Environ. Sci. Technol. 1999, 33, 49-54
Temporal Change in Fallout 137Cs in Terrestrial and Aquatic Systems: A Whole Ecosystem Approach J A M E S T . S M I T H , * ,† SERGUEI V. FESENKO,‡ BRENDA J. HOWARD,§ A. DAVID HORRILL,§ NATALYA I. SANZHAROVA,‡ ROUDOLF M. ALEXAKHIN,‡ DAVID G. ELDER,† AND CHRISTOPHER NAYLOR§ Institute of Freshwater Ecology, East Stoke, Wareham, Dorset, BH15 2DH, U.K., Russian Institute of Agricultural Radiology and Agroecology, Kievskoe Street, 249020, Obninsk, Russia, and Institute of Terrestrial Ecology, Grange over Sands, Cumbria, LA11 6JU, U.K.
During the years after a nuclear accident, the bioavailability and environmental mobility of radiocesium declines markedly, resulting in large changes in contamination of foodstuffs, vegetation, and surface waters. Predicting such changes is crucial to the determination of potential doses to affected populations and therefore to the implementation of radiological countermeasures. We have analyzed 77 data sets of radiocesium (137Cs) activity concentrations in milk, vegetation, and surface waters after the Chernobyl accident. Our results show that the rate of decline in 137Cs during the years after Chernobyl is remarkably consistent in all three ecosystem components, having a mean effective half-life, Teff ≈ 2 years. By comparing changes in 137Cs availability with rates of diffusion of 40K (a close analogue) into the lattice of an illitic clay (1) we have, for the first time, directly linked changes in the environmental availability of 137Cs to fixation processes at a mechanistic level. These changes are consistent with declines in the exchangeable fraction of 137Cs in soils (2, 3).
Introduction Laboratory studies on the sorption of radiocesium to soils and sediments have shown that Cs is selectively bound to specific sorption sites (“Frayed Edge Sites”, FES) on illitic clay minerals (4). On these sorption sites, radiocesium is available for ion-exchange with ions which have a similar hydrated radius, specifically potassium and ammonium (4, 5). Over time, however, radiocesium slowly diffuses into the illite lattice (6) becoming unavailable for direct ion-exchange, a process commonly known as “fixation”. Clearly, rates of radiocesium fixation in soils and sediments are crucial to its long-term mobility and bioavailability, and a number of laboratory studies have been carried out to determine fixation rates (6-10). Most such experiments, however, are carried out over a period of months at most * Corresponding author phone: +44 (0)1929 462314; fax: +44 (0)1929 462180; e-mail:
[email protected]. † Institute of Freshwater Ecology. ‡ Russian Institute of Agricultural Radiology and Agroecology. § Institute of Terrestrial Ecology. 10.1021/es980670t CCC: $18.00 Published on Web 11/20/1998
1998 American Chemical Society
FIGURE 1. Schematic diagram indicating time scales, τ, of transfers of radiocesium from soils to terrestrial and aquatic ecosystems during the years after a fallout event. The time scale of “fixation” in soils is significantly longer than rates of retention and release of activity in the other parts of the ecosystem. Thus, in the long term, changes in the soil-soil water partitioning of activity controls changes in activity in surface waters, vegetation, etc. and cannot therefore reliably determine the temporal changes on a time scale of years which are required for long-term planning. Time Scales of Processes Determining Changes in Radiocesium Contamination of the Ecosystem. As shown in Figure 1, radiocesium transfers through a number of different ecosystem components are controlled by processes which operate on widely different time scales. Following a fallout event, some of these processes will achieve equilibrium more rapidly than others. For example, the rate at which radiocesium activity in vegetation is transferred to milk has a time scale of days. If the activity concentration in vegetation is constant or changing on a time scale . days, then the activity concentration in milk will rapidly reach equilibrium with respect to the activity concentration in vegetation. Thus, the activity concentration in milk will be a constant multiple of the (slowly changing) activity concentration in vegetation (often described by a “concentration factor”, CF: the equilibrium ratio of activity concentration in milk to that in vegetation). These concepts may be understood in terms of the firstorder differential equations which describe transfers of activity between the different ecosystem components. Let the activity concentration in vegetation, Cv, for example, decline exponentially with rate constant rv, so that Cv ) Cv(0)e-rvt. The uptake and removal of activity from milk may be described by uptake and removal rate constants, rmu and rmr, the activity concentration in milk, Cm, being given by
dCm ) rmuCv - rmrCm dt
(1)
Equation 1, with Cv ) Cv(0)e-rvt, has solution
Cm )
rmuCv(0)
(e-rvt - e-rmrt)
(rmr - rv)
(2)
