lable at ScienceDirect

Journal of Environmental Radioactivity 146 (2015) 27e34

Contents lists avai

Journal of Environmental Radioactivity

journal homepage: www.elsevier .com/locate/ jenvrad

Estimation of vertical migration velocity of 137Cs in the MountIDA/Kazdagi, Turkey€Ozlem Karadeniz a, *, Rukiye Çakır b, Hidayet Karakurt c

a Department of Physics, Faculty of Sciences, Dokuz Eylül University, 35160 Tınaztepe, _Izmir, Turkeyb Department of Medical Physics, Institute of Health Sciences, Dokuz Eylül University, 35340 _Inciraltı, _Izmir, Turkeyc South-eastern Anatolian Forestry Research Institute, 23049 Elazı�g, Turkey

a r t i c l e i n f o

Article history:Received 29 January 2015Received in revised form27 March 2015Accepted 27 March 2015Available online

Keywords:Cs-137Vertical migrationMean annual velocityExternal gamma-dose rateForest

* Corresponding author. Tel.: þ90 232 3018675; faxE-mail addresses: ozlem.karadeniz@deu.edu.tr (€O.

deu.edu.tr (R. Çakır), hkarakurt@yahoo.com (H. Karak

http://dx.doi.org/10.1016/j.jenvrad.2015.03.0320265-931X/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

This paper presents the results obtained from a radioecological study carried out in the forest sites ofMount IDA (Kazdagi)/Edremit, Turkey. For 118 soil profiles, the depth distribution of 137Cs activity wasestablished by fitting the experimental points to an exponential, a gaussian or a log-normal function. Therelaxation lengths were in the range of 1.09e16.7 cm with a mean of 5.73 cm, showing a slow transportand a strong retention capacity of 137Cs even after the 26-y period of Chernobyl accident. From the datafor the vertical distribution of 137Cs in soil profiles, the mean annual migration velocity of 137Cs was in therange of 0.11e0.62 cm year�1 with a mean of 0.30 cm year�1. Statistically significant correlations betweenthe thickness of the humus layer and the mean annual velocity of 137Cs were found for both coniferousand mixed forest sites. The mean annual velocity of 137Cs in the forests sites with Pinus nigra var pal-lasiana was significantly higher than sites with Pinus brutia. External dose-rates from the 137Cs in forestsoils were estimated using a conversion factor used in many studies and comprised with the externaldose-rates determined according to the vertical distribution of 137Cs within the soil depth profiles. It isclearly seen that both levels and spatial distribution patterns of the external dose-rates from 137Cs wereinfluenced considerably with the vertical migration rate and the vertical distribution of 137Cs.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

From the radiological point of view, the major fission productradionuclide of concern in radiation dose is 137Cs because of itsrelatively long half-life (30.17 y), its abundance in the fallout andhigh biological mobility. After the deposition of 137Cs onto the soil,the resulting external gamma-radiation may contribute signifi-cantly to the total exposure of the population living in thecontaminated area. As the 137Cs migrates slowly into deeper soillayers, part of its external gamma radiation is attenuated by theoverlying soil layers and the corresponding dose rate in air willchange with time. Thus, the external dose rate in air dependsessentially on the vertical distribution of the 137Cs activity in thesoil. Knowledge on the vertical migration of 137Cs is also importantin predicting the fate of present and future fallout, root uptake byplants and ground water contamination. In addition, because

: þ90 232 4534188.Karadeniz), rukiye.cakir@ogr.urt).

radiocaesium activities in soil exhibit great variations comparedwith natural radionuclides, it is concluded that contribution of 137Cshas to be taken into consideration during the estimation of gammaradiation dose rate (Dołha�nczuk-�Sr�odka, 2012). Therefore, verticalmigration of 137Cs and resulting external exposure in forests hasbeen subject to research for a long time (Bunzl et al., 1997; Almgrenand Isaksson, 2006; Clouvas et al., 2007; Karadeniz and Yaprak,2008a).

Recent determinations of the vertical distributions of 137Cs inundisturbed soils point out that radiocaesium is still distributedmainly through the upper 10 cm or in the humicrich layers of forestsoil and shows very little vertical migration (Rafferty et al., 2000;Bunzl et al., 2000; Livens et al., 1991; Ramzaev et al., 2006;Karadeniz and Yaprak, 2007a). Some hypotheses have been sug-gested to explain such a slow migration (Staunton et al., 2002;Nikolova et al., 2000), and there have been many studies thatdemonstrate the parameters affecting the behaviour of radio-caesium in soil (Lee et al., 1997; Cha et al., 2006; Kim et al., 1998;Zygmunt et al., 1998; Albers et al., 1998; Karadeniz and Yaprak,2008b). With reference to the above mentioned studies, accumu-lation and migration behaviour of radiocaesium in soil, particularly

€O. Karadeniz et al. / Journal of Environmental Radioactivity 146 (2015) 27e3428

of 137Cs, was greatly influenced by the physicochemical forms of Cs,texture, mineralogy, organic matter content, various exchangeablecations and type of the soil, biological activity of microorganisms insoil, the hydrological regime, meteorological circumstances (suchas precipitation, temperature or humidity) and the ecologicalconditions of the contaminated area. Accordingly, its migrationsand associated profile distributions differ from area to area anddepend strongly on the landscape concerned, with the relationsbetween the 137Cs content in soils and the different parametershaving to be determined at each location.

