Determination of arsenic and cadmium uptake by rice from topsoil

Topsoil is the main sink of various pollutants and the direct source of heavy metal uptake by crops. In a 2‐year field experiment conducted in central China from 2016 to 2017, the significant role of topsoil in pollutant absorption, in particular the arsenic (As) and cadmium (Cd) uptake in rice was examined quantitatively. Two soil treatments, removing half of the topsoil (TR) and a control (CK), were applied alongside two rice varieties ‐ Huanghuazhan (HHZ, indica inbred) and Yangliangyou6 (YLY6, indica hybrid). On average, TR reduced the As and Cd concentrations in grain by 10.0% and 56.2%, respectively. The reduction was linked to lower pollutant concentrations in soil above the hardpan and rice straw. TR also decreased shoot As and Cd accumulation by 30.7% and 67.8% at maturity, with topsoil at 12–15 cm depth contributing an average of 2372 μg As m−2 (30.1%) and 138 μg Cd m−2 (66.1%) to total uptake of As and Cd, respectively. Contribution of topsoil to total As uptake remained consistent across growing stages, while for Cd, topsoil contributed 24.8% before heading and 81.2% after. Moreover, varietal differences were observed, with TR significantly reducing grain As and Cd concentrations of YLY6 but no difference in HHZ. This study quantifies topsoil's impact on As and Cd uptake in rice, and underscores genotypic variations in their response to topsoil removal.


INTRODUCTION
It is a great accomplishment that China has successfully fed 22% of the population with only 7% of the farmlands of the world. 1 However, soil contamination in China has attracted considerable public attention both domestically and internationally due to rapid urbanisation and industrialisation in the recent decades. 2 How to ensure the food safety of agricultural products is currently one of the most important research subjects.Heavy metals have -1 of 11 https://doi.org/10.1002/moda.25     become an important soil contaminant because of their toxicity and difficulty in degradation, 3 and they can be inadvertently taken up by crop roots, posing a serious threat to the health of consumers. 4The Chinese government and researchers are anxious to comprehensively understand the pollution status of heavy metals in farmland soil and the transfer ability of toxic elements from soil to edible parts of crops.
According to a nationwide survey carried out from 2005 to 2013 in China, 19.4% of the agricultural soil samples were contaminated based on environmental quality standard (GB15618-1995). 5The dominant contaminants in soil were heavy metals and metalloids, with cadmium (Cd) and arsenic (As) ranking first and third, respectively.Huang et al. 6 conducted a metaanalysis of 336 articles published from 2005 to 2017, which reported heavy metal concentrations in agricultural soils in China, and the results showed that the concentrations of most heavy metal elements in agricultural soils decreased, while the concentrations of As and Cd obviously increased between 2012 and 2017.Our recent study showed that approximately half of 110 rice samples contained excessive concentrations of total As (>200 μg kg −1 ) in brown rice. 7He et al. 8 reported that 20.7% of 169 rice samples collected from a former electronic-waste dismantling region exceeded the Cd threshold value (200 μg kg −1 ).Therefore, designing strategies for lowering the concentrations of As and Cd in agricultural soils and products is of great importance for food safety.
Rice is the staple food of Chinese population and the second largest food crop in terms of harvest area in China. 9However, rice is also the main dietary source of As and Cd to consumers who feed on rice. 10Previous studies reported that rice accounts for 60% and 56% of total dietary inorganic As and Cd intake of the general population in China, respectively. 11,12Compared with other food crops such as maize, wheat and barley, rice exhibits a higher capacity to accumulate As and Cd. 13,14On the one hand, the bioavailability of the two elements in paddy fields is generally high.Specifically, long-time submerged conditions promote the mobility of soil As. 15 Acidic soil properties of paddy fields, especially in southern China, enhance the solubility of soil Cd. 16 On the other hand, rice inherently has a strong absorbing and translocating ability for As and Cd.For example, rice roots take up arsenite (As(III)) and arsenate (As(V)) through the uptake pathway of silicon and phosphate, respectively, 17,18 and primarily take up Cd via the manganese transporter OsNRAMP5. 19inimizing the bioavailability of As and Cd in paddy soil and planting varieties with low accumulation capacity of As and Cd are the main strategies for decreasing the As and Cd concentrations in rice grain. 10egulating water management is an effective way to control the bioavailability of As and Cd in paddy soil.However, the effect of water management on the bioavailability of As and Cd in soil is opposite, making it difficult to simultaneously reduce As and Cd accumulation in rice though a single water regime, especially in soils with combined pollution of them. 20,21There are large differences across varieties in As and Cd concentrations in rice grain. 22With the identification of numerous quantitative trait loci (QTLs) controlling grain As and Cd concentrations, it is promising to select new rice varieties with low As and Cd accumulation through molecular breeding. 10Up to now, several varieties with low Cd accumulation have been developed, 23,24 while the development process of varieties with low As accumulation is slow.Thus, exploring other effective agronomic measures to simultaneously reduce As and Cd accumulation in rice is still important.
Topsoil is the direct source of heavy metal absorption by plant roots.Due to the inputs of atmospheric deposition, livestock manures, fertilizers and agrochemicals, sewage irrigation and sewage sludge, the contaminants are increasing accumulated in agricultural soils. 25Previous studies indicated that heavy metals exhibit different distribution patterns at different soil layers, with the concentrations of heavy metals in the topsoil being higher than that in the subsoil, especially for Cd. 26,27Zhang et al. 28 reported that plough tillage significantly reduced Cd concentrations in grains of rice and wheat in comparison with rotary tillage and no-tillage.However, the direct information about the effects of topsoil removal on As and Cd accumulation in rice is limited.Moreover, the differences among rice varieties in the responses of As and Cd accumulation to topsoil removal need to be investigated.
To address these knowledge gaps, a 2-year field experiment was conducted in a farmer's field in Hubei Province of central China.In this study, the soil was removed at about half the topsoil depth, and an inbred and a hybrid rice variety were planted.The objectives of this study were to: (1) investigate the effects of topsoil removal on As and Cd accumulation in rice; (2) compare the differences in responses of As and Cd accumulation to topsoil removal between inbred and hybrid varieties; and (3) quantify the contribution of topsoil to the accumulation of As and Cd in rice.We hypothesized that topsoil removal would reduce the accumulation of As and Cd in rice, and hybrid rice has a higher response in As and Cd accumulation to topsoil removal than inbred rice.

