1 Introduction

Arid and semi-arid lands represent 40% of the Earth’s land area, and it is assumed that more than one third of the world’s human population depends directly on these agroecosystems. These lands are under increasing threat of soil degradation by erosion, which is being exacerbated by improper soil management associated with subsistence farming, deforestation, and nutrient mining (Prosdocimi et al. 2016; Vaezi et al. 2017).

Annual soil loss of cropland is reported to be 10 million hectares worldwide (Pimentel and Burgess 2013). In Iran, recent investigations carried out by the Soil Conservation and Watershed Management Research Institute documented a mean erosion rate of ca. 16.5 Mg ha−1 year−1 which is expected to increase further over the next decades in response to climate change and to the increase of human population (see Yang et al. 2003; Nearing et al. 2004). The value of damages caused by erosion in Iran is estimated to be $56 billion per year (Nikkami et al. 2010). More specifically, Rastgar et al. (2015), investigating on-site costs of soil erosion in a watershed in north of Iran, reported that soil erosion and associated losses of N, P, and K led to significant financial losses of ca. 71.5 US$ ha−1 year−1.

Precise information on on-site soil redistribution rates could help landowners to develop effective conservation measures by applying targeted mitigation approaches. In order to complement existing conventional methods, the search for reliable tools for evaluating soil erosion has led to the development of nuclear techniques based on the use of fallout radionuclides (FRNs). FRNs, mainly cesium-137 (137Cs) and unsupported lead-210 (210Pbex), can be very helpful in evaluating the effectiveness of agricultural conservation measures, such as minimum tillage, contour farming, terraces, mulching, use of cover crops, and the establishment of agro-forestry.

Cesium-137 (half-life = 30.2 years) is an anthropogenic radioisotope that was released into the stratosphere and then deposited worldwide following the past nuclear weapon tests. Lead-210 (half-life = 22.3 years) is a natural geogenic radioisotope, generated by the uranium (238U) decay series, and produced by the gaseous radon-222 (222Rn), a daughter of radium 226 (226Ra). The proportion of 210Pb produced from the lithogenic escape of 222Rn is deposited on the soil surface via precipitation and is called “excess” or “unsupported” 210Pb (210Pbex). Because it is not in equilibrium with its parent 226Ra, 210Pbex can be used as a soil movement tracer. Like 137Cs, 210Pbex is rapidly adsorbed onto fine soil particles, and then redistributed through agricultural tillage, soil erosion, and sediment transport processes (Mabit et al. 2008). Soil redistribution rates derived from 137Cs and 210Pbex determinations are based on a comparison between the total inventory from an individual sampling point and a local reference inventory obtained in stable (without erosion and deposition) and undisturbed areas (see Mabit et al. 2008).

Although a minor proportion of 137Cs or 210Pbex can be mobilized by chemical and biological processes, it is demonstrated that the use of these two FRNs allows assessing soil erosion rates for periods up to 60 and 100 years, respectively (Walling and He 1999; Walling et al. 2003; Porto et al. 2006).

Over the past four decades, the 137Cs method has been used with success worldwide (e.g., Mabit et al. 2012, 2013; IAEA 2014). However, 210Pbex as tracer for use in soil redistribution assessment requires specific analytical equipment, i.e., “N”-type HPGe detector and skilled staff (Shakhashiro and Mabit 2009; Mabit et al. 2014). 210Pbex may not be applicable or reliable under some specific agro-environmental conditions, such as those occurring locally in Austria (Mabit et al. 2009) and Mongolia (Kato et al. 2010). In contrast, several successful applications of the 210Pbex technique have been reported in Zambia (Walling et al. 2003), China (Zhang et al. 2006), Italy (Porto et al. 2009), Spain (Gaspar et al. 2013), and Morocco (Benmansour et al. 2013).

