods of any of the radionuclide inventories in the literature.
Field studies of plutonium rarely distinguish between the two isotopes, 239 Pu and 240 Pu, but rather tend to report plutonium as a sum of all isotopes. Some limited data suggest a typical isotope distribution of 60% 239 Pu, 30% 240 Pu, and the remaining 10% divided among short-lived plutonium isotopes and decay chain radionuclides, but that does not seem to compromise the usefulness of the data.
As noted, most radionuclides entering the Great Lakes move to sediments as their final repositories. The separation of radionuclide fractions between water column and sediments depends on a distribution parameter, K D (defined as the ratio of the activity of a radionuclide per unit weight of sediment to the activity of that same radionuclide per unit weight of the bulk water phase or water column), which quantifies the distribution of activities of a radionuclide in a bulk sediment and bulk water column. Edgington has pointed out to the Task Force that a K D of less than 1,000 often means that there is a detectable fraction of the activity of a radionuclide in the water column, a K D between 1,000 and 10,000 often means that the distribution of radioactivity between sediment and water column is unpredictable, and K D of greater than 10,000 assures that the radionuclide is almost totally incorporated into the sediment (greater than 99.99%). Although the parameter, K D , is empirical, its use has some theoretical basis in surface chemistry. Values of K D depend on sediment properties: consolidation, pore water, strength of bottom currents that cause scour, the chemical environment ( i.e. , alkalinity, pH, Eh, ionic strength, etc.), and the biological activity of benthic organisms. In undisturbed sediments, radionuclides that bind to particulates or bottom materials tend to remain in the upper few centimeters of surficial sediments.
One of the most difficult components of an inventory is the assessment of the radionuclide content of the biota. Organisms are continuously exposed to radiation and radioactivity, but the extent to which they act as repositories for radioactive isotopes of various elements involves a complex set of metabolic and physiological processes, which have not been intensively studied for purposes of establishing an inventory. Most of the research entails the use of radioactive versions of selected elements or compounds (tracers), which are important in the physiological functioning of various species in order to understand the pathways and mechanisms of those physiological processes and functions. Almost none of the studies extended the data from tracer studies to establish biological compartmental inventories of radionuclides.
This section addresses individual elements and radionuclides with respect to bioaccumulation and biomagnification factors for freshwater biota. Bioaccumulation refers to the retention in a biological compartment of material from the external environment or a non-biological or non-living source material to an extent or in an amount that exceeds on a relative weight basis the presence of the material in the source. Biomagnification refers to the retention in a biological compartment of a material that originated from another biological compartment in an amount that exceeds on a relative weight basis its presence in the originating biological compartment. Thus, bioaccumulation applies to the uptake of materials from water and sediments as non-living source materials, and biomagnification applies to the uptake of materials from living source materials, as in predator-prey interactions. The quantification of bioaccumulation and biomagnification is through biological transfer factors, which are defined for radionuclides in a manner identical to the definition of the K D of sediments:
a ratio of activity of a given radionuclide on per unit weight basis for biological tissue to non-living source material, and a ratio of activity of a given radionuclide on per unit weight basis for a biological repository to that of its biological (living) source material .
The work emphasizes studies with stable radionuclides, but some data derived from radionuclides appear, mainly radionuclides of cesium and potassium. Table 25 lists those radionuclides in increasing order of atomic weight and number, which because of their half-lives, could be of possible interest with respect to the limnological cycling of elements through the biota of the Great Lakes. Several radionuclides appear as "combinations," notably 140 Ba/ 140 La, 95 Zr/ 95 Nb, and 99 Mo/ 99 Tc. Some reports from nuclear power plants and other dischargers, as well some research and monitoring reports, present the given combinations as single radionuclide entries. These sources report the total activity of the combination (a sum of the activities of the two radionuclides), but either do not or cannot disaggregate the data and assign the separate activities to each radionuclide in the pair. The second radionuclide in each pair is a decay product of the first, and the pair has assumed a special status in the radionuclide literature. However, most of those elements even when released to the water column do not enter biota. Therefore, only a small subset of the elements of the Table 25 are discussed individually.