Site description
During the rice-growing season from May to October 2016, a field experiment was conducted in a farmer's fields at Dajin Township, Wuxue County, Hubei Province, central China (29°51 0 N, 115°33 0 E).The experiment was repeated in an adjacent field in 2017.The study area is located in the subtropical monsoon humid climate region, where the soil has a silty loam texture.

Experimental design and crop management
The experiments were arranged in a split-plot design with soil treatments as main plots and varieties as subplots with four replicates.The plot size was 23.5 and 23.2 m 2 in 2016 and 2017, respectively.The two soil treatments were removing half of the topsoil (TR) and control (CK).Before the start of field experiments, the topsoil depth above the hardpan was measured.Then, half of the topsoil was removed under TR treatment and used to build bunds for separating subplots.To ensure the precision of the removal process, the topsoil depth was measured every square metre, and the error in soil depth was controlled within 1 cm.Finally, the topsoil depths of CK and TR plots were 30 and 15 cm in 2016, respectively.In 2017, the topsoil depths of CK and TR plots were 25 and 13 cm, respectively.There were no soil bunds within each CK plot, and plastic baffles were used to prevent the exchange of water and fertilizers across the subplots.Soil samples were collected each year before applying basal fertilizers to analyse soil chemical properties, 29 and total As and total Cd concentrations (Table 1).The depth of soil sampling was 20 cm within CK plots, while soil samples were taken from surface to hardpan (only 12-15 cm) within TR plots.Two high-yielding and widely planted varieties, namely Huanghuazhan (HHZ, indica inbred) and Yan-gliangyou6 (YLY6, indica hybrid), were used in this study.Inbred and hybrid lines are two distinct rice types in terms of breeding methods.There are large differences in the genetic background and growth characteristics between HHZ and YLY6.
Pre-germinated seeds were sown into seedbeds on 13 May 2016 and 14 May 2017.Seedlings of 29-and 33-d-old were transplanted into the paddy field on 11 June 2016 and 16 June 2017, respectively.Transplanting was done at a density of 25 hills m −2 and a hill spacing of 13.3 � 30.0 cm with two seedlings per hill.A total of 100 kg N ha −1 was applied in three splits: 40% as basal fertiliser, 30% at 10 d after transplanting, and 30% at panicle initiation (PI).A total of 40 kg P ha −1 was applied as basal fertiliser.A total of 100 kg K ha −1 was split equally as basal fertiliser and at PI.The N, P and K fertilizers were applied in the form of urea, calcium superphosphate and potassium chloride, respectively.The field was irrigated immediately after transplanting and a floodwater depth of 3-5 cm was maintained until 1 week before maturity except for drainage during the maximum tillering period to reduce unproductive tillering.Weeds, pests and diseases were strictly controlled during the rice growing season.