If the time scale (τ ) 1/r) of changes in vegetation activity is significantly greater than the time scale of uptake and removal of activity from milk, then, as time increases, the activity concentration in milk will change at the same rate VOL. 33, NO. 1, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
49
FIGURE 2. Illustration of changes in radiocesium in milk in a system with slowly declining activity concentrations in vegetation (rv ) 0.001 d-1) and relatively rapid rates of uptake (rmu ) 0.2 d-1) and removal (rmr ) 0.02 d-1) from milk. Parameter values are for illustrative purposes only. as the activity concentration in vegetation (i.e. exp(-rmrt) f 0 therefore Cm ∝ exp(-rvt)). This is illustrated graphically in Figure 2. Similar principles have been shown to apply to radiocesium activity concentrations in lake waters after the Chernobyl accident (11). Fallout of radiocesium to a lake leads to an initial “spike” of activity to the lake water by direct deposition to the lake surface and “fast” transfers of activity in runoff water before it is sorbed to catchment soils. This “spike” is removed from the lake water by flushing and transfers to bottom sediments. The time scale for this removal process is, for most lakes, of order weeks to months, though for large lakes, it may be years or more (12). On longer time scales than those on which this removal takes place, lake water activity concentrations are determined by secondary contamination by runoff from catchment soils and remobilization from bottom sediments. Let the rate of removal of activity from the lake water by flushing and transfers to sediments be k1 (d-1). The longterm transfer of activity from the catchment soils to the lake, per square meter of lake surface, may be described by ic(t) (Bq.m-2) where ic(t) ) ic(0) exp(-k2t): k2 represents a slow decline in mobility of 137Cs in catchment soils (11). The activity concentration of radiocesium in the lake, CL (Bq.m-3) (ignoring physical decay), is given by
dCL ic(0)e-k2t ) - k lC L dt δ
(3)
where δ (m) is the lake mean depth. Equation 3 has solution
CL ) Ae-k1t + Be-k2t
(4)
where A and B are constants representing the initial lake concentration from direct fallout and initial catchment runoff contribution, respectively. A more complete derivation of these equations may be found in (11). Equation 4 represents a “double exponential” decline in lake water radiocesium activity concentrations. As illustrated in Figure 3, if k1 . k2, then at times . 1/k1, the radiocesium activity concentrations decline at rate k2. In this study, we confine ourselves to those lakes in which secondary contamination is dominated by transfers from the catchment and to those lakes in which the removal rate is great enough that, during our chosen 50
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 1, 1999
study period, the lake water radiocesium activity concentration is determined by the secondary contamination process (i.e. for times . the removal residence time of the particular lake). The Controlling Influence of Soil-Solution Partitioning of Radiocesium in Long-Term Contamination of the Ecosystem. During the weeks to months after a fallout event, radiocesium activity concentrations in both vegetation and surface waters are determined by short-term processes. Activity concentrations in plants are determined by interception and washoff rates of the initial fallout as well as uptake by the roots (13). Similarly, in rivers and lakes, activity concentrations are initially high as a result of direct deposition to the water surface and rapid runoff of 137Cs before it is sorbed to catchment soils (11, 14). Activity concentrations then decline over a period of weeks to months as a result of reduced runoff from catchments and, for lakes, loss of 137Cs through the outflow and deposition to bottom sediments (11, 12). On long time scales (years), however, the processes which determine radiocesium transfers to and from many different ecosystem components (for example, between plants and animals) are fast in comparison with the slow decline in radiocesium availability in soil (Figure 1). Thus, we hypothesize that the change in radiocesium activity concentration in the main environmental compartments should be controlled by slow changes in its soil-soil solution partitioning. To test this hypothesis, we have analyzed many longterm field studies of temporal changes in radiocesium in three different ecosystem components: vegetation, surface waters (dissolved phase), and milk following the Chernobyl accident (examples shown in Figure 4). We have then linked these studies to measured changes in soil chemical availability after the Chernobyl accident and to rates of diffusion of ions into the illite lattice. Long term rates of change in the radionuclide content of vegetation and biota are commonly described by assuming that the decline in radioactivity concentration, C, is exponential:
C ) C(0)e-λt
(5)
The rate of decline, λ (y-1) is often quoted as an effective ecological half-life, Teff (years) where Teff ) ln(2)/λ. λ may be identified with rv in eq 2 and with k2 in eqs 3 and 4.