Because of its importance to the Great Lakes, tritium is discussed separately in this section. For radionuclides other than tritium, the Task Force has prepared Table 27 bioaccumulation and biomagnification factors. These factors are based mainly on data for the stable isotopes of the elements, and where possible, from studies on the radioactive isotopes of the elements. All data were from studies of organisms found in lakes and especially the Great Lakes. Among the early available collection of freshwater data, Cowgill's work from 1973 to 1980 in Linsley Pond North Brantford, Connecticut, although not on the Great Lakes, provides one of the largest and most systematic studies of the uptake of chemical elements in plants and some animal species. Cowgill studied all detectable elements in the plants, animals, waters, and landscape materials of her study area (Cowgill 1970, 1973 a , 1973 b , 1974 a , 1974 b , 1976; Hutchinson 1975). She reported data as total elemental concentration (stable plus unstable) with no consideration as to isotopes. She compared her data with marine and oceanic studies because of a shortage of other data sets on elemental accumulations in plants and animals. Following her successful work on Linsley Pond, she turned her attention to similar studies in many aquatic systems ranging from laboratory cultures to entire marshlands in locations all over the world (Cowgill and Prance 1982; Cowgill et al 1986). The Task Force gratefully acknowledges her guidance in assessing the relationships among elements in aquatic species.
The Task Force also had access to a classic data set of Copeland et al (1973) for fish in the vicinity of nuclear power plants on Lake Michigan, and reporting on all of the elements that were quantifiable. Many reports cite this special compendium of Lake Michigan data, including the National Council on Radiation Protection (NCRP, #76 and #126), but often through secondary sources.
As previously noted, tritium is not routinely monitored in biological compartments. An early study, Rosenthal and Stewart (1971), and an important research project of the International Atomic Energy Agency (IAEA) with three studies, Blaylock and Frank (1979), Kirchmann et al. (1979), and Adams et al. (1979) provide the basic information.
Rosenthal and Stewart (1971) examined two algal species, an aquatic macrophyte, a Daphnia , and three species of freshwater snails in a small pond. Their work has two important characteristics: first, that the tritium which bound to biological tissues ("tissue bound tritium" or TBT) never exceeded 7% of the tritium taken up by the biota, and second, that the species studied occur in the Great Lakes Basin. The experiments ran a sufficiently long time relative to the life cycles of the biota to assure that the results could likely represent a steady-state bioaccumulation level. Their data also show that the levels of TBT may not show species selectivity between plants and animals, although plants seem to accumulate slightly more tritium than animals.
Rosenthal's work shows the need to understand the "tritium terminology." Tritium moves environmentally mainly as a tritiated water molecule, HTO. Discussions of biota tend to emphasize the different forms of HTO in organisms. Text Box 1 gives a short lexicon.
From the Text Box, the combustion HTO is a form of organically bound tritium. (This form of tritiated water appears to exert most of the adverse toxicological effects on biota.) Organisms can rapidly excrete unbound HTO and mobilize physiological systems to excrete less rapidly bound HTO. However, organisms retain organically bound HTO indefinitely and lose tritium by radioactive decay.
Blaylock and Frank (1979) examined some plants and animals from a pond on the Oak Ridge Reserve (Tennessee). Kirchmann et al. (1979) added several freshwater and marine species to those previously studied, including a salmonid, and some information on tritium apportionment in cellular compartments. Adams et al. (1979) reported on tritium uptake by Great Lakes aquatic biota in a freshwater marsh system in Lake Erie, near the Davis-Besse nuclear power plant. These studies used mutually agreed upon protocols and methods that assured that their data were statistically compatible (could be pooled statistically, because they come from the same "universe of data"); reported tritium bound to tissues on a dry weight basis, providing consistency and continuity with previous data of early investigators on bioaccumulation; and identified clearly the species studied taxonomically and tissues and substructures examined. Tables 26a-d present the data of Blaylock and Adams (1979), Rosenthal and Stewart (1971), Kirchmann et al. (1979), and Adams et al. (1979). The Task Force recalculated the bioaccumulation factors for Blaylock and Adams from the original paper. The Task Force only presents the data from North American species from Kirchmann's study.
Tritium levels in organisms track the tritium levels of the environment. Bioaccumulation factors of unity for tritium are conservative.