Several investigations have also reported the usefulness of 137Cs and 210Pbex in evaluating the influences of conservation measures on soil erosion magnitudes (e.g., Belyaev et al. 2009; Rafiq et al. 2011; Dercon et al. 2012). Current knowledge on soil erosion rates in Iran is limited, especially when using the 137Cs technique (Gorji et al. 2008; Abbaszadeh et al. 2010; Ayoubi et al. 2012; Emadodin and Bork 2012; Rahimi et al. 2013), and, up to now, there are no published contributions reporting the use of 210Pbex to assess agro-ecosystem soil erosion magnitude in this country.

In Iran, large areas covered by steep hillslopes, naturally under rangeland land use, have been commonly converted to dry-farming lands (Scherr and Yadav 1996). Following this land use change, soil losses are estimated to increase by almost seven times (Soil and Water Conservation Institute, pers. comm.). Possible explanations of the increase in soil erosion rates in these dry farms can be attributed to down-slope cultivation, leaving the soil surface bare, burning crop residuals, inappropriate crop rotation, and intensive tillage. Therefore, dry farming lands, occupying ca. 54% of the Iran’s agricultural land, can be considered one of the most fragile agro-ecosystems (Siadat et al. 1997).

In spite of the high rates of soil erosion occurring in Iran, many watershed management measures have been sparsely adopted throughout the country, e.g., contour farming, conservational tillage, terraces, check dams, etc. (see Sadeghi 2017). However, no practical, extensive country-wide plan has been employed to control soil losses yet and, more important, the effectiveness of soil conservation measures has not been scientifically and accurately assessed so far (Sadeghi 2017).

In Kouhin, a region in the Iranian province of Qazvin, soil erosion is a serious agro-environmental threat and requires urgent attention and effective mitigation actions. In this region, soils are prone to erosion due to steep topography and heavy rainfalls occurring during late winter and early spring. The inappropriate past and current land management practices also exacerbates the risk for soil erosion.

The present investigation was undertaken to gain information on the usefulness of soil conservation strategies on soil degradation of a rainfed hillslope in the Kouhin region, using 137Cs and 210Pbex. The objectives of this study were (i) to assess soil redistribution rates in managed and unmanaged hillslopes and (ii) to deepen our knowledge on the effectiveness of soil conservation measures in controlling soil erosion.

2 Materials and methods

2.1 Study area

The area under investigation (36°22′26″ to 36°22′33″N and 49°34′53″ to 49°35′07″E) is located in the Kouhin region, ca. 40 km northeast from Qazvin (Fig. 1). The experimental site is located in the soil and water conservation research station of Tehran University, Iran (Fig. 1). The average altitude is about 1354 m a.s.l., and the landscape topography ranges from gentle to fairly steep hillslopes (4–22%). Average annual precipitation and air temperatures are 330 mm and 12.1 °C respectively (Iran Meteorological Organization). The soils of the area, classified as Vertic Calcixerepts, were developed on the alluvial deposits of marls and brown to gray limestones (Gorji et al. 2004). The topsoil texture is dominated by clay with 14–17% of CaCO3. The soil temperature and moisture regimes are mesic and xeric, respectively. Soil conservation practices such as controlled grazing, broad-base terraces, and contour farming have been implemented for about 50 years at the research station (i.e., managed site) (Gorji et al. 2004). Outside that station (i.e., unmanaged site), there are no such conservation measures and the soil is being cultivated along the main slope. At both sites, the tillage depth is around 20 cm.

Fig. 1
figure 1

The location of the study fields at Kouhin, in Qazvin Province, Iran. (NB: the position of the reference site, sampling points inside and outside the research station and Broad-Base Terraces are also indicated)

Full size image

Two adjacent hillslopes—both with northern exposures—were selected for this study: (a) a 7-ha managed hillslope with an average slope steepness of 21% situated inside the research station and (b) a 3.5-ha un-managed cultivated field with a slope steepness of 22% located outside the station (see Figs. 1 and 2a).