The external gamma-dose rate at 1 m height above a flat area for137Cs can be determined from its contamination density, thegamma ray absorption properties of the soil, and the observedvertical distribution of 137Cs within the soil depth profiles. Thesedepth profiles are obtained by analysing the different sections ofsample cores. Simple models have often been successfully used toquantify the downward migration of radionuclides. For example,the negative exponential depth function (Baeza et al., 1993), loga-rithmic-polynomial equation (Barisic et al., 1999) or that radionu-clide transport is controlled by a diffusion-like mechanism(Schuller et al., 2004). The vertical distribution of 137Cs was thendescribed by variations of a general exponential function by fittingthe function to the field data.

Studies are required on the vertical distribution of radiocaesiumin the forest ecosystem to assess the total radiation exposure ofpeople from the contaminated areas. Therefore, a radioecologicalstudy was carried out at the forest sites in Mount IDA (Kazdagi)/Edremit that were contaminated by deposition after the Chernobylaccident. The purposes of this article are to examine the depth-related distributions of 137Cs, to determine the vertical migrationof 137Cs and to establish the external dose-rates from 137Cs in forestsoils.

2. Materials and methods

2.1. Sample collection and processing

A map of the studied area was presented in our previous study(Karadeniz et al., 2015). As the highest mountain on the Bigapeninsula, it was most likely to intercept contaminant plumes fromChernobyl.

The sampling was done according to the characterisation ofdistinct soil horizons and approximately 3e4 kg of sample of theeach forest soil layers (OL, OF þ OH, A and other mineral horizons)were systematically taken separately in each grid from 150 rectan-gles, in 1 site per 4 km2 each in 2010e2012. The location of eachsample site was determined by global positioning system, GPSGarmin Model 12XL. The soil from each horizon was weighted andthen dried to a constant weight at 60 �C for 24 h in an electric oven,reweighed and sieved through a 2-mm sieve to eliminate impu-rities such as stones and roots. The loss of weight after drying (d.w.loss) was calculated for each soil samples. Each dried sample(200e1850 g) was placed in 1000 mL Marinelli beaker prior toanalysis.

2.2. Gamma-spectrometric measurements

The activity concentration of 137Cs in the soil samples weremeasured with a high resolution HPGe gamma-ray spectrometrysystem. A detailed description for detector characteristics, systemoperation and calibration has been presented elsewhere (Karadenizet al., 2015).

The sample containers were placed on detector endcap forcounting. The accumulating time of the sample spectra was rangedbetween 10 000 and 20 000 s to obtain a gamma spectrum with

good statistics. The activity concentration of 137Cs was estimatedusing single gamma-peaks of 661.6 keV. The statistical errors wereconsidered only for the counting statistical uncertainty, whichwerein the order of 1e3% for high activities and more than 10% for thesmall activities at the 95% level of confidence. The minimumdetectable activity (MDA) based on Currie (1968) for the countingtime of 20 000 s was 0.03 Bq kg�1 for 137Cs.

Caesium-137 concentrations per unit mass in Bq kg�1 dryweight (d.w.) and per square metre in kBq m�2 dry weight (d.w.)were determined in the soil samples. Deposition values of 137Cs(kBq m�2) were calculated as the product of the activity per unitmass (Bq kg�1) and the mass depth of each component (kg m�2).The mass depth (kg m�2) for each of the soil horizons was calcu-lated such that the soil density (kg m�3) multiplied by the depth,from the surface down to midpoint of each layer. The bulk density(kg m�3) of all soil samples was determined as the ratio of weightafter drying to fresh soil volume. All measured activities werecorrected for the radioactive decay to the sampling date.

2.3. Vertical distribution of 137Cs in soil

Knowing the vertical distribution is important for calculation ofthe penetration and original deposition of 137Cs and also estimationof the dose rate from 137Cs. Namely, these calculations are based onthe measured activity in each layer and the analytical expressionfor the depth distribution found in a fit. Many models to describethe vertical distribution of 137Cs in soil have been suggested. In thecase of fresh fallout in the first years after deposition, the depthdistribution of 137Cs in undisturbed soil is usually described as anexponential function of depth (Almgren and Isaksson, 2006;Takahashi et al., 2015), which can be written as follows:

CðzÞ ¼ a$e�b$z (1)

where C(z) is the activity concentration of the 137Cs at depth z; aand b are experimentally determined parameters and a the con-centration of the 137Cs as z approaches zero. As the parameter a isdetermined with assuming a uniform density for the whole soilprofile (Isaksson and Erlandsson, 1998), the depth can be expressedeither as linear depth z (cm or m) or as mass depth z$r (kg m�2),where r is the density of the soil. The reciprocal of b is referred to asthe relaxation length and can be described either as cm or kg m�2.In general, the relaxation length describes the shape of the tail ofthe depth distribution and it can be a good indicator for the esti-mation of the 137Cs long-term vertical distribution in the soil(Schuller et al., 2002). The depth given asmass depth, an alternativeexpression can be written as follows:

CðzÞ ¼ a$e�b0$z0 (2)

where z0 is the mass depth, determined from field samples,expressed in kg m�2 and hence b0 is expressed in m2 kg�1. Thereciprocal of b0 is referred to as the relaxation mass depth whichcan be described as kg m�2 and describes the shape of the initialdepth distribution of activity concentration of the 137Cs in the soil(Schuller et al., 2007). The greater is the value of the relaxationmass depth, the deeper the 137Cs penetrates into the soil profile(Kato et al., 2012).