Tritium uptake in aquatic biota usually tracks the environmental levels. Some investigators note that tritium in organismal tissues quickly "equilibrate" with the tritium content of the surrounding water, but that view is simplistic. The data show that biota can excrete unbound tritium as HTO rather rapidly, and bound tritium as HTO more slowly, but may not be able to excrete tritium as organically bound HTO. Thus the view of "equilibration" is simplistic since tritium targets DNA and RNA, and through isotope exchange, tritium replaces a stable hydrogen on a nucleotide as organically bound HTO in the cell nucleus or extranuclear DNA in cytoplasm or organelles such as plasmids and chloroplast. "Equilibration" masks the isotope exchange on biomolecules because the exchange process occurs after uptake and at a much slower time scale.
From the available information, the Task Force affirms a bioaccumulation factor of unity (1) is conservative and confirms previous recommendations from marine studies.
The retention of 10% of the tissue-bound tritium level in organisms suggests a method of estimating an inventory of tissue-bound tritium in biota.
The previous information suggests a method to estimate a tritium inventory for aquatic organisms: as the product of the biomass and 10% of the environmental level of tritium. However, such an inventory calculation has a practical limitation in that most environmental tritium data for waters are reported at or below a level of detection, variably 100-200 pCi/L, depending on the instrumentation. Thus, the method only provides an upper bound estimate of the biological inventory of tritium.
The univalent elements sodium, potassium, lithium, rubidium, cesium, silver, chlorine, bromine, and iodine are found in freshwater lakes and their biota. The Task Force was especially interested in potassium, rubidium, cesium, and iodine.
The isotope 40 K is a major natural source of radioactivity. It accounts for as much as 40% of the radioactivity present in biological tissues (humans included). Potassium is a macro element with important physiological functions, and therefore, its presence exerts an important control function on the cycling of other elements. Measurements of 40 K in Great Lakes biota appear in several documents including the annual reports of various nuclear power plants. Since geological data do not clearly show a selective or differential enrichment of potassium isotopes in geological repositories ( i.e. , glaciers, mineralized soils, and volcanic rock), a simple inventory estimate for 40 K depends on the assumption that the ratio of radioactive potassium to stable potassium in biological compartments and substrates is numerically the same as the ratio of radioactive potassium to stable potassium in geological strata.
Since freshwaters are impoverished in sodium (by definition), isotopes of sodium although limnologically important are not discussed. The naturally occurring radioactive isotopes of sodium have not been reported on in biota. An inventory procedure similar for the one 40 K does not apply to radioactive isotopes of sodium.
Rubidium accumulates in plants often with the ability to substitute for potassium and sodium in selected chemical matrices. Isotopes of rubidium are emitted directly by nuclear power plants to the Great Lakes as well as forming as decay products of noble gas radionuclides
Although cesium chemistry parallels sodium and potassium chemistries, cesium does not easily replace either sodium or potassium in various chemical matrices because its atomic and ionic radii are larger than those of sodium and potassium. The element's rarity in nature also limits its biological availability.
The most extensive data base for bioaccumulation of an artificial radionuclide exists for 137 Cs. This long-lived isotope (half-life: 30 years) was the first long-lived isotope in nuclear fallout studied in biota as part of biological monitoring and the study of the effect of radio isotopes on biota. The important compilations on radiocesium uptake are the studies of Blaylock (1982), Joshi (1984), and Elliott et al. (1984). The latter two references discuss Great Lakes fishes. Most of the compilations from these authors were from research components of monitoring studies. Further, dischargers of radionuclide materials in the Great Lakes Basin must monitor for 137 Cs in biological substrates, although the data have deficiencies that limit its use for biocompartment inventory calculations.
The Task Force has also examined monitoring data from nuclear power plants for 137 Cs in fishes. Tables 28 and 29 present data for the bioaccumulation factors for 137 Cs for Lakes Huron, Erie, and Ontario. All of the bioaccumulation factors in Table 29 have the same order of magnitude, although fishes in Lake Huron have higher bioaccumulation factors than fishes in Lake Ontario. These factors should not be assumed to hold for the two Lakes over all periods of time. Joshi (1984) calculated bioaccumulation factors from several fish studies going back to 1976. The data in Table 29 show that the ratio of 137 Cs to 40 K is about 0.01. This suggests a way to calculate an inventory for 137 Cs. From the estimate of the inventory for 40 K, estimate the inventory for 137 Cs by adjusting the estimate for 40 K by 0.01.