Methods A literature survey was carried out of rates of long-term decline in radiocesium activity concentrations in surface waters, vegetation, and milk after the Chernobyl accident. As shown in Figure 4, when plotted on a logarithmic scale, the activity concentration data is linear as a function of time confirming the validity of eq 5 for characterizing these data over the period of our study. Values of λ include a radioactive decay component, though for 137Cs (decay constant: 0.023 y-1) this is negligible over our study period. Where possible, we have re-analyzed original data to obtain values of λ, though in 13 cases quoted λ (or Teff) values were used. Since the rate of decline, λ, in radiocesium decreases over time after a fallout event (2,18) (i.e. Teff increases over time, so the exponential model only applies for a limited period), we have confined our study to a five year period after contamination (after Chernobyl, this means autumn 19861991, though in some cases measurements in 1992 and1993 were included). As discussed above, data from the early period after the accident (spring-summer 1986) is influenced by short-term processes, and so vegetation and milk data from the first few months after the accident was ignored. Similarly, river water data from the first weeks (“fast”
FIGURE 3. Change in dissolved phase 137Cs activity concentration in Ennerdale Water after Chernobyl (adapted from ref 11, with permission from the Health Physics Society). The plot shows the “double exponential” decline in 137Cs concentration, CL. The dotted line indicates the second (long-term) component of the decline, the slope of which, k2, is the subject of the present study. in lake waters. Thus measurements of rates of removal of the initial fallout from lakes (e.g. ref 12) were not relevant to our study. We have further restricted the study to areas contaminated by aerosol deposition, excluding the “near zone” of Chernobyl where the breakdown of fuel particles can significantly affect temporal changes in radionuclide availability (2).
Results and Discussion
FIGURE 4. Examples of changes of 137Cs activity in different ecosystem components after Chernobyl. Data are from refs 11, 15-17. catchment loss phase, ref 14) after the accident was ignored, as was data covering the first phase of activity concentrations
Decline in 137Cs Activity Concentrations in the Environment. Rates of decline in radiocesium activity concentrations are found to be remarkably consistent over this period (examples are shown in Figure 4), with mean values λ ) 0.46 ( S.D. 0.18 y-1 (n ) 30) for vegetation (grasses, including hay, straw and grains), λ ) 0.38 ( S.D. 0.17 y-1 (n ) 23) for lakes (n ) 9) and rivers (n ) 14) (dissolved activity) and λ ) 0.39 ( S.D. 0.12 y-1 (n ) 24) for milk. It is known (11) that, in at least seven of the nine lakes we have studied, long term 137Cs activity concentration in water is determined by runoff from the catchment soils. At one site (Peat Ranker 1 of ref 19), no change in 137Cs in vegetation was observed over our study period. This was partly due to the influence of large seasonal variation in activity measurements: over a longer period (1987-1996) a slow decline (λ ) 0.18 y-1) was observed (16). Combining results for all three ecosystem components gives λ ) 0.41 ( S.D. 0.16 y-1 (n ) 77), corresponding to a mean Teff ) 1.7 years, with 91% of all measurements falling within the range 1-4 years. Figure 5 shows a histogram of measured Teff values in the three different ecosystem components: vegetation, surface waters, and milk. One source of variation in rates of decline is error in the estimation of λ. Typically, standard errors in the fitted value of λ were (25%, though in 10 cases single measurements from only three different years were available, leading to errors of around (50%. Environmental variation in rates of decline of 137Cs availability may be due to a number of factors including soil type and mineralogy, seasonal effects, agricultural practices, and application of countermeasures (2, 30). A number of authors (2, 19, 28) have suggested that the rate of decline in 137Cs in vegetation is related to soil type, in particular that rates of decline are higher in vegetation grown on soils containing significant clay fractions. The data we have assembled covers a number of different soil types (sandy soils, loams, clays, and peats), from upland, meadow, VOL. 33, NO. 1, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
51
FIGURE 5. Frequency distribution of effective ecological half-lives in different ecosystem components during the first five years after Chernobyl. Surface water data was obtained from refs 11, 14, 17, 20-22; vegetation data from refs 2, 13, 16, 19, 24-28; and milk data from refs 15, 23, 29.