After longer time periods of the initial deposition, this approachis not reasonable because of the subsequent vertical migration of137Cs. When the profile tends to have a maximum at another depththan at the surface, the source activity concentration as a functionof depth in soil for a Gaussian distribution is presented as follows by(Likar et al., 1998):

€O. Karadeniz et al. / Journal of Environmental Radioactivity 146 (2015) 27e34 29

AðzÞ ¼ A0$e�ðz�z0Þ2=2s2

(3)

where A0 is the specific activity of the top layer of soil in Bq kg�1, z0the mean Gaussian curve and s the dispersion.

Another type of the observed vertical distribution of 137Cs is log-normal, which can be given as follows:

AðzÞ ¼ A0$e�

�ln z

z0

�2

2s (4)

where A0 is the specific activity of the top layer of soil in Bq kg�1, z0and s log-normal distribution parameters.

In order to investigate the vertical migration, a good way seemsto be the evaluation of the vertical velocity, which is the shift, withtime, of the centre of the distribution in the radionuclide's con-centration along the soil profile (Arapis and Karandinos, 2004). Thevertical migration of 137Cs was investigated in the studied areaaccording to following formula:

x ¼Xi

xigi (5)

where: x is the weighted mean depth of vertical concentrationdistribution in centimetres, xi is the mean depth of the layer incentimetres, and gi is the percentage of 137Cs activity concentration(Bq kg�1) in the corresponding depth xi.

2.4. Estimation of the external gamma dose rate

One of the pathways by which radioactive isotopes prevailing inthe soil affect man is by external irradiation. The contribution ofradionuclides to the absorbed dose rate in air depends on theconcentrations of the radionuclides in the soil. Various approachesto estimation of the external gamma dose rate for 137Cs in soil canbe found in the literature (Saito et al., 2015; Saito and Petoussi-Henss, 2014; Petrovic et al., 2013; Mabit et al., 2012; Miller et al.,1990). To estimate the gamma external dose, the dose conversionfactor should be determined. In this paper, the absorbed dose ratesfor 137Cs were estimated using the conversion factor of0.03 nGy h�1/Bq kg�1 published by several authors (Jibiri et al.,2007; Nada et al., 2009; Dołha�nczuk-�Sr�odka, 2012). In addition,themethodology given by Likar et al. (1998) was used, which can beeasily used to calculate the dose rate for any distribution of 137Csencountered in practice. These calculations rely on the observedvertical distribution of 137Cs, themass depth of the each soil layer ateach site and the conversion factors for the plane sources atdifferent depths, making this approach both site and temporallyspecific.

2.5. Statistical analysis

All statistical evaluations were carried out with SPSS 13.0version. Statistical analyses for possible significant correlationsbetween the parameters were performed with nonparametricPearson correlation analysis. The frequency distribution of data setswas tested against a normal or lognormal distribution by the Kol-mogoroveSmirnov test (significance level p > 0.05).

3. Results and discussion

Based on 118 soil profiles, the results of the 137Cs measurementsin the forested areas at the Mount IDA (Kazdagi)/Edremit werepresented in our previous study (Karadeniz et al., 2015). In the

investigated soil profiles, about 27e99% (with a mean of 80%) of the137Cs total activity is contained in the first 10 cm of soil. The pres-ence of 137Cs in soils was derived from both the past atmosphericnuclear explosion tests and from the Chernobyl accident. However,during the measurements of the soil samples, 26 years after theChernobyl accident, the 137Cs cannot be separated by whether itsorigin is related to Chernobyl or global fallout.

The 137Cs distribution with depth was variable and threedifferent profiles can be identified (Fig.1). Namely, the distributionsof the 137Cs radionuclide can be classified as exponential, gaussianor log-normal. For 118 soil profiles, equation (1) and (3) or (4) wassatisfactorily fitted to the experimental values of the 137Cs activityconcentration and related parameters were found. If the modelcould not be successfully fitted to the measured data for a profile,then they were not use in the calculations. Fig. 1 shows the depthdistributions of 137Cs-activity (Bq kg�1) in the soil horizons of sixprofiles and lines indicate the fitted curve. The fits were made withthe depth expressed as linear depth (cm) as well as mass depth(kg m�2). To evaluate whether the depth dependence was equallywell described as linear depth in cm or mass depth in kg m�2, thecorrelation between the b and b

0values was investigated. Relatively

high correlation coefficient (rsp ¼ 0.925, a < 0.01) indicating thatwith this method of sampling, thicknesses of the samples by bothdefinitions of linear depth and mass depth were well-defined.