Given the small bioaccumulation rates for radionuclides of cesium, the final repository for radiocesium from fallout and liquid and gaseous discharges to the Great Lakes is sediments. Although cesium salts are relatively soluble in aquatic media, the attachment of the cesium compounds to other particulates provides a mechanism to reach sediments with minimal physicochemical interaction with aquatic media.
One radionuclide of silver, 110m Ag (half-life: 233 days), consistently appears in the nuclear discharges to the Great Lakes. The radionuclide has been reported every year in the gaseous and liquid discharges of at least eight nuclear power plants in the Great Lakes over the period of 1980-1993. It persists long enough for silver to cycle through Great Lakes biota following its discharge. However, there is no procedure derivable from the data with which the Task Force could estimate and inventory of this isotope in biological compartments.
There are no radionuclides of fluorine that need to be discussed. Radionuclides of chlorine are mainly cosmological, although 36 Cl sometimes appears as a fission product. The radionuclides of bromine are fission products and are associated with fuel rods and would not be expected to be discharged routinely to the Great Lakes.
Freshwater and marine biota both accumulate iodine. The environmental levels of iodine are greatest in the oceans, and marine plants, invertebrates and vertebrates accumulate it to levels greater than freshwater or terrestrial organisms. The low levels of iodine in freshwater limit its uptake almost uniquely to plants and vertebrates, the latter biota having thyroid glands. There are very limited data on the uptake of iodine by freshwater invertebrates. Bioaccumulation factors for iodine of about 500 for plants and biomagnification factors of about 0.05 for zooplankton and insects seem appropriate.
The divalent elements beryllium, magnesium, calcium, strontium, and barium are all found in the waters, sediment, and biota of the Great Lakes. The studies of two investigators, Cowgill (1973 a , 1973 b , 1974 a , 1974 b ) and Yan et al. (1989) are most important.
Cowgill's data are the most comprehensive for beryllium. Because beryllium isotopes produced cosmogenically remain in the atmosphere for a long time, the appearance of radioactive beryllium in Great Lakes biota suggests that the beryllium probably originates from artificial sources, mainly reactions in a nuclear reactor. 7 Be has been occasionally documented in gaseous and liquid effluents of nuclear reactors, but as a very minor radionuclide.
The radionuclides of calcium and magnesium are too short lived to consider biological inventories. However, the Task Force notes that chemical data for both elements are needed in discussing the behaviour of the radionuclides of other elements cycling within the Great Lakes, especially radionuclides of divalent elements in the same or adjacent families of the Periodic Table: beryllium and barium. Calcium also correlates with many other elements: titanium, phosphorus, iron, cerium, lanthanum, and most other rare earth elements. The correlation suggests that it exerts a control function on the cycling of many other elements in freshwater biota.
Strontium has several important radionuclides. The element accumulates in plants, can biomagnify up the food chain, and thus presents a health hazard to humans through intake of food. Isotopes of strontium can be found in the water column because many strontium compounds are water soluble. Although strontium has bone as its target organ in vertebrates, the Task Force could not confirm that calcium has a control influence on either strontium uptake or biological cycling.
140 Ba is a fission product of some importance. It accumulates in biota, but the available data on barium levels in biota do not permit a separate calculation of the radioactive barium content. The Task Force did find that levels of barium in biota are correlated with levels of calcium and that high environmental levels of calcium and high tissue levels of calcium both block uptake and accumulation of tissue levels of barium.
Zinc, manganese, iron, technetium, ruthenium, chromium, nickel, cobalt, and molybdenum belong to different chemical groups, but it is often easier to address them together. Their radionuclides are mainly activation products, but they are also produced by fission. Because their nuclear reactions involve inter-conversions, their methods of analysis usually provide data on all them virtually simultaneously. All of these elements accumulate in biota, but the pathways for technetium are unknown. That element has not been studied in aquatic biota despite the fact that its major ion, pertechnate (TcO 4 ) 1- is water soluble and allows potential direct uptake by biota from the water column.
The important radionuclide of zinc is 65 Zn (half-life: 244 days), which can cycle within the Great Lakes. Zinc is an essential trace element in nutrition, assuring its bioaccumulation in all species. From the available date on zinc levels in Great Lakes biota, the Task Force could not calculate a separate inventory for this radioisotope.