Fig. 2
figure 2

A view of the study sites (a), the reference site (b), and incremental soil sampling using the scraper plate device (c)

Full size image

2.2 Soil sampling design

Finding a suitable reference site is the first critical step for using FRN techniques (Mabit et al. 2008, 2012). An undisturbed, flat site, which had a cover of perennial grass, was selected within the research station. This reference site is located ca. 800 m from the managed study area (see Figs. 1 and 2b). Three sampling campaigns have been undertaken in this area to establish the basic levels of 137Cs and 210Pbex. During the first campaign, 11 bulk cores and a sectioned soil core were collected to determine initial fallout of 137Cs (see Gorji et al. 2004). A second sampling campaign was undertaken to update and validate the initial inventory of FRNs at this benchmark site in 2012. In this case, three sectioned cores were collected with a 1-m distance from the same area. Samples were collected to a depth of 40 cm at 2-cm increments using a scraper plate sampling devise as suggested by Campbell et al. (1988) (Fig. 2c). In the third campaign, five bulk cores were collected in undisturbed areas of the reference site in 2017 to further check the basic levels of FRNs.

At the two cultivated sites, a multi-transect sampling approach—validated already by several researchers such as Benmansour et al. (2013)—was implemented. In the managed site, a total of 44 bulk soil cores have been collected along four parallel transects in the main slope direction (Fig. 1). Sampling points along each transect were spaced ca. 25 m apart. In the unmanaged site, 55 soil cores were collected along 6 transects, each with a different distance according to the topography variability. Soil sampling was performed with a soil column cylinder auger (inner diameter of 8.3 cm) up to 40 cm (a 50-cm depth was used for the samples located at the bottom of the slope, where sedimentation processes were expected) to make sure that all FRNs in the soil profile were included.

2.3 Soil sample pre-treatment and gamma spectrometry analysis

The sample pre-treatment was carried out according to Pennock and Appleby (2002). All soil samples have been air dried, manually disaggregated, sieved to < 2 mm, ground, and then homogenized. Soil samples collected in the managed site were analyzed for 137Cs, 210Pb, and 226Ra by gamma spectrometry using HPGe detectors. Gamma spectrometry measurements were performed at the Nuclear Science and Technology Research Institute of Iran. Two high-resolution coaxial detectors, “P”-type with relative efficiency of 80% for measuring 137Cs and “N”-type with relative efficiency of 30% for measuring 210Pb and 226Ra, were used. Due to a long waiting list of the P-type detector with relative efficiency of 80%, the soil samples collected in the unmanaged site were analyzed by another P-type HPGe detector having a relative efficiency of 10%. In order to account for the uncertainty related to the use of different detectors, a preliminary analysis consisting of measuring a number of samples with both detectors was performed. The two types of equipment did not show substantial differences between the two series of the same samples.

Following the recommendations provided by Shakhashiro and Mabit (2009), the calibration of both detection systems was established with certified multi-gamma sources. The calibration accuracy was confirmed by using IAEA reference materials (Soil 6 and Soil 375) in the same geometry used for the measured samples.

Prior to analysis of 210Pb, sub-soil samples were sealed for one month to reach the equilibrium between 226Ra and 222Rn (IAEA 2011). 137Cs, total 210Pb, and 226Ra (from 214Bi) activities were determined at 661.6, 46.5, and 609 keV, respectively. Counting times up to 24 h allowed obtaining a precision of 5 to 20% at the 95% level of confidence. The self-absorption correction was taken into account based on transmission measurements for reporting total 210Pb activity from the emitted gamma rays at 46 keV (Cutshall et al. 1983; Khater and Ebaid 2008; Robu and Giovani 2009). To correct for self-absorption, a point source with a similar energy than 210Pb (241Am ~ 59.5 keV) was utilized. The 210Pbex mass activity was determined by deducting 226Ra from total 210Pb measurement. The detection limits were 0.5 and 5 Bq kg−1 for 137Cs and total 210Pb, respectively.