Sampling and measurements
A total of 12 hills were sampled from each plot at midtillering (MT), PI, heading (HD) and maturity (MA).The plant samples were separated into stems (culm plus sheath), leaves and panicles (when present).Panicles were hand-threshed and divided into filled grains and the remaining parts of panicles.The dry weights of each organ at all growth stages were determined after ovendrying at 80°C to constant weight.
Soil samples were naturally dried at room temperature in the laboratory, crushed and sieved using a 0.2 mm (100 mesh) nylon sieve.Filled grains were ground to fine powder using a stainless steel grinder (MM400, Retsch GmbH, Germany).Other plant organs were ground to fine powder with a laboratory mill (Model 4, THOMAS-Wiley, USA).By using a microwave digester (MARS6, CEM Microwave logy Ltd., USA), soil samples (0.5 g each) were digested with 10 mL of aqua regia (mixed solution of high-purity concentrated HCL and HNO 3 , 3:1 of v/v) and plant samples (0.5 g each) were digested with 10 mL of high-purity concentrated HNO 3 .The enter method parameters of digestion programme were shown in Table S1.In each batch of 40 samples, there were two certified reference materials of rice flour (GBW 100358 and 100,362) and three reagent blanks for quality assurance.After digestion, the digests were dissolved in 2% HNO 3 .Throughout the entire experiment process, all glassware used for analysis was soaked in 10% HNO 3 for more than 12 h in advance, and all experimental reagents used for analysis were of guaranteed grade and deionised water was used.Before the determination of As concentration, a mixed solution of 1% thiourea and 5% HNO 3 was added to the original solution to reduce arsenate to arsenite.The total As concentration was determined using an atomic fluorescence spectrophotometer (AFS-933, Beijing Titan Instruments Co., Ltd., China).The Cd concentration was determined using an atomic absorption spectrometer (PinAAcle 900T, Perki-nElmer, USA).The accumulation of As or Cd in each organ was calculated by multiplying the dry weight by the As or Cd concentration in the organ.The total aboveground accumulation of As or Cd was the sum of the accumulation of As or Cd in each organ.The transfer factor (TF) of As or Cd was calculated as follows: where C straw and C grain are the concentrations of As or Cd in straw (stem plus leaf) and filled grain, respectively.The As or Cd uptake from topsoil was calculated by the difference in total As or Cd uptake between CK and TR based on the "N-difference method". 30

Data analysis
Analysis of variance was performed using Statistix 9.0 (Analytical Software).Mean comparison was based on the least significant difference (LSD) test at the 0.05 probability level.