The important radionuclide of chromium is 51 Cr. Chromium cycles through biological compartments in several valence states, two of which are very important in aquatic systems. The hexavalent (+6) state is water soluble and very toxic to most organisms. The trivalent (+3) state has a low water solubility and behaves as a trace micronutrient in certain tissue and organismal systems (Mertz 1967; see also other papers by Walter Mertz). Most environmental studies report a total chromium level without specification of the valence states. From the available date on chromium levels in Great Lakes biota, the Task Force could not calculate a separate inventory for this radioisotope.
Molybdenum is a trace element nutrient needed by plants. It occurs in several enzymes associated with nitrogen fixation and the utilization of iron and sulfur in cellular metabolism. Two radionuclides, 95 Mo and 99 Mo, form as fission and activation products. Although 99 Mo has a half -life of only 6 hours, it decays to 99 Tc (half-life: 212,000 years) and thus plays an important role in estimating inventories for technetium. 99 Mo and 99 Tc are sometimes treated as a combined pair. From the available date on molybdenum levels in Great Lakes biota, the Task Force could not calculate a separate inventory for this radioisotope.
The important radionuclides are 55 Fe (half-life: 2.6 years), 59 Fe (half-life: 45 days), 54 Mn (half-life: 303 days), and 56 Mn (half-life: 2.6 hours). The two radionuclides of iron and lower atomic weight radionuclide of manganese last long enough to cycle through Great Lakes ecosystems.
Depending on the pH of the freshwater system or the "local pH" (the acidity of surface materials) and the level of oxygenation of the system (aerobic versus anaerobic environment), iron and manganese hydroxides can form precipitates on the external surfaces of biological materials. Many analyses of plants that report very high iron and manganese levels may actually have reported crystalline ferric and manganese oxides as surface contaminants. Despite these chemical artifacts, both elements have major metabolic roles in organisms and approach a status of macronutrient rather than simple trace nutritional requirement. From the available date on iron and manganese levels in Great Lakes biota, the Task Force could not calculate separate inventories for the listed radioisotopes.
Without nuclear technology, ruthenium would be of little interest to the Task Force. Two radionuclides are produced by fission and appear mainly in nuclear fuel processing operations: 103 Ru (half-life: 41 days) and 106 Ru (half-life: 1 year). The radionuclides can appear in both gaseous and liquid effluents. They were detected in the atmospheric fallout to the Great Lakes after the accident at Chernobyl, but not from the liquid discharges of radionuclides from nuclear power operations. Both radionuclides persist long enough for possible cycling within the biota of the Great Lakes, but the element has rarely been detected analytically in freshwater biota and reported in the literature. It has no known biological role, and despite its persistence, the available studies suggest that it moves mostly to sediments. The Task Force has not estimated a biocompartment inventory for ruthenium given its lack of observation in freshwater biota.
All radionuclides of cobalt are activation products. Four radionuclides, 57 Co, 58 Co, 59 Co, and 60 Co, are discharged to the Great Lakes, but two of them, 60 Co (half-life: 5.26 years) and 57 Co (half-life:270 days), last long enough to cycle biologically. The former has commercial use, and the latter is used in research.
Cobalt is an essential micronutrient that activates vitamin B 12 . Therefore, cobalt uptake occurs in all aquatic biota. Some algae, macrophytes, and invertebrates can substitute cobalt for zinc in essential enzyme systems (Price and Morel 1990; see also other papers of François Morel), suggesting that Zn:Co geochemical ratios may be important in certain water bodies. From the available data on cobalt levels in Great Lakes biota, the Task Force could not calculate a separate inventories for cobalt radioisotopes.
Except for polonium, a decay product of transuranic elements, none of the Group VIb elements require inventories. The Task Force has previously discussed the cosmogenically produced radionuclides of sulfur. All radionuclides of selenium produced by nuclear activities in the Great Lakes Region except for 79 Se are very short lived, but the residuals of 79 Se are important only in considerations of the high-level waste inventories for fuel elements. The Task Force has no documentation that selenium radionuclides are released by nuclear facilities to the Great Lakes Basin. Despite the toxicity of tellurium, biota only show limited accumulation. Nor are radionuclides of tellurium documented in the releases from nuclear power plants, although such radionuclides would require consideration in the high-level waste inventories for fuel elements. Several isotopes of tellurium form as fission products. The volatility of many tellurium compounds is the factor explaining their appearance in the gaseous radionuclide emissions of fuel reprocessing operations.