3 Results

3.1 Vertical distribution and inventory of 137Cs and 210Pbex in the reference site

At the benchmark site, the vertical distribution of both FRNs was similar and decreased exponentially with increasing depth (Fig. 3). Most of the FRN activity was found in the top soil (i.e., 0–10 cm) and no 137Cs or 210Pbex were detectable below 35 and 30 cm, respectively, or their concentration was lower than their detection limits. The mean reference inventory value for 137Cs was estimated to be 2195 ± 120 Bq m−2 with a CV of 5.5%. The 137Cs inventory at the reference site is consistent with that previously measured by Gorji et al. (2004) at the Kouhin research station (2380 ± 214 Bq m−2, decay corrected) and was used as mean reference inventory value for 137Cs.

Fig. 3
figure 3

Depth distribution of 137Cs (a) and 210Pbex (b) at the reference site (NB: The bars represent the estimated error of the measurement and the determination for 137Cs and 210Pbex respectively)

Full size image

The mean reference inventory value for 210Pbex was estimated to be 5825 ± 297 Bq m−2 with a CV of 5.1%. Considering constant fluxes and radioactive decay, the annual deposition flux of 210Pbex was estimated to be 181 Bq m−2 for our study area, a level which is in the range of the reported global annual deposition fluxes of 210Pbex (see details in the discussion section).

3.2 Radionuclide inventories in the study area

The 137Cs inventories at the managed site ranged from 662 to 3287 Bq m−2 with a 30% coefficient of variation (CV). The 137Cs inventories at the unmanaged site ranged from 832 to 2608 Bq m−2 with a CV of 29% (see Table 1). Most of the 137Cs activities in soil samples at the cultivated field were smaller than those recorded at the benchmark location as erosion in the managed and unmanaged sites occurred in 62 and 88% of hillslope, respectively (Table 2). These high values of CVs were not unexpected. In fact, at the managed site, the main soil conservation measures were based on the establishment of broad-base terraces which shorten the slope length. These practices have produced a drastic change in the flow pattern, and this has affected as well the soil redistribution inducing high variability in the inventories. Concerning the unmanaged site, where tillage operations were usually made before the rainfall season, it is possible that a portion of 137Cs fallout may have escaped from the homogenization produced within the tillage depth. However, similar results are documented in the literature. For example, Gaspar et al. (2013) reported a large variability in 137Cs inventories in heterogeneous Mediterranean environments. Another possible reason of this high variability could be due to the impact of the Chernobyl accident. In the study area, this impact was assumed to be negligible even if there is very limited information on whether Iran was affected by Chernobyl fallout or not as well as distribution and possible magnitude of that. ArabKhedri (2014) investigated 137Cs fallout data throughout Iran and concluded that some parts of the country could have been affected by the 137Cs Chernobyl fallout. In this regard, some Chernobyl-derived radiocaesium fallout was detected by the authors in northern and western parts of the country where the amount of precipitation is much higher (unpublished data). However, the study site is located in the central part of Iran where precipitation is very low (Fig. 1) and based on local information, the resulting fallout would not have affected for more than 4% the basic 137Cs level in this area.

Table 1 Summary of 137Cs and 210Pbex inventories at managed and unmanaged sites
Full size table
Table 2 Sediment budget based on 137Cs and 210Pbex data set at managed and unmanaged sites
Full size table

At the managed site inside the research station, the 210Pbex inventories, which were measured along three transects, ranged between 1929 and 10,131 Bq m−2 with a CV of 45%. This high CV can be also explained by the analytical precision of the gamma measurement because the relative uncertainties obtained for 210Pbex determinations reached 15 to 20% while relative uncertainties associated with 137Cs measurements ranged from 5 to 7%.

3.3 Assessment of soil redistribution rates estimated by 137Cs at the managed and unmanaged hillslope sites

Calculation of soil erosion and deposition rates from 137Cs measurements requires the use of a conversion model able to derive loss or gain of 137Cs into values of soil loss or deposition. However, prior to this calculation, it was necessary to establish if there was any significant difference between the managed and the un-managed sites in order to evaluate the effect of soil conservation measures. A rigorous statistical test based on a one-way analysis of variance (ANOVA) was performed on the 137Cs inventories obtained in the two datasets. A preliminary logarithmic transformation was necessary to normalize data. All statistical analyses were carried out using SPSS software ver. 16 (IBM, Armonk, NY, USA). The results highlight that the two datasets are significantly different at the p-level < 0.01 (see Table 1).