The relaxation lengths/relaxation mass depths were computedas the reciprocal of parameters b/b0 from fits with depth andexpressed in centimeter or kg m�2 in Equations (1) and (2). For thegaussian and log-normal distributions of 137Cs, parameters weredetermined with excluding the litter layer, as mentioned byTakahashi et al. (2015). The relaxation lengths were in the range of1.09 cme16.7 cm with a mean of 5.73 cm. These values agree withrelaxation lengths previously reported (Isaksson and Erlandsson,1998; Schuller et al., 2002; El-Reefy et al., 2006; Karadeniz andYaprak, 2008a). These findings showed a slow transport and astrong retention capacity of 137Cs even after a relatively long time.The relaxation mass depths were 21e250 kg m�2 with a mean of90 kgm�2. According to the International Commission on RadiationUnits and Measurements (ICRU), relaxation mass depth values forforests are expected to range from20 to 200 kgm�2 after more than10 years (ICRU, 1994). The averaged relaxation mass depth values atthe sites were consistent with the ICRU data. Some of the values arehigher than those reported by ICRU (1994), which may be attrib-uted to the ever deeper penetration of the 137Cs following 26-yperiod after Chernobyl accident.

Relaxation parameters of 137Cs are affected by factors such assoil properties, rainfall rates, the elapsed time since deposition andtype of ecosystem, as well as type of forest litter (Matsuda et al.,2015; Shcheglov et al., 2014; Koarashi et al., 2012; Kato et al.,2012). For example, relatively weak but significant correlationsbetween rainfall rates and relaxation lengths and a negative cor-relation between particle size and relaxation length were found byGraham and Simon (1996). Schuller et al. (2002) found that therelaxation length shows a tendency to increase with decreasingclay content and increasing soil coarse-pore volume. Koarashi et al.(2012) showed that relaxation length correlates positively withorganic C content and bulk density. Takahashi et al. (2015) obtainednegative correlation between relaxation length and 137Cs inter-ception potential (RIP), indicating that the penetration distance of137Cs in the very early stages decreased with an increasing RIP ofthe topsoil. In addition, penetration process of 137Cs relates withwater flow and soil structure such as rate of fine particle content,amount of dispersible particles. Downward migration of 137Cs alsoincreases with temperature because of microbial organic matterdecomposition. According to Hao et al. (2013), the strong retentionof 137Cs in surface soils is attributed to the presence of amorphous

€O. Karadeniz et al. / Journal of Environmental Radioactivity 146 (2015) 27e3430

mineral phases, organic matter and high CaeMg status in the soil.However, some investigations found that no evidence for correla-tion with any of the soil parameters and the relaxation length (El-Reefy et al., 2006; Isaksson and Erlandsson, 1998).

Mean depth values were calculated based on Equation (5) andranged from 2.75 cm to 15.5 cmwith a mean of 7.58 cm. The meanannual migration velocity of 137Cs was estimated from the meandepth values and time after the Chernobyl accident. Accordingly,the mean annual velocity of 137Cs was in the range of0.11e0.62 cm year�1 with a mean of 0.30 cm year�1. These valuesare similar with previous studies (Arapis and Karandinos, 2004;Almgren and Isaksson, 2006).

Fig. 1. Depth distribution of 137Cs in soil profi

It is suggested that migration from the humus layers into theunderlying soil might be influenced with the thickness of thehumus layer (Karadeniz et al., 2015 and references cited there in).Therefore, possible relation between the thickness of the humuslayer with the relaxation mass depths and the mean annual ve-locity of 137Cs were investigated for coniferous and mixed forestsites (mixed stand with coniferous and deciduous tree species).Analysis of the correlation between the thickness of the humuslayer and the relaxation mass depths of 137Cs is shown in Fig. 2. Thebest-fitting relation between the thickness of the humus layer withthe relaxation mass depths of 137Cs was of linear and positive typefor coniferous forest sites (Fig. 2a), with a statistically significant

les for IDA (54, 67, 80, 84, 102 and 103).

Table 2Summary statistics for the external dose-rates from the 137Cs of Mount IDA.

Da (nGy h�1) Db (nGy h�1)

Median 1.61 4.10Mean ± S.E. 1.95 ± 0.13 4.17 ± 0.21S.D. 1.46 2.25

Fig. 2. Relationship between the thickness of the humus layer and the relaxation mass depths of 137Cs for (a) coniferous and (b) mixed forest sites; between the thickness of thehumus layer and mean annual velocity of 137Cs for (c) coniferous and (d) mixed forest sites with linear regression lines (**p < 0.01).

€O. Karadeniz et al. / Journal of Environmental Radioactivity 146 (2015) 27e34 31

correlation (a < 0.01, rp ¼ 0.502). The relation for mixed forest sitesalso had a positive trend (Fig. 2b) but these were not statisticallysignificant. There were statistically significant correlations be-tween the thickness of the humus layer with the mean annualvelocity of 137Cs for both coniferous and mixed forest sites (Fig. 2c,d).

It is well known that coniferous litter decomposes very slowlyand forms a thick humus horizon that richer in 137Cs than those inmixed forest soils. In addition, plant roots are located mainly in theOh horizon of coniferous forest soils, whereas roots are mainly inthe Ah horizon in mixed forest soils (Konopleva et al., 2009). Asmentioned in the previous studies, the relaxation mass depthswere significantly higher in the forest soils than in the grasslandand the cultivated soils, indicating that 137Cs penetration is rapid

Table 1KruskaleWallis test table showing the mean annual velocity (cm year�1) of 137Cs forPinus nigra var pallasiana tree species is significantly higher than that of Pinus brutia.