The trivalent elements include boron, aluminum, gallium, indium, and thallium. Boron and aluminium bioaccumulate in organisms, but there is no indication that they are discharged to the Great Lakes. Their congener elements of gallium, indium, and thallium do not have radionuclides that are produced in nuclear systems that are likely to be discharged to the Great Lakes.
The Group IIIa elements of the Periodic Table are rather unusual. All are rare in nature. Scandium often shows up in particulate matter sampled in the upper atmosphere, and has two radionuclides which form cosmogenically. But while scandium can bioaccumulate, there is no indication that its radionuclides are related to either natural background levels or radioactive discharges to the Great Lakes.
Yttrium, lanthanum, and the other rare earth elements, are well represented among the radionuclides formed in the nuclear fuel cycle. Many of these elements have primordial radionuclides, and almost all of the isotopes of rare earth elements are mildly radioactive. The behaviour of these elements in biota is not well understood. Cowgill's studies suggested that organisms may exert considerable selectivity on which elements they accumulate: besides cerium and lanthanum, and occasionally europium, the freshwater plants seem to favour the elements of even atomic number. They also accumulate in biota greatly compared with source materials, which often have levels undetectable by present methods.
Yttrium : Radionuclides of yttrium are fission products and the decay products of radioactive strontium. Two radionuclides of yttrium occur in discharges to the Great Lakes, 90 Y and 91 Y. The former has a very short half life, but the latter has a sufficiently long half-life to be of interest. Yttrium can accumulate in organisms, but the available data base is sparse; Cowgill's data are the most comprehensive. Yttrium's environmental cycling appears to follow its congener elements.
Lanthanum and Cerium : Lanthanum, although very toxic to aquatic biota, is detected in small amounts in aquatic biota along with other rare earth elements. Lanthanum phosphate, the compound expected in most fresh waters, is very insoluble, limiting its bioavailability. The decay of 140 Ba to 140 La makes lanthanum of biological interest.
Cerium is the only other rare earth element besides lanthanum that is usually detected in biota. Cowgill's data probably form the most complete set on the stable forms of the element in freshwater biota. Cerium isotopes except 138 Ce, which is cosmogenic, are fission products and have been detected in Great Lakes waters. The two major radionuclides, 141 Ce and 144 Ce, usually appear in the aerosol content of gaseous emissions from nuclear power plants and occasionally in the liquid emissions. In nuclear fuel reprocessing operations, these radionuclides can appear equally likely in both gaseous and liquid emissions. Both isotopes were detected in the Great Lakes waters following the incident at Chernobyl.
Calcium appears to control the biological uptake of lanthanum and cerium as well as other lanthanides. The mechanism is unclear. It may block uptake directly by a mechanism similar to the one for barium or indirectly because of the relationship with phosphorus. Where uptake occurs, the Task Force has no indication that radionuclides of other rare earth elements behave differently with respect to calcium, and thus further discussion of individual lanthanides does not appear warranted.
The Task Force derived bioaccumulation factors of 20,000 for both lanthanum and cerium in freshwater plants; biomagnification factors were 2.4 in zooplankton and insects for lanthanum and 1.0 in zooplankton and insects for cerium.
The quadrivalent elements are carbon, silicon, germanium, tin, and lead. Despite their common chemical grouping, their chemistry differs very markedly from element to element.
Carbon has already been discussed as 14 C. Silicon bioaccumulates in some species as an essential element in the skeletal structure of diatoms and foraminifera, usually in the amorphous mineralogical form of opaline phytoliths. There is no indication that radioisotopes of silicon are emitted to the Great Lakes. The two cosmogenically produced radionuclides of silicon, 31 Si and 32 Si, have unknown impact on the Great Lakes. Germanium isotopes can result from fission, but the discharge of germanium to the Great Lakes is largely undocumented.
Radionuclides of lead result from the decay of transuranics. The long life of 210 Pb and the concern about lead as a toxicant and air pollutant have led researchers to study this element and this particular radionuclide intensively. Thus, the information bases on radionuclide and stable lead are rather large, but the biological data on the radionuclide form of lead is not very extensive.