As and Cd concentrations in grain and straw, and TF straw-grain
There was a significant difference in As concentration in grain and straw between TR and CK, and this was also true for Cd (Table 2).Averaged across years and varieties, compared with CK, TR significantly reduced As concentrations in grain and straw by 10.0% and 16.5%, and Cd concentrations in grain and straw by 56.2% and 62.5%, respectively.However, the TF straw-grain of As and Cd was significantly higher in TR than in CK.
Compared with CK, TR increased TF straw-grain of As and Cd by 16.7% and 18.2%, respectively.Among the two varieties, no significant difference was observed in grain As concentration.However, on average, YLY6 had 5.9% higher straw As concentration while 14.3% lower TF straw-grain than HHZ.There was a significant difference in grain Cd concentration between HHZ and YLY6.On average, YLY6 had 77.7% higher grain Cd concentration than HHZ.Moreover, the straw Cd concentration and TF straw-grain of YLY6 were 34.3% and 28.6% higher than those of HHZ, respectively.
However, there was a significant interaction between varieties and soil treatments in terms of grain As and Cd concentrations (Table S2).Specifically, TR significantly reduced As and Cd concentrations in grain of YLY6, but this was not true for HHZ.For YLY6, the decrease in the concentrations of As and Cd in grain under TR was attributed to the decrease in the concentrations of As and Cd in straw, and no difference was observed for TF straw-grain .Whereas for HHZ, TR significantly reduced As and Cd concentrations in straw but increased the TF straw-grain (except for As in 2017), resulting in similar grain As and Cd concentrations with CK.
Overall, the average As and Cd concentrations in the grain were 532.2 and 30.6 μg kg −1 , respectively.The grain As concentration exceeded the national food safety standard (200 μg kg −1 , GB 2762-2017).The concentration of As in straw was much higher than that of Cd, whereas the TF straw-grain of As was much lower than that of Cd.

As and Cd uptake in shoot
There were significant differences in As uptake in shoots at different growing periods between CK and TR (Table 3).Across years and varieties, compared with CK, TR reduced total As uptake at MA by 30.7%.TR significantly reduced the As uptake from TP to MT, from MT to There was a significant difference in As uptake in shoots at different growing periods between HHZ and YLY6.Across years and soil treatments, the total As uptake at MA was 15.3% higher in YLY6 than that in HHZ.Compared with HHZ, YLY6 had 37.5%, 22.2%, and 59.1% higher As uptake from TP to MT, from MT to PI, and from PI to HD, respectively, but 47.4% lower As uptake from HD to MA.Moreover, compared with other growing periods, the As uptake from PI to HD was the highest, accounting for an average of 51.5% of the final total As uptake.Except for TP to MT, there were significant differences in Cd uptake in shoots at different growing periods between CK and TR (Table 4).Across years and varieties, compared with CK, TR reduced total Cd uptake at MA by 67.8%.The Cd uptake from MT to PI, from PI to HD, and from HD to MA was 22.6%, 29.6%, and 82.0%lower in TR than that in CK, respectively.Significant differences in Cd uptake from TP to MT and from HD to MA were observed between HHZ and YLY6.Across years and soil treatments, the total Cd uptake at MA was 60.0% higher in YLY6 than that in HHZ.Compared with HHZ, YLY6 had 45.2% and 103.4% higher Cd uptake from TP to MT and HD to MA, respectively.Moreover, compared with other growing periods, the Cd uptake from HD to MA was the highest, accounting for an average of 66.0% of the final total Cd uptake.
Overall, the uptake capacity of As in shoots was much higher than that of Cd.Hybrid rice showed a higher uptake ability of As and Cd than inbred rice.The periods from PI to HD and from HD to MA were the main periods for As and Cd uptake in shoots, respectively.

As and Cd concentrations in stem and leaf
Figure 1 shows the As concentration in stem and leaf at different growing stages.Along with the growth process of rice, stem As concentration first decreased and then increased, while leaf As concentration gradually increased.Overall, the stem As concentration in TR was significantly lower than that in CK at most growing stages.For leaf As concentration, significant differences were observed between TR and CK at MA. Figure 2 shows the Cd concentration in stem and leaf at different growing stages.The stem Cd concentration was relatively lower before HD, and then obviously increased, especially under CK.For leaf Cd concentration, there was an obvious increase from PI to MA.The leaf Cd concentration under TR was significantly lower than that under CK at MA. Correlation analysis (Figure 3) shows that grain As concentration was significantly and positively correlated with leaf As concentration (R 2 = 0.91), but not with stem As concentration (R 2 = 0.01).However, grain Cd concentration was closely related to stem Cd concentration (R 2 = 0.70), rather than leaf Cd concentration (R 2 = 0.10).