, 5
and agricultural ecosystems as well as two lysimeter experiments (25, 27). Where possible, we have divided soil types into “organic” and “mineral” and compared λ values for vegetation grown on these two different categories of soil. The mean value of λ for organic soils (0.42 ( S.E. 0.064 y-1) was lower than that for mineral soils (0.51 ( S.E. 0.046 y-1), but the difference was not statistically significant at the 95% confidence level (t-test). Using measurements from 11 upland sites in Cumbria U.K. (13, 19), where there was good information on soil type and organic matter content, we have compared rates of decline in radiocesium activity concentrations, λ, in vegetation with organic matter content of the soil. To be consistent, λ values were calculated for the time period 1987-end 1989, giving a common sampling period for all 11 sites. For these sites, with a range of organic matter content of 19-88%, we observed a significant (r ) 0.8, p < 0.001) negative correlation between λ and the organic matter content. Rates of decline in 137Cs activity concentrations were twice as high in vegetation grown on soils with organic matter content 80% (mean λ ) 0.29 ( S.E. 0.13 y-1, n ) 5). It should be noted, however, that the highly organic soils showed very high variability in rates of decline (S.D. ) 0.28). Clearly, there is natural variation in λ values which is linked to soil characteristics. However, to our knowledge, there is currently no systematic way of predicting such variation. Nevertheless, the range in rates of decline we have observed (Teff ) 1-4 y) is relatively small: a value of Teff ) 2 y will adequately describe most cases. The narrow range in λ (Teff) values and the excellent agreement between values estimated for different environmental compartments is consistent with the controlling influence of soil-solution partitioning on changing 137Cs availability. Time Changes in 137Cs Exchangeability. The link between soil radiocesium availability and the effective ecological halflife is further demonstrated by direct measurements of changes in the chemical exchangeability of radiocesium over time. The exchangeably sorbed radiocesium is commonly measured by extraction with 1 M ammonium acetate solution (7). In studies on a hydromorphous and an automorphous soil, Fesenko and co-workers (2) showed that rates of decline in 137Cs in vegetation of λ ) 0.2 and 0.54 y-1, respectively, were similar to rates of decline in the exchangeable fraction 52
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 1, 1999
,
FIGURE 6. Diffusion of 40K into two differently prepared samples of a pure illite, calculated from results given in (1).
TABLE 1. Summary of Mean Values of Rate of Decline in 137 Cs Activity Concentrations in Different Environmental Compartments and Comparison with Rate of Diffusion of 40K into the Illite Lattice
description
mean rate of decline, λ (y-1)
mean effective half-life (y)
rivers and lakes (n ) 23) vegetation (n ) 30) milk (n ) 24) exchangeable 137Cs (n ) 3) 40K sorption to illite (n ) 2)
0.38 0.46 0.39 0.35 0.55
1.8 1.5 1.8 2.0 1.3
of λ ) 0.18 and 0.64 y-1, respectively. Experiments carried out by Shand and co-workers (3) on a peaty podzol also showed a slow decline in the exchangeably sorbed fraction, from 25% to 10% between 1 month and 4 years after the time of contamination. Interpreting this as a first-order transfer of activity from exchangeable to fixed sites again gives a similar rate of decline (λ ) 0.22 y-1) to those observed in vegetation, milk, and surface waters. Fixation of 137Cs in Illite. It has been shown (4, 9) that in most soils (with the exception of soils with a very high organic matter content), 137Cs is specifically sorbed to FES on illitic clay minerals. It is further reported (6) that the long-term fixation of 137Cs is a result of diffusion into the illite lattice where it exchanges with interlattice K. To our knowledge, there have been no laboratory studies of cesium diffusion into illite on the time scale of years required for comparison with our results, but a 16 month study has been made of diffusion of 40K into an illite (1). Using data obtained in this study (1), we have calculated the rate of reduction of 40K activity in solution as a result of isotopic exchange with interlattice K in an illite. For two differently prepared samples, this reduction can be modeled by an exponential function with rate constants λ ) 0.60 and 0.50 y-1 (Figure 6). The results show good quantitative agreement with our observed rates of reduction of environmental 137Cs activity concentrations (see Table 1), particularly for the mean value estimated for vegetation grown on mineral soils (λ ) 0.51 y-1).