Tree species Na Median SDb c2c pd

Pinus nigra var pallasiana 78 0.31 1 5 0.023Pinus brutia 36 0.27

a Cases.b Standard deviation.c Chi-square value.d p-Value < 0.05 is significant.

through the organic rich soils with the lower bulk density (Grahamand Simon, 1996; Koarashi et al., 2012; Kato et al., 2012). Therefore,correlations between the thickness of the humus layer with therelaxationmass depths/themean annual velocitymay be attributedto the deeper depth of 137Cs penetration in the thick humus layerwith low bulk density.

GM 1.47 3.52CV (%) 74.7 54.0GSD 2.30 1.89Range 0.03e7.65 0.36e11.72Skewness 1.53 0.77Kurtosis 2.47 0.52Frequency distribution Log-normal Normal

Median, mean (arithmetic mean), standard error of arithmetic mean (S.E.), standarddeviation (S.D.), geometric mean (GM), coefficient of variation (CV), geometricstandard deviation (GSD), range, expressed in nGy h�1 and skewness, kurtosis of thefrequency distributions of the external dose-rates for 137Cs.

a External dose-rates estimated using the conversion factor of 0.03 nGy h�1/Bq kg�1.

b External dose-rates determined from the observed vertical distribution of 137Cs.

€O. Karadeniz et al. / Journal of Environmental Radioactivity 146 (2015) 27e3432

A KruskaleWallis test was used to evaluate the significance ofthe difference between both the relaxation mass depths and themean annual velocity of 137Cs obtained for forest stand types(coniferous, mixed stand with coniferous and deciduous tree spe-cies), but statistically significant differences were not found. Inaddition, the presence of the difference between the mean annualvelocities of 137Cs obtained for the forests with dominant treespecies (Pinus nigra var pallasiana, Pinus brutia) were investigated.Test results (mean rank, standard deviation, chi-square value andp-values) are provided in Table 1. Statistical analysis showed thatthere was a statistically significant difference in the annual velocityof 137Cs between dominant tree species, c2 ¼ 5.192, p ¼ 0.023, witha median value of 0.31 cm year�1 for P. nigra var pallasiana and0.27 cm year�1 for P. brutia. Thus, it is seen that the mean annualvelocity of 137Cs for P. nigra var pallasiana was significantly higherthan that of P. brutia (p < 0.05).

The external dose-rates from the 137Cs in forest soils throughoutthe region were estimated using the conversion factor mentionedabove. Estimations were based on the 137Cs activity concentrationsaveraged over the surface soil layers (OL, OF þ OH and A horizons),because the sources in the few inches of soil (~10 cm) account formost of the exposure rate (Beck et al., 1972) and deposited radio-caesium is still distributed mainly through the upper 10 cm or in

Fig. 3. Spatial distribution of external effective gamma dose rates (D) estimated using the codistribution of 137Cs (b) in the soils of the Mount IDA.

the humicrich layers of forest soil (Karadeniz and Yaprak, 2008a;Petrovic et al., 2013). As shown in Table 2, the external dose-ratesfrom the 137Cs varied between 0.03 nGy h�1 and 7.65 nGy h�1

with a geometric mean of 1.47 nGy h�1. Radiological mapping dataof the external dose-rates from the 137Cs for the region is shown inFig. 3 (a) as a contour map.

The external gamma-dose rates at 1 m height above a flat areafor 137Cs were also estimated from the observed vertical distribu-tion of 137Cs within the soil depth profiles. Because the soil densityaffects the radiation attenuation properties of the soil, the massdepth is a more fundamental quantity to use for purposes of eval-uating the aboveground radiation dose rate. Therefore, these cal-culations were based on conversion factors for the plane sources atdifferent depths and the mass depth of the soil layers. From thevalues of the 137Cs activity and of the bulk density in each soil layerat each site, simple numerical integration of the functions for thecases of exponential, Gaussian or log-normal distributions wasmade as described by Likar et al., 1998. In this connection, theexternal dose-rates from the 137Cs in the investigated area rangedbetween 0.36 nGy h�1 and 11.7 nGy h�1 with a mean of4.17 nGy h�1.

The frequency distributions of the external dose-rates bothestimated from the observed vertical distribution of 137Cs and

nversion factor of 0.03 nGy h�1/Bq kg�1 (a) and determined from the observed vertical