The 137Cs inventories obtained in the two areas were then converted into rates of soil redistribution using the conversion model MBM2 (Walling et al. 2002). In this respect, a value of 4 kg m−2 for the relaxation mass depth (H), a value of 0.5 for the proportion factor (γ), a bulk density of 1300 kg m−3 (reflecting the use of minimum tillage), a plow mass depth of 260 kg m−2, and a particle size factor of 1 have been set based on our experimental condition (Walling et al. 2002; IAEA 2011, 2014). For more details on the conversion model MBM2 and the definition of its parameters, the readers are referred to Walling et al. (2002) and IAEA (2014). In our study, the Excel Add-in “radiocalc.xla” developed at the University of Exeter (see http://www-naweb.iaea.org/nafa/swmn/models-tool-kits.html) was used to estimate the soil redistribution rate at each sampling point.

Soil erosion rates at the managed site ranged between 0.01 and 44.2 Mg ha−1 year−1; deposition values in the same site ranged between 2.3 and 22.5 Mg ha−1 year−1. The corresponding estimates obtained at the unmanaged site, based on a bulk density of 1250 kg m−3 (reflecting the use of conventional tillage), ranged between 0.3 and 35 Mg ha−1 year−1 for soil erosion and between 0.2 and 8.5 Mg ha−1 year−1 for soil deposition (see Table 2). The net soil loss from the unmanaged hillslope was 12 Mg ha−1 year−1 inducing a sediment delivery ratio (SDR) of 97%. The SDR, here calculated as the ratio of net erosion to gross erosion rate (IAEA 2014), highlights that most of the eroded soil has been moved out of the field. These results were expected due to performing soil cultivation along the main slope direction, slopes that can reach 22%. In contrast, the net soil erosion rate at the managed site resulted 4.6 Mg ha−1 year−1, a value significantly lower than the erosion rate evaluated at the unmanaged hillslope, and this produced a value of SDR equal to 52%. This result suggests that soil conservation measures, such as terracing of fields, reduced significantly net soil losses from these lands. Indeed, channel terraces with reduction of slope length have a key function in trapping the eroded particles. The above results are in line with previously reported studies that also showed a reduction in soil losses by erosion after introducing a terrace system hillslope (e.g. Hatch 1981; Chow et al. 1999, 2010; Golosov et al. 2011; Dercon et al. 2012).

3.4 Determination of the soil redistribution rates using 210Pbex at the managed site

The MBM2 was also used to convert the 210Pbex inventories into soil redistribution rates at the managed site. This calculation provided values of soil erosion rate ranging from 5.8 to 62.8 Mg ha−1 year−1 and deposition values from 0.9 to 59.2 Mg ha−1 year−1 (Table 2). The corresponding values of net soil erosion and SDR resulted equal to 12.3 Mg ha−1 year−1 and 60%, respectively. These estimates are generally higher than those obtained with 137Cs and suggest possible discrepancies between the results obtained by the two radionuclides. Possible explanations of these contrasting results will be provided later in the discussion section.

3.5 Soil redistribution pattern in the managed and unmanaged fields

The existence of channel terraces led to high spatial variations in 137Cs inventory; therefore, erosion and deposition patterns resulted into highly variable values at short distances. Also, geostatistical analyses did not highlight the presence of any significant spatial structure for both 137Cs and 210Pbex datasets. Therefore, as recommended by Mabit and Bernard (2007), to produce soil redistribution maps from the interpolation of the soil erosion and deposition rates obtained from both radionuclides, the Inverse Distance Weighting power 2 (IDW2) spatialization option of the Surfer software package 8.0 was employed (see Fig. 4). In fact, the IDW interpolation technique involves a weighted average interpolator which attributes more weight to adjacent points, so the weights are inversely proportional to a power of the distance (Webster and Oliver 2007). The value of the power is commonly chosen to be 2 (Isaaks and Srivastava 1989).