The measurements of de Haan and co-workers (1) show good agreement with a 15 day study of Cs diffusion into a K-saturated illite (6). The value of λ ) 1.39 ( 0.5 y-1 obtained in this latter study is significantly higher than we have observed over our much longer time period but is comparable with values we have estimated for the early period (first 60 days) of the 40K diffusion experiment (1) (λ ) 2.66 ( 0.64 and 0.92 ( 0.29 y-1). Although we have observed an exponential decline in 137Cs availability over time (eq 5) during the period of our study, it appears from these results (1, 6) that fixation proceeds more rapidly over shorter time periods after contamination. In addition, a study on the mobility and availability of both Chernobyl and weapons-test derived 137Cs in bottom sediments of two lakes (31) has shown that the fixation process is not irreversible, tending to an equilibrium on a time scale of around a decade. This implies that the rate of fixation will decrease during the years after contamination and is consistent with observations (2, 16, 18) that the decline in 137Cs activity concentrations in vegetation slowed during the decade after Chernobyl. On longer time scales than that covered by our study, changes in 137Cs activity concentrations in the environment may tend toward steady-state, declining only as a result of physical decay with a half-life of 30.2 years. For simplicity and consistency with other studies, we have used a simple exponential decline (eq 5) to characterize our data, but we note that the applicability of this function is dependent on our (necessarily restricted) study period (ca. 0.5-5 years after contamination). Our results should not be extrapolated significantly beyond this period of time after contamination. Previous work (9) has shown that in soils of low to medium organic matter content (80%) that the majority is associated with regular ion-exchange sites. These results indicate that low illite contents in highly organic soils will lead to a slower rate of reduction (i.e. lower λ) in 137Cs availability, in agreement with our observations. They are also consistent with our observations that in the majority of soils, rates of fixation are similar. As long as there is a significant illite mineral component, the majority of the 137Cs will be sorbed to FES (9). Thus, rates of fixation will be controlled by the same diffusion rate into the illite lattice, regardless of soil type. In “mineral” soils (those in which the majority of the radiocesium is on the FES) soils which are relatively high in FES capacity would still be expected to have a higher solidssolution partitioning of radiocesium (Kd) than those of lower FES capacity, but the rates of fixation would be expected to be the same. Thus, we would expect to see differences in levels of radiocesium uptake by plants, and releases to runoff water between these soil types, but little difference in changes over time. Potential Changes in 137Cs in the Absence of Slow Fixation. Two other processes have been hypothesised which may reduce radiocesium activity concentrations following a fallout event: (1) the reduction in inventory of radiocesium as a result of losses of activity from soils in runoff water and (2) transport of activity down the soil profile. We will test the ability of each of these processes to explain our results. If there is no slow fixation of radiocesium, the activity concentration in soil water would be a constant fraction of that sorbed to the soil. Thus, the activity concentration in runoff water would decline only as a function of declining catchment 137Cs inventory. In this case, the rate of change, λ, in 137Cs activity concentrations in runoff water would be given by λ ) fc, where fc is the fractional loss of 137Cs from the catchment per year (y-1). Using measured values of radiocesium loss from typical catchments of ca. 0.5-2% per year (11, 32) gives values of λ ) 0.005-0.02 y-1 (Teff ) 139-35