€O. Karadeniz et al. / Journal of Environmental Radioactivity 146 (2015) 27e34 33

determined from the conversion factor were studied. Themeasuredhistograms were compared with the normal and log-normal dis-tribution functions using KolmogoroveSmirnov test values for thegoodness-of-fit. Application of the KolmogoroveSmirnov testshows that in both cases, a normal as well as a log-normal distri-bution cannot be rejected (p > 0.05) for the external dose-ratesestimated from the observed vertical distribution of 137Cs.Regarding the external dose-rates determined from the conversionfactor, KolmogoroveSmirnov test shows that a normal distributionwas rejected (p < 0.05). Another way of testing the similarity be-tween the measured distribution and the normal or log-normaldistributions is to compare the measured median with the arith-metic mean or the geometric mean. If the distribution is normal,then the arithmetic mean should be equal to themedian (Karadenizand Yaprak, 2007b). Table 2 lists the statistical data for the externaldose-rates from 137Cs. Accordingly, application of the Kolmogor-oveSmirnov test, the approximate null value of the skewness co-efficient and the values of the arithmetic mean and the medianfrom Table 2 show that distribution of the external dose-ratesdetermined from the observed vertical distribution of 137Cs fol-lows a normal distribution, while other estimation fit to a log-normal distribution. Because the results for the external dose-rates determined from the conversion factor fit to a log-normaldistribution fairly well, it is convenient to use the geometricmean values as a mean rather than arithmetic mean (€Oztürk et al.,2013).

Interpolated radiological map of the external dose-rates deter-mined from the observed vertical distribution of 137Cs is shown inFig. 3 (b) as a contour map. As seen from Fig. 3 (b) and Table 2, theexternal dose-rate levels were higher compared other estimation.In addition, the spatial distribution pattern of the external dose-rates from 137Cs shows differences with Fig. 3 (a). It is clearlyseen that the external dose-rates from the 137Cs in the forested areawere influenced considerably with the vertical migration rate andthe vertical distribution of 137Cs.

4. Conclusions

This study was carried out to examine the depth-related dis-tributions of 137Cs, to determine the vertical migration of 137Cs andto establish the external dose-rates from 137Cs in the forested areasat the Mount IDA (Kazdagi)/Edremit. Cs-137 activity concentrationswere determined in the forest soil layers separately by using a high-resolution gamma-spectrometer system. The depth distributions of137Cs were exponential, gaussian or log-normal. The relaxationlengths were in the range of 1.09e16.7 cmwith a mean of 5.73 cm,showing a slow transport and a strong retention capacity of 137Cseven after the 26-y period of Chernobyl accident. The mean annualmigration velocity was in the range of 0.11e0.62 cm year�1 with amean of 0.30 cm year�1. The best-fitting relation between thethickness of the humus layer with the relaxation mass depths wasof linear and positive type for coniferous forest sites, while therewere statistically significant correlations between the thickness ofthe humus layer with the mean annual velocity for both coniferousand mixed forest sites. In addition, statistical analysis showed thatthe mean annual velocity of 137Cs for the forests with P. nigra varpallasiana was significantly higher than that of P. brutia.

External dose-rates from the 137Cs in forest soils were estimatedusing the conversion factor used in many studies and comprisedwith the external dose-rates determined according to the verticaldistribution of 137Cs within the soil depth profiles. It is clearly seenthat both levels and spatial distribution patterns of the externaldose-rates from 137Cs were influenced considerably with the ver-tical migration rate and the vertical distribution of 137Cs. Therefore,further study is required to clarify the mechanisms affect the

penetration of 137Cs in soil depth profile in order to establish theexternal dose-rates from the 137Cs in the future and to know howlong 137Cs will remain within the surface layers where it is readilyavailable for root uptake and transfer to crops.

Acknowledgement

Grateful thanks are offered to the provider of financial supportfor the research presented here: The Scientific and TechnicalResearch Council of Turkey (TUB_ITAK) (Project no: 109Y336). Theauthors are also grateful to Prof. Dr. Günseli Yaprak for professionaladvice on several aspects of the Gamma spectroscopy, to Assoc.Prof. Dr. Cüneyt Akal for his indispensable help in drawing radio-logical maps and to Mr. Hüseyin Atay, Mr. Fatih Çoban, Mr. EmirBüyükok and Ms. Sabiha Vurmaz for assisting in sample collection,preparation of soils.

References

Albers, B.P., Rackwitz, R., Schimmack, W., Bunzl, K., 1998. Transect survey of radi-ocesium in soils and plants of two alpine pastures. Sci. Total Environ. 216,159e172.

Almgren, S., Isaksson, M., 2006. Vertical migration studies of 137Cs from nuclearweapons fallout and the Chernobyl accident. J. Environ. Radioact. 91, 90e102.

Arapis, G.D., Karandinos, M.G., 2004. Migration of 137Cs in the soil of sloping semi-natural ecosystems in Northern Greece. J. Environ. Radioact. 77, 133e142.

Baeza, A., Del Rio, M., Miro, C., Paniagua, J., 1993. Surface and depth fallout distri-bution of 137Cs and 90Sr in soils of Caceres (Spain). Dose commitments due toexternal irradiation. J. Radioanal. Nucl. Chem. 175 (4), 297e316.

Barisic, D., Vertacnik, A., Lulic, S., 1999. Caesium contamination and vertical dis-tribution in undisturbed soils in Croatia. J. Environ. Radioact. 46, 361e374.

Beck, H.L., DeCampo, J., Gogolak, C., 1972. In situ Ge(Li) and NaI(Tl) Gamma-raySpectrometry. Report HASL-258 (US-DOE; Environmental Measurements Lab.,New York).

Bunzl, K., Jacob, P., Schimmack, W., Alexakhin, R.M., Arkhipov, N.P., Ivanov, Y.,Kruglov, S.V., 1997. 137Cs mobility in soils and its long-term effect on theexternal radiation exposure. Radiat. Environ. Biophys. 36, 31e37.