Fig. 4
figure 4

Maps of soil redistribution rates (Mg ha−1 year−1) provided by 137Cs (a), 210Pbex (b) at the managed site, and by 137Cs (c) at the unmanaged site using IDW2. (In a and b, red lines are Broad-Base Terraces positions. In c, brown and blue lines indicate short embankment and road position, respectively)

Full size image

The resulting soil redistribution pattern for the managed site was considerably affected by the position of the broad-base terraces. Soil erosion generally was found just below a terrace, and deposition occurred at the bottom of each terrace (see Fig. 4a, b).

Broad-base terraces reduce the slope length and are reported to trap the eroded soil particles, thus reducing both downslope runoff and sediment movement. Golosov et al. (2011), Li and Lindstrom (2001), Li et al. (2011), and Gaspar et al. (2013) observed similar soil redistribution patterns in areas with channel terraces using 137Cs and 210Pbex radionuclides as soil tracers. The fluctuating soil movement magnitudes along each transect, estimated by both FRNs, demonstrated that the terraces exert a major control on soil redistribution processes. At the bottom of each transect, both radionuclides indicated sediment deposition. However, the rates derived from 210Pbex were appreciably higher than those obtained with the 137Cs approach. The concordance between the estimated soil distribution and the topography feature confirmed that both 137Cs and 210Pbex radionuclides can be used successfully for investigating soil redistribution.

In contrast to the results above, which dealt with the managed hillslope, the conventionally cultivated hillslope highlighted severe 137Cs losses at most individual sampling points. However, at the unmanaged hillslope, a short embankment was made allowing deposition of material as highlighted by the 137Cs technique (Fig. 4c).

4 Discussion

As reported above, the mean reference inventory values were estimated to be 2195 ± 120 Bq m−2 for 137Cs and 5825 ± 297 Bq m−2 for 210Pbex. The 137Cs reference value obtained in our study area is consistent with that previously measured by Gorji et al. (2004) at the Kouhin research station (2380 ± 214 Bq m−2, decay corrected) and was used as a mean reference value in the conversion model. This result is encouraging and confirms that potential fallout associated with Chernobyl event is negligible in this area.

About the reference value indicated by 210Pbex, to our knowledge, no reference inventories or fallout fluxes of 210Pbex have been reported for Iranian reference sites. Different values of 210Pbex for undisturbed sites have been reported worldwide, e.g., 5170 Bq m−2 in UK (Walling and He 1999), 5730 and 12,860 Bq m−2 in China (Zhang et al. 2003, 2006), 19,703 and 8453 Bq m−2 in Japan (Fukuyama et al. 2008; Wakiyama et al. 2010), 6310 Bq m−2 in Mongolia (Kato et al. 2010), 5266 and 7794 Bq m−2 in Italy (Porto et al. 2006; Porto and Walling 2012, respectively), 2000 Bq m−2 in Zambia (Walling et al. 2003), 3305 Bq m−2 in Morocco (Benmansour et al. 2013), 620 Bq m−2 in Austria (Mabit et al. 2009), 2020 Bq m−2 in Spain (Gaspar et al. 2013), and 34,000 Bq m−2 in Taiwan (Huh and Su 2004). This wide range of 210Pbex values for undisturbed sites could be associated with the difference in substratum, annual rainfall, latitude, and location of the investigated areas (Mabit et al. 2008, 2014; Matisoff 2014). Worldwide, the annual deposition fluxes of 210Pbex ranged between 23 and 367 Bq m−2. The 210Pbex value of 5825 Bq m−2 obtained in our reference site lies within the results obtained in the above cited studies. Based on this value, the annual deposition flux of 210Pbex was estimated to be ca. 181 Bq m−2 for our study area, a level that is in the range of the reported global annual deposition fluxes.