y) which is more than 1 order of magnitude slower than the rate of decline we have observed. It is important not to confuse the rates of change in activity concentrations we have been studying with rates of removal of activity concentration from the soil. The parameter fc has also been measured by, for example, Linsley and co-workers (33), who found even lower removal rates (fc) than those observed by (11, 32). Over the period of our study, the decline in inventory of catchment soils as a result of removal in runoff water is negligible. Thus the rates of change in activity concentrations in surface waters (Teff ∼ 2 years) cannot be due to loss of inventory but must be due to a reduction in radiocesium mobility. In the absence of slow fixation, the change in activity concentration in vegetation would decline only as a result of declining activity in the rooting zone. The vertical transport of radionuclides in soils may be modeled by the advectiondispersion equation
∂C ∂2C ∂C )D 2-ν ∂t ∂x ∂x
(6)
where D (cm2‚y-1) is the dispersion coefficient ν (cm‚y-1) is the rate of advection and x is depth from the soil surface. For an impermeable upper boundary (i.e. diffusion can only take place downward from the soil surface), eq 6 was solved for a decay-corrected tracer, input as a short term “spike” at the soil surface (34-36). Using typical parameters (36) for advection (ν ) 0.8 cm‚y-1) and dispersion (D ) 0.45 cm2‚y-1), we have calculated the transport of 137Cs through the soil profile for a period of 5 years after Chernobyl. After this period, 82% of the 137Cs remained in the surface 5 cm of soil and 100% was within the surface 10 cm, implying that transport out of the rooting zone is not significant in comparison with the rates of decline in vegetation activity concentrations we have observed. We have shown that neither vertical migration of activity nor transfers of activity in runoff water result in significant declines in 137Cs activity concentrations in comparison with those we have observed. We have, for the first time, quantitatively linked diffusion of radiocesium into the illite lattice to changes in its long-term environmental mobility and bioavailability. We have shown that this diffusive “fixation” of 137Cs had a controlling influence over rates of change in 137Cs contamination of terrestrial and aquatic ecosystems in the first few years after the Chernobyl accident.
Acknowledgments We would like to thank Nick Beresford, John Hilton, Mike Hornung, and Ed Tipping for constructive criticism of earlier versions of this manuscript. This work was supported by the CEC Nuclear Fission Safety Program (ECOPRAQ project, F14P-CT95-0018), the CEC Inco-Copernicus program (Reclaim project ERBIC15CT960209), and by the U.K. Natural Environment Research Council (CEH Integrating fund).
Literature Cited (1) Haan, F. A. M. de.; Bolt, G. H.; Pieters, B. G. M. Soil Sci. Soc. Am. Proc. 1965, 29, 528-530. (2) Fesenko, S. V.; Spiridonov, S. I.; Sanzharova, N. I.; Alexakhin, R. M. J. Environ. Radioactivity 1997, 34, 287-313. (3) Shand, C. A.; Cheshire, M. V.; Smith, S.; Vidal M.; Rauret, G. J. Environ. Radioactivity 1994, 23, 285-302. (4) Cremers, A.; Elsen, A.; De Preter, P.; Maes, A. Nature 1988, 335, 247-249. (5) Comans, R. N. J.; Middelburg, J. J.; Zonderhuis, J.; Woittiez, J. R. W.; De Lange, G. J.; Das, H. A.; Van Der Weijden, C. H. Nature 1989, 339, 367-369. (6) Comans, R. N. J.; Hockley, D. E. Geochim. Cosmochim. Acta 1992, 56, 1157-1164. VOL. 33, NO. 1, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