Bunzl, K., Albers, B.P., Schimmack, W., Belli, M., Ciuffo, L., Menegon, S., 2000. Ex-amination of a relationship between 137Cs concentrations in soils and plantsfrom Alpine pastures. J. Environ. Radioact. 48, 145e158.

Cha, H., Kang, M., Chung, G.H., Choi, G.S., Lee, C.W., 2006. Accumulation of 137Cs insoils on different bedrock geology and textures. J. Radioanal. Nucl. Chem. 267(2), 349e355.

Clouvas, A., Xanthos, S., Takoudis, G., Antonopoulos-Domis, M., Zinoviadis, G.,Vidmar, T., Likar, A., 2007. Twenty-year follow-up study of radiocesiummigration in soil. Radiat. Prot. Dosim. 124 (4), 372e377.

Dołha�nczuk-�Sr�odka, A., 2012. Estimation of external gamma radiation dose in thearea of Bory Stobrawskie forests (PL). Environ. Monit. Assess. 184, 5773e5779.

El-Reefy, H.I., Sharshar, T., Zaghloul, R., Badran, H.M., 2006. Distribution of gamma-ray emitting radionuclides in the environment of Burullus Lake: I. Soils andvegetations. J. Environ. Radioact. 87, 148e169.

Graham, J.C., Simon, S.L., 1996. A study of 137Cs in soil profiles from the MarshallIslands. Sci. Total Environ. 183, 255e268.

Hao, L.C., Nitta, M., Fujiyoshi, R., Sumiyoshi, T., Tao, C.V., 2013. Radiocesium fallout insurface soil of Tomakomai experimental forest in Hokkaido due to theFukushima Nuclear Accident. Water Air Soil Pollut. 224, 1428.

ICRU, 1994. Gamma-ray Spectrometry in the Environment. ICRU Report 53. Interna-tional Commission on Radiation Units and Measurements, Bethesda, Maryland.

Isaksson, M., Erlandsson, B., 1998. Investigation of the distribution of 137Cs fromfallout in the soils of the city of Lund and the province of Skane in Sweden.J. Environ. Radioact. 38, 105e131.

Jibiri, N.N., Farai, I.P., Alausa, S.K., 2007. Estimation of annual effective dose due tonatural radioactive elements in ingestion of foodstuffs in tin mining area of Jos-Plateau, Nigeria. J. Environ. Radioact. 94, 31e40.

Karadeniz, €O., Karakurt, H., Çakır, R., Çoban, F., Büyükok, E., Akal, C., 2015. Persis-tence of 137Cs in the litter layers of forest soil horizons of Mount IDA/Kazdagi,Turkey. J. Environ. Radioact. 139, 125e134.

Karadeniz, €O., Yaprak, G., 2007a. Dynamic equilibrium of radiocesium with stablecesiumwithin the soil-mushroom system in Turkish pine forest. Environ. Pollut.148, 316e324.

Karadeniz, €O., Yaprak, G., 2007b. Distribution of radiocesium and natural gammaemitters in pine needles in coniferious forest sites of Izmir. Appl. Radiat. Isot. 65(12), 1363e1367.

Karadeniz, €O., Yaprak, G., 2008a. Vertical distributions and gamma dose rates of 40K,232Th, 238U and 137Cs in the selected forest soils in Izmir, Turkey. Radiat. Prot.Dosim. 131 (3), 346e355.

Karadeniz, €O., Yaprak, G., 2008b. Geographical and vertical distribution of radio-cesium levels in coniferous forest soils in Izmir. J. Radioanal. Nucl. Chem. 277(3), 567e577.

€O. Karadeniz et al. / Journal of Environmental Radioactivity 146 (2015) 27e3434

Kato, H., Onda, Y., Teramage, M., 2012. Depth distribution of 137Cs, 134Cs, and 131I insoil profile after Fukushima Dai-ichi Nuclear Power Plant Accident. J. Environ.Radioact. 111, 59e64.

Kim, C.S., Lee, M.H., Kim, C.K., Kim, K.H., 1998. 90Sr, 137Cs, 239þ240Pu and 238Puconcentrations in surface soils of Korea. J. Environ. Radioact. 40 (1), 75e88.

Koarashi, J., Atarashi-Andoh, M., Matsunaga, T., Sato, T., Nagao, S., Nagai, H., 2012.Factors affecting vertical distribution of Fukushima accident-derived radio-cesium in soil under different land-use conditions. Sci. Total Environ. 431,392e401.

Konopleva, I., Klemt, E., Konoplev, A., Zibold, G., 2009. Migration and bioavailabilityof 137Cs in forest soil of southern Germany. J. Environ. Radioact. 100, 315e321.

Lee, M.H., Lee, C.W., Boo, B.H., 1997. Distribution and characteristics of 239,240Pu and137Cs in the soil of Korea. J. Environ. Radioact. 37 (1), 1e16.

Likar, A., Vidmar, T., Pucelj, B., 1998. Monte Carlo determination of gamma-ray doserate with the GEANT System. Health Phys. 75 (2), 165e169.