Based on the above reference values, the application of the 137Cs technique highlighted lower net soil erosion at the managed site as compared to the un-managed site. This demonstrated the effectiveness of conservation practices in reducing soil losses in semi-arid farm lands of Kouhin. When compared to the cultivated hillslope without soil conservation measures, the terracing of managed hillslopes reduced the net soil losses and the SDR by 38 and 54%, respectively. Broad-base terraces shorten the slope length, trap the eroded soil particles along the hydraulic path, and reduce downslope runoff and sediment movement. Our results are in good agreement with previous studies highlighting the reduction of soil losses after introducing a terracing system (e.g., Hatch 1981; Chow et al. 1999, 2010).

However, under our experimental condition, the soil redistribution rates derived from 137Cs and 210Pbex are not similar (Fig. 5). The soil redistribution rates derived from 210Pbex were generally higher than those obtained from 137Cs (both methods were tested only at the managed site) which are in line with the results obtained by Walling et al. (2003) and Kato et al. (2010). Effectively, these two FRNs are unlikely to yield the same findings since the mean soil redistribution rates obtained from 137Cs inventories are associated with a time period from 1963 to the time of sampling. In contrast, the mean soil redistribution rates calculated from 210Pbex inventories estimate rates across the past 100 years (Walling et al. 2003). But, as highlighted by Porto et al. (2013, 2014), current inventories of 210Pbex, are expected to be more sensitive to erosive events that occurred during the last 15 to 20 years. In Iran, Balling et al. (2016) reported a positive trend for extreme precipitation events over the period 1951–2007. Despite that we have not been able to establish such a trend in our study area, the intensity and pattern of precipitation might have changed significantly (Fig. 6a, b). There is then a possibility that the higher soil losses estimated by the 210Pbex approach reflect an increasing climate variability during the last two decades. But more investigation would be needed to confirm this hypothesis. Indeed, changes in land uses (Zhang et al. 2006; Gaspar et al. 2013) and, to some extent, the rainfall characteristics during the different timescales could have a significant impact on the estimated soil erosion magnitudes obtained by these FRNs. Such a discrepancy was also found by Wakiyama et al. (2010) in Japan.

Fig. 5
figure 5

Distribution of soil erosion and deposition rates (Mg ha−1 year−1) derived from 137Cs and 210Pbex along the three transects of the managed site

Full size image
Fig. 6
figure 6

Annual maximum one-day precipitation (mm) (a) and number of days with precipitation higher than 10 and 20 mm (b) at the study area based on daily records from the Iran Meteorological Organization

Full size image

The high soil erosion rates established for the managed terraced hillslope area are most likely due to the past agricultural mismanagement of the field (prior to the establishment of channel terraces), as well as its steep topography and heavy rainfall occurring throughout the late winter and early spring seasons when the soil remain almost bare.

5 Conclusions

The Kouhin research station in Qazvin, where soil conservation measures have been adopted for ca. 50 years, was selected for evaluating the effectiveness of soil conservation measures using estimates of soil redistribution rates based on 137Cs measurements. This well-known technique highlighted that the conservation measures (i.e., broad-base terraces and contour farming) were effective in reducing the magnitude of soil erosion. Our results also showed the potential of using the 210Pbex approach for estimating soil redistribution at terraced hillslope in semi-arid climatic condition. Estimates of soil redistribution provided by 137Cs and 210Pbex were, however, not similar. This likely reflects the fact that soil erosion processes in the agricultural field had changed substantially during the last four decades, when channel terrace systems were introduced around 50 years ago. It should be noted that the 137Cs technique is likely to provide the more reliable results, relative to 210Pbex due to its well-known higher gamma measurement precision. However, 210Pbex can be exploited as a substitute soil tracer in agricultural areas which experienced low 137Cs fallout. Further studies and researches are required to improve the accuracy of 210Pbex determinations to establish more reliable evaluation of soil redistribution magnitude in agro-ecosystems.