53
(7) Evans, D. W.; Alberts, J. J.; Clark, R. A. Geochim. Cosmochim. Acta 1983, 47, 1041-1049. (8) Robbins, J. A.; Lindner, G.; Pfeiffer, W.; Kleiner, J.; Stabel, H. H.; Frenzel, P. Geochim. Cosmochim. Acta 1992, 56, 2339-2361. (9) Valcke, E.; Cremers, A. Sci. Tot. Environ. 1994, 157, 275-283. (10) Absalom, J.; Young, S. D.; Crout, N. M. J. European J. Soil Sci. 1996, 46, 461-469. (11) Smith, J. T.; Leonard, D. R. P.; Hilton, J.; Appleby, P. G. Health Phys. 1997, 72, 880-892. (12) Santschi, P. H.; Bollhalder, S.; Zingy, S.; Luck, A.; Farrenkother, K. Environ. Sci. Technol. 1990, 24, 519-527. (13) Horrill, A. D.; Howard, D. M. J. Radiol. Prot. 1991, 11, 249-257. (14) Monte, L. J. Environ. Radioactivity 1995, 26, 71-82. (15) Mu ¨ ck, K. Health Phys. 1997, 72, 659-673. (16) Pomeroy, I. R.; Nicholson, K. W.; Branson, J. R. Monitoring Chernobyl radiocaesium in Cumbrian grass, 1992-1995. AEA Technology report AEAT/18142002/REMA-197, AEA Abingdon, U.K., 1996. (17) Vakulovsky, S. M.; Nikitin, A. I.; Chumichev, V. B.; Katrich, I. Yu.; Voitsekhovitch, O. A.; Medinets, V. I.; Pisarev, V. V.; Bovkum, L. A.; Khersonsky, E. S.. J. Environ. Radioactivity 1994, 23, 103122. (18) Ehlken, S.; Kirchner G. J. Environ. Radioactivity 1996, 33, 147182. (19) Sandalls, J.; Bennett, L. J. Environ. Radioactivity 1992, 16, 147165. (20) Håkanson, L.; Brittain, J. E.; Monte, L.; Bergstro¨m, U.; Heling, R. J. Environ. Radioactivity 1996, 33, 1-26. (21) Nyle´n, T. Uptake, turnover and transport of radiocaesium in boreal and forest ecosystems. FOA-R-96-00242-4.3-SE. National Defense Research Establishment, Umea˚, Sweden, 1996. (22) Saxe´n, R. In The transfer of radionuclides through Nordic Ecosystems to Man; Dahlgaard, E. H., Ed.; Elsevier: Amsterdam, 1996. (23) Alexakhin, R. M.; Firsakova, S.; Rauret, G.; Arkhipov, N.; Vandecasteele, C. M.; Ivanov, Y.; Fesenko, S. V.; Sanzharova, N.
54
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 1, 1999
(24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36)
I. In The radiological consequences of the Chernobyl accident; Karaoglou, A., Eds.; Desmet, G., Kelly, G. N., Menzel, H. G., Eds.; EUR 16544 EN, European Commission, Luxembourg, 1996. Belli, M.; Tikhomirov, F. Behaviour of radionuclides in natural and semi-natural environments. EUR 16531 EN, European Commission, Luxembourg, 1996. Krouglov, S. V.; Filipas, A. S.; Alexakhin, R. M.; Arkhipov N. P. J. Environ. Radioactivity 1996, 34, 267-286. Noordijk, H.; van Bergeijk, K. E.; Lembrechts, J.; Frissel, M. J. J. Environ. Radioactivity 1992, 15, 277-286. Pietrzak-Flis, Z.; Krajewski, P.; Krajewska, G.; Sunderland, N. R. Sci. Tot. Environ. 1994, 141, 147-153. Squire, H. M.; Middleton L. J. Radiation Botany 1966, 6, 413423. Karle´n, G.; Johanson, K. J.; Bertilsson, J. J. Environ. Radioactivity 1995, 28, 1-15. Sanzharova, N. I.; Fesenko, S. V.; Alexakhin, R. M.; Anisimov, V. S.; Kuznetsov, V. K.; Chernyayeva, L. G. Sci. Tot. Environ. 1994, 154, 9-22. Smith J. T.; Comans R. N. J. Geochim. Cosmochim. Acta 1996, 60, 995-1004. Helton, J. C.; Muller, A. B.; Bayer, A. Health Physics 1985, 48, 757-771. Linsley, G. S.; Haywood, S. M.; Dionan, J. IAEA Report 1982 IAEA-SM-257/8, 615-633. Crank, J. C. The Mathematics of Diffusion, 2nd ed.; Clarendon Press: Oxford, 1975. Kudelsky, A. V.; Smith, J. T.; Ovsiannikova, S. V.; Hilton, J. Sci. Tot. Environ. 1996, 188, 101-113. Smith, J. T.; Elder, D. G. European J. Soil Sci. 1998, Submitted.
Received for review July 6, 1998. Revised manuscript received September 30, 1998. Accepted October 7, 1998. ES980670T