As and Cd uptake in shoot from topsoil
Along with the growth process of rice, the As and Cd uptake from topsoil gradually increased (Figure 4a,d).During the entire growing season, the topsoil provided an average of 2372 � 468 μg As m −2 (30.1% � 4.6%) and 138 � 32 μg Cd m −2 (66.1% � 3.5%) for total T A B L E 3 As uptake at different growing periods of two varieties under different soil treatments in 2016 and 2017.

Year Variety Treatment
As uptake (μg m −2 ) MODERN AGRICULTURE uptake of As and Cd, respectively.There was no significant difference in As uptake from topsoil and the ratio of As uptake from topsoil to plant As uptake among different growing periods (Figure 4b,c).Specifically, the average ratio of As uptake from topsoil to plant As uptake during the period from TP to MT, from MT to PI, from PI to HD, and from HD to MA was 29.6%, 29.9%, 15.7%, and 37.6%, respectively.However, the Cd uptake from topsoil was significantly higher from HD to MA than that from other growing periods (Figure 4e).The ratio of Cd uptake from topsoil to plant Cd uptake was also the highest between HD and MA (Figure 4f).Before and after heading, the ratio of Cd uptake from topsoil to plant Cd uptake was 24.8% and 81.2% on average, respectively.

DISCUSSION
The impact of topsoil on crop productivity has been widely studied in previous research studies.For example, 5-10 cm of topsoil removal reduced soybean