Tin is the most difficult quadrivalent element to consider. Very few data exist on its occurrence in aquatic systems and even less in biological compartments, Cowgill's data being the most comprehensive. Analytical methods to quantify tin in environmental media and substrates call for considerable skill and instrumental sophistication. The important radionuclides are 113 Sn (half-life: 115 days) and 117m Sn (half-life: 14 days), and 126 Sn (half-life: 100,000 years). The first two radionuclides originate by fission and activation processes, the third by activation processes only. The first two have been occasionally documented in the Great Lakes, while third one has not, although its appearance would signal trouble. The activation sources are the zircalloy (a zirconium-tin alloy) cladding for nuclear fuel elements, making them important radionuclides from the nuclear fuel cycle. From the available data on tin levels in Great Lakes biota, the Task Force could not calculate a separate inventories for tin radioisotopes.
The quinquevalent elements include arsenic, phosphorus, antimony, and bismuth. Transition quinquevalent elements are vanadium, niobium, and tantalum. Quadrivalent transition elements, specifically titanium, zirconium, and hafnium have biological characteristics that follow the quinquevalent elements. Data for biological uptake of antimony, bismuth, and tantalum are very rare or non-existent. Titanium and hafnium accumulate in plants, but there is no indication that their radionuclides are discharged to the Great Lakes. Titanium data assist in establishing inventories for many other elements including phosphorus, calcium, and the lanthanides.
Of all of the elements of the Periodic Table, phosphorus holds a special place in the limnology of the Great Lakes. Starting in 1972, the removal of phosphorus from point sources and its quantification and management in non-point sources and sediments has guided programs to control eutrophication (the proliferation of aquatic plants which are stimulated by excess or luxuriant levels of nutrients, mainly phosphorus) of the Great Lakes. The chemistry and biology of phosphorus in the Great Lakes is rather complex subject to understand, and the control of eutrophication in the Great Lakes, while highly successful, has not achieved results sufficient to reduce the current control programs.
Phosphorus has two radioisotopes, both of which have cosmogenic as well as other sources. Although the isotopes have half-lives of the order of a few days to a few weeks, the chemical dynamics of phosphorus make all sources of phosphorus a special concern. Further, the early work on understanding the dynamics of phosphorus chemistry and biology in lakes used both of these isotopes as tracers.
In the section on sources of radioactivity, the data presented on secondary sources ( e.g. , hospitals, university laboratories and reactors, commercial uses) indicated widespread use and discharge of both radioisotopes of phosphorus. For these reasons, phosphorus has received special attention in the Inventory of Radionuclides.
Zirconium and niobium are associated with the cladding material of fuel elements in nuclear power plants. Their radionuclides are both neutron activation and fission products. Radionuclides of zirconium, mainly 95 Zr, occurred in the fallout from atmospheric testing of nuclear weapons prior to 1963. Radionuclides of zirconium decay to radionuclides of niobium, and the combination 95 Zr/ 95 Nb sometimes appears as a single radionuclide in the emission reports of nuclear facilities. Both 95 Zr (half-life: 65 days) and 95 Nb (half-life: 35 days) are of potential environmental concern. The radionuclide, 94 Nb (half-life: 20,000 years), is one of the longest lived, but there are no indications that it is either discharged directly to the Great Lakes region or forms as a decay product of another radionuclide discharged to the Great Lakes. It has a very small fission yield and is included among the radionuclide inventories for spent nuclear fuel.
Both elements bioaccumulate. Cowgill's data are the most extensive and suggest niobium accumulates in plants more than zirconium. From the available data on zirconium and niobium levels in Great Lakes biota, the Task Force could not calculate separate inventories for their radioisotopes.
Vanadium accumulates. It is an essential micronutrient for certain plants and fungi. 50 V is a primordial radionuclide, and its inventory in biological tissue can be estimated by a similar procedure used for 40 K. Other radionuclides of vanadium form as both fission and activation products. Recent interest in vanadium by a number of researchers has provided a data base on vanadium uptake in plants and animals that offers possibilities in producing inventories for the biological compartments. However, there is only limited indication that vanadium radionuclides are released to the Great Lakes, and the only inventory that can be calculated for a radioactive isotope of vanadium is for its primordial radionuclide 50 V by multiplying the vanadium level in biota by 0.0025.