Livens, F.R., Horril, A.D., Singleton, D.L., 1991. Distribution of radiocesium in the soil-plant systems of upland areas of Europe. Health Phys. 60 (4), 539e545.

Mabit, L., Dornhofer, P., Martin, P., Toloza, A., Zupanc, V., 2012. Dept distributionof selected geogenic radionuclides (40K, 226Ra, 232Th) and anthropogenic 137Csin an undisturbed forest soil in East Slovenia. Indian J. Pure Appl. Phys. 50,45e48.

Matsuda, N., Mikami, S., Shimoura, S., Takahashi, J., Nakano, M., Shimada, K., Uno, K.,Hagiwara, S., Saito, K., 2015. Depth profiles of radioactive cesium in soil using ascraper plate over a wide area surrounding the Fukushima Dai-ichi NuclearPower Plant, Japan. J. Environ. Radioact. 139, 427e434.

Miller, K.M., Kuiper, J.L., Helfer, I.K., 1990. 137Cs fallout depth distributions in forestversus field sites: implications for external gamma dose rates. J. Environ.Radioact. 12 (1), 23e47.

Nada, A., Abd-ElMaksoud, T.M., Abu-ZeidHosnia, M., El-Nagar, T., Awad, S., 2009.Distribution of radionuclides in soil samples from a petrified wood forest in El-Qattamia, Cairo, Egypt. Appl. Radiat. Isot. 67, 643e649.

Nikolova, I., Johanson, K.J., Stephen, C., 2000. The accumulation of 137Cs in thebiological compartment of forest soils. J. Environ. Radioact. 47, 319e326.

€Oztürk, B.C., Çam, N.F., Yaprak, G., 2013. Reference levels of natural radioactivity and137Cs in and around the surface soils of Kestanbol pluton in Ezine region ofÇanakkale province. Turk. J. Environ. Sci. Health A 48, 1522e1532.

Petrovic, J., Cujic, M., Ðordevic, M., Dragovic, R., Gajic, B., Miljanic, S., Dragovic, S.,2013. Spatial distribution and vertical migration of 137Cs in soils of Belgrade

(Serbia) 25 years after the Chernobyl accident. Environ. Sci. Process. Impacts 15,1279e1289.

Rafferty, B., Brennan, M., Dawson, D., Dowding, D., 2000. Mechanisms of 137Csmigration in coniferous forest soils. J. Environ. Radioact. 48, 131e143.

Ramzaev, V., Yonehara, H., Hile, R., Barkovsky, A., Mishine, A., Sahoo, S.K.,Kurotaki, K., Uchiyama, M., 2006. Gamma-dose rates from terrestrial andChernobyl radionuclides inside and outside settlements in the Bryansk Region,Russia in 1996e2003. J. Environ. Radioact. 85, 205e227.

Saito, K., Petoussi-Henss, N., 2014. Ambient dose equivalent conversion coefficientsfor radionuclides exponentially distributed in the ground. J. Nucl. Sci. Technol.51 (10), 1274e1287.

Saito, K., Tanihata, I., Fujiwara, M., Saito, T., Shimoura, S., Otsuka, T., Onda, Y.,Hoshi, M., Ikeuchi, Y., Takahashi, F., Kinouchi, N., Saegusa, J., Seki, A.,Takemiya, H., Shibata, T., 2015. Detailed deposition density maps constructed bylarge-scale soil sampling for gamma-ray emitting radioactive nuclides from theFukushima Dai-ichi Nuclear Power Plant accident. J. Environ. Radioact. 139,308e319.

Schuller, P., Bunzl, K., Voigt, G., Ellies, A., Castillo, A., 2004. Global fallout 137Csaccumulation and vertical migration in selected soils from South Patagonia.J. Environ. Radioact. 71, 43e60.

Schuller, P., Voight, G., Handl, J., Ellies, A., Oliva, L., 2002. Global weapons' fallout137Cs in soils and transfer to vegetation in South-central Chile. J. Environ.Radioact. 62, 181e193.

Schuller, P., Walling, D.E., Sepulveda, A., Castillo, A., Pino, I., 2007. Changes in soilerosion associated with the shift from conventional tillage to a no-tillage sys-tem, documented using 137Cs measurements. Soil Tillage Res. 94 (1), 183e192.

Shcheglov, A.I., Tsvetnova, O.B., Klyashtorin, A., 2014. The fate of Cs-137 in forestsoils of Russian Federation and Ukraine contaminated due to the Chernobylaccident. J. Geochem. Explor. 142, 75e81.

Staunton, S., Dumat, C., Zsolnay, A., 2002. Possible role of organic matter in radio-cesium adsorption in soil. J. Environ. Radioact. 58, 163e173.

Takahashi, J., Tamura, K., Suda, T., Matsumura, R., Onda, Y., 2015. Vertical distribu-tion and temporal changes of 137Cs in soil profiles under various land uses afterthe Fukushima Dai-ichi Nuclear Power Plant accident. J. Environ. Radioact. 139,351e361.

Zygmunt, J., Chibowski, S., Klimovicz, Z., 1998. The effect of sorption properties ofsoil minerals on the vertical migration rate of cesium in soil. J. Radioanal. Nucl.Chem. 231 (1e2), 57e62.