F I G U R E 3 Correlation analysis between As or Cd concentration in grain and
As or Cd concentration in stem and leaf.
MODERN AGRICULTURE and maize yields by 5%-9% and 10%-13%, respectively. 31In addition, removing 20 cm of topsoil reduced wheat yield by 53%. 32In our previous research on rice, 11-15 cm of topsoil removal resulted in a yield decrease of 16.7%, and the decrease in total N uptake was responsible for the yield reduction. 29However, limited information is available on the effect of topsoil removal on the crop uptake of heavy metal elements.In the present study, on average, removing 12-15 cm topsoil reduced As and Cd concentrations in rice grain by 10.0% and 56.2%, respectively (Table 2).Therefore, removing topsoil is a potential strategy to simultaneously reduce As and Cd concentrations in agricultural products, especially for protected agriculture.However, considering the limitations of increasing economic costs and reducing soil fertility, topsoil removal may not be suitable for large-scale applications. 29veraged across years and varieties, topsoil removal reduced As and Cd uptake in shoots by 30.7% and 67.8%, respectively (Tables 3 and 4).Under controlled conditions, the period from PI to HD had the highest As uptake in shoots compared to other growing periods, followed by the period from HD to MA.However, the Cd uptake in shoots from HD to MA was much higher than that of other growing periods, accounting for 66.0% of the total Cd uptake in shoots.Yamaji and Ma 33 indicated that before and after heading of rice was the main stage for As uptake due to the enhanced expression of Lsi1, a transporter for As.Moreover, the grain filling stage was proved to be the most critical period for Cd uptake in rice. 20During these growing periods, topsoil removal largely reduced the As and Cd accumulation in shoots, ultimately resulting in a significant decrease in total accumulation of As and Cd in shoots.
The reduction in As and Cd uptake in shoots due to topsoil removal was mainly attributed to two aspects.
Firstly, removing half of topsoil reduced the As and Cd concentrations in the soil above the hardpan.In this study, the As and Cd concentrations in soil under TR were 7.9% and 7.7% lower than that under CK, respectively (Table 1).Secondly, topsoil removal greatly restricted the root growth of rice.This was because that paddy fields in most Asian rice-producing regions typically had a hardpan with a thickness of 5-10 cm, located 10-40 cm below the soil surface. 34,35In the two experimental fields of our study, the hardpan was located 25-30 cm and 12-15 cm below the soil surface under CK and TR treatments, respectively.Thus, the root penetration would be largely limited above the hardpan under TR. 36owever, there was a difference in the response of As and Cd concentrations in grain to topsoil removal between HHZ and YLY6 (Table 2).For example, topsoil removal significantly reduced As and Cd concentrations in grain of YLY6, while no difference was observed in As and Cd concentrations in grain of HHZ between TR and CK.Although topsoil removal significantly reduced As and Cd concentrations in straw for both varieties, the TF straw-grain of As and Cd of HHZ significantly increased under TR, which was different from that of YLY6.These results suggested that the As and Cd concentrations in grain of YLY6 were more responsive to topsoil removal than that of HHZ.Moreover, YLY6 had a higher Cd concentration in grain than HHZ.It has been reported that there were large genotypic variations in heavy metal accumulation in rice grains. 37,38Previous studies indicated that indica rice tends to accumulate more Cd than japonica rice, 39 and hybrid rice appears to have a higher Cd accumulation capacity than inbred rice. 40owever, Duan et al. 22 investigated the grain As and Cd concentrations of 471 rice varieties, and the results showed that there was no significant difference in the As or Cd concentration in grain between indica and japonica rice, as well as between hybrid and inbred rice.Therefore, the interaction between genotype and environment on heavy metal accumulation in rice should not be neglected. 38n the present study, the grain As concentration was relatively high, especially in 2016, while the grain Cd concentration was very low in both years (Table 1).Water management practices greatly affect As and Cd concentrations in rice grain. 20,21During the rice growing season in 2016 and 2017, the experiment fields were kept flooded from transplanting to 1 week before maturity except for drainage during the maximum tillering period to reduce unproductive tillering.This conventional irrigation regime greatly promotes the As accumulation, whereas conversely prevents the Cd accumulation in rice. 10Another possible reason was that the field soil had relatively high total As concentration but low total Cd concentration (Table 1).The grain As concentration was much higher in 2016 than in 2017, which might be due to the cultivation of dryland crops in the fields before this experiment.Rice planted in newly converted paddy soils or in dryland-paddy rotations is generally prone to accumulate As. 41 In this study, the As and Cd uptake from topsoil and the contribution of topsoil to total As and Cd uptake in rice were quantified (Figure 4).The As or Cd uptake from topsoil was estimated by the difference in total As or Cd uptake between CK and TR.This calculation method was based on the estimation of the contribution of soil-derived N to total plant N uptake under N-applied treatments, which was known as the N-difference method. 30In our previous study, we quantified the contribution of topsoil to total N uptake in rice, and the results showed that 11-15 cm of topsoil contributed 34.4-68.5 kg N ha −1 . 29The results of this study showed that the topsoil at the depth of 12-15 cm contributed an average of 2372 μg As m −2 (30.1%) and 138 μg Cd m −2 (66.1%) to total uptake of As and Cd, respectively.Moreover, the contribution of topsoil to total As uptake during different growing periods was similar, but the topsoil contributed an average of 81.2% of total Cd uptake during HD to MA.However, there were some limitations for the "difference method" to estimate the contribution of topsoil.On the one hand, it was notable that the N-difference method would overestimate the contribution of fertiliser-N to total plant N uptake due to the "added N interaction", with an overestimation rate of approximately 7%. 42On the contrary, topsoil removal might underestimate the contribution of topsoil to total As and Cd uptake in rice because the uptake of As and Cd from the residue soil under TR treatment was likely increased.On the other hand, the coefficient of variation of the contribution of topsoil was relatively high, which was attributed to the combined effects of experimental factors, such as year, soil and variety, in this study.
It is a great challenge to protect arable soil from heavy metal contamination, which directly relates to food safety and human health.Excessive exposure to heavy metals through dietary intake can lead to serious healthy diseases, such as skin lesions, diabetes, cardiovascular disease, renal dysfunction and cancer. 43,44 this study, we emphasised the importance of topsoil in heavy metal accumulation in food crops.Therefore, the most important thing is to identify and stop the sources of contamination in the topsoil, which requires stricter monitoring and effective implementation of environmental protection policies, especially for industrial emission sources, irrigation water quality, fertiliser quality, etc. 2 Then, it is necessary to explore other agronomic mitigation strategies to reduce heavy metals enrichment in topsoil.For example, firstly, phytoremediation is regarded as a low-cost and environmentally friendly technology for cleaning contaminated soils, and planting hyperaccumulators can be used to remove heavy metals instead of removing topsoil. 45urakami et al. 46 reported that 2 years of phytoextraction resulted in a reduction of 38% and 47% in Cd concentrations in soil and subsequently planted rice, respectively.Secondly, considering the concentration of As and Cd in the topsoil as demonstrated in present and previous studies, 26,27 mixing topsoil and subsoil through deep plough tillage is an alternative strategy to reduce heavy metal accumulation of crops, 28 which indicated that plough tillage significantly reduced Cd concentration in rice grain by 69% in comparison with rotary tillage.Thirdly, managing the water content in the topsoil is also effective in controlling the bioavailability of As and Cd.According to our results on the absorption pattern of As and Cd during different rice growing periods from the topsoil, it is recommended to reduce irrigation before heading, and to keep flooded from heading until a week before harvest to simultaneously control the accumulation of As and Cd in rice grain.Moreover, although our study focuses on the importance of topsoil in heavy metal accumulation in rice, future research can focus on investigating the impact of topsoil removal on other crop species and examining its long-term effects on soil health.

CONCLUSION
The accumulation of As and Cd in the grain and shoot of rice were significantly reduced by removing half of the topsoil.The decrease in As and Cd concentrations in soil above the hardpan and rice straw under TR was responsible for the reduction in grain As and Cd concentrations.However, the response of grain As and Cd concentrations to TR was different between the two varieties, which was attributed to the difference in the transfer process from straw to grain of As and Cd.The periods from PI to HD and from HD to MA were the main uptake periods for As and Cd in rice shoots, respectively.During these growing periods, the As and Cd accumulation in shoots was largely reduced under TR.Moreover, we quantified the As and Cd uptake in rice from topsoil on the basis of "difference method".The contribution of topsoil to total As uptake during different growing periods was similar.From HD to MA, the topsoil had the greatest contribution to the total Cd uptake.The study highlighted the importance of topsoil on As and Cd accumulation in rice, and demonstrated genotypic variation in response of As and Cd MODERN AGRICULTURE

F I G U R E 4
As and Cd uptake from topsoil at different growing stages and periods, and the ratio of topsoil As and Cd to plant As and Cd uptake.Data were averaged across four combinations of varieties and years.Error bars indicate SE (n = 4).HD, heading; MA, maturity; MT, mid-tillering; PI, panicle initiation; TP, transplanting.
Total As and Cd concentrations in soil of experimental fields in 2016 and 2017.
11A B L E 1Note: Within a column for each year, means followed by different letters are significantly different according to LSD (0.05).Abbreviations: CK, control; TR, topsoil removal.MODERN AGRICULTURE-3 of11

Table 1
shows the total As and Cd concentrations in the soil above the hardpan under different soil treatments.Compared with CK, on an average, TR significantly reduced the total As concentration in soil by 7.9%.The effect of soil treatment on total Cd concentration in soil

Variety Treatment Conc. in grain (μg kg −1 ) Conc. in straw (μg kg −1 ) TF straw-grain
As and Cd concentrations in grain and straw, and transfer factor of As and Cd from straw to grain of two varieties under different soil treatments in 2016 and 2017.
T A B L E 2Note: Within a column for each variety, means followed by different lowercase letters are significantly different according to LSD (0.05).Different uppercase letters indicate significant difference among varieties or treatments according to LSD (0.05).Abbreviations: CK, control; HHZ, Huanghuazhan; TR, topsoil removal; YLY6, Yangliangyou6.PI, from PI to HD, and from HD to MA by 32.3%, 34.7%, 21.8% and 43.0%, respectively.
Within a column for each variety, means followed by different lowercase letters are significantly different according to LSD (0.05).Different uppercase letters indicate significant difference among varieties or treatments according to LSD (0.05).
Cd uptake at different growing periods of two varieties under different soil treatments in 2016 and 2017.Within a column for each variety, means followed by different lowercase letters are significantly different according to LSD (0.05).Different uppercase letters indicate significant difference among varieties or treatments according to LSD (0.05).