Phytoextraction by Rice Capable of Accumulating Cd at High Levels

Environ. Sci. Technol. 2009, 43, 5878–5883

Downloaded by UNIV OF CALIFORNIA SANTA BARBARA on August 27, 2015 | http://pubs.acs.org Publication Date (Web): July 2, 2009 | doi: 10.1021/es8036687

Phytoextraction by Rice Capable of Accumulating Cd at High Levels: Reduction of Cd Content of Rice Grain M A S A H A R U M U R A K A M I , * ,† FUMIHIKO NAKAGAWA,‡ N O R I H A R U A E , †,| M A S A S H I I T O , § A N D TOMOHITO ARAO† Soil Environment Division, National Institute for Agro-Environmental Sciences, 3-1-3, Kannondai, Tsukuba, Ibaraki 305-8604, Japan, Department of Agro-Environmental Research, Yamagata Agricultural Research Center, 6060-27 Minorigaoka, Yamagata, Yamagata 990-2372, Japan, Department of Production and Environment, Agricultural Experiment Station, Akita Prefectural Agriculture, Forestry and Fisheries Research Center, Genpachisawa 34-1, Yuwaaikawa, Akita, Akita 010-1231, Japan

Received December 24, 2008. Revised manuscript received June 2, 2009. Accepted June 18, 2009.

In accordance with a new international standard set by the Codex Alimentarius Commission for cadmium (Cd) content in rice grain, Japan must perform large-scale remediation of paddy fields polluted with low to moderate levels of Cd. Phytoextraction using hyperaccumulating wild plants has been proposed as a low-cost, environmentally friendly restoration technology. However, because of difficulties with sowing, weed and disease control, and harvesting, hyperaccumulators may not be suitable for large-scale phytoextraction in polluted paddy fields. Here, we demonstrated phytoextraction using Indicatype rice cultivars capable of accumulating Cd at high levels. Phytoextraction with the Indica rice Chokoukoku grown for 2 years without irrigation after drainage removed 883 g Cd ha-1, reduced the total soil Cd content by 38%, and reduced the grain Cd content in subsequently grown Japonica food rice by 47% without decreasing yield. The results suggest that phytoextraction with Chokoukoku can remove Cd from paddy fields polluted with low to moderate levels of Cd and reduce the grain Cd concentration of Japonica food rice cultivars to below the Codex standard within a reasonable time frame. This approach will help reduce the risk of Cd contamination of rice from paddy fields.

Introduction Cadmium (Cd) is more mobile and bioavailable than other metals, and it is toxic to humans at concentrations lower than those toxic to plants because its effects on humans are cumulative (1). Soil pollution by Cd has been a public concern * Corresponding author phone: +81-29-838-8308; fax: +81-29838-8313; e-mail: [email protected]. † National Institute for Agro-Environmental Sciences. ‡ Yamagata Agricultural Research Center. § Akita Prefectural Agriculture, Forestry and Fisheries Research Center. | Present address: Graduate School of Agricultural Science, Kobe University, 1-1 Rokkodai, Nada, Kobe, Hyogo 657-8501, Japan. 5878

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 15, 2009

since daily ingestion of high-Cd rice (Oryza sativa L.) was found to be the main cause of Itai-Itai disease in the 1970s (2). Since then, Japan has used engineering techniques in Cd-polluted paddy fields to combat this condition (3). However, in accordance with a new international standard set by the Codex Alimentarius Commission for the Cd content of rice grains0.4 mg kg-1 (4), which is stricter than the Japanese standard of 1 mg kg-1sJapan must perform large-scale remediation of paddy fields that cover more than 40 000 ha and are polluted with Cd at low to moderate levels (5, 6). The use of engineering techniques may satisfy this requirement, but such methods are extremely environmentally disruptive and expensive and are impractical for treating large areas (7). Although phytoextraction by hyperaccumulating wild plants such as Thlaspi caerulescens has been proposed as a low-cost, environmentally friendly restoration technology for soils contaminated with toxic metals, the difficulties with sowing, weed and disease control, and harvesting of such hyperaccumulators suggest that they may be unsuitable for large areas (8–10). The use of agricultural species adapted to these growing conditions may therefore be a better alternative. In previous studies, we found several Indica-type rice cultivars that are capable of accumulating Cd at high levels (11, 12), and we reported the effects of phytoextraction by high-Cd-accumulating rice on the seed Cd content of subsequently grown soybean (13). In this study, we show that phytoextraction using these Indica-type rice cultivars was capable of accumulating Cd at high levels. We examined the reduction in soil Cd concentration in a paddy field with a moderate level of Cd contamination and the decrease in grain Cd concentration of a Japonica food cultivar after phytoextraction with these Indica-type rice cultivars.

Materials and Methods Experimental Field. The field experiment was performed in a paddy field located in the Tohoku District of eastern Japan during the summers of 2004 to 2007. The physicochemical properties of this paddy soil are shown in Table S1 (Supporting Information). The main source of Cd contamination appeared to have been irrigation wastewater from an abandoned copper mine; the mine had been operated from 1937 to 1979. The Cd concentration of irrigation water since this mine closed down has been below the Japanese environmental quality standards for river water pollution by Cd (0.01 mg kg-1; (14)). The geometric means (and range) of background metal levels in Japanese agricultural soils (mg kg-1) are 0.295 (0.056-0.801) for Cd, 19.0 (7.95-44.0) for Cu, 17.2 (9.25-41.8) for Pb, and 59.9 (16.0-105) for Zn (15). In the soils studied, the concentrations of Cd (1.6 mg kg-1), Pb (66.8), and Zn (133.9) were considered to represent low to moderate levels of contamination. Phytoextraction using Rice Cultivars Capable of Accumulating Cd at High Levels (Experiment 1). The experimental paddy field consisted of two plots (each 19 × 6.5 m), and each plot was divided into five subplots (one 7 × 6.5 m subplot, three 3.5 × 6.5 m subplots, and one 1.5 × 6.5 m subplot). The Indica-type rice cultivars IR8 (Indica) and Milyang 23 (an Indica-Japonica hybrid) both showed the ability to accumulate Cd at high levels in our previous study (11, 12). Moretsu is an Indica rice cultivar used as a highbiomass forage crop in Japan, whereas Akitakomachi is a Japonica food rice cultivar that shows low Cd accumulation (12). In 2004, we cultivated three Indica-type rice cultivars (IR8, Milyang 23, and Moretsu) in the 3.5 × 6.5 m subplots, and we used the 7 × 6.5 m subplot as a control with no plants. We also cultivated the Japonica cultivar Akitakomachi 10.1021/es8036687 CCC: $40.75

 2009 American Chemical Society

Published on Web 07/02/2009


in the 1.5 × 6.5 m subplot to compare shoot Cd uptakes by the Indica-type cultivars with that by the Japonica rice cultivar. In 2005, the subplot used as a control in 2004 was subdivided into two 3.5 × 6.5 m subplots. In our previous study the Indica rice cultivar Chokoukoku had shown the ability to accumulate Cd at high levels (Ito et al., unpublished). We used the two 3.5 × 6.5 m subplots to cultivate the Chokoukoku and as a control. In 2006, the Indica-type IR8, Milyang 23, Moretsu, and Chokoukoku were cultivated in the same subplots as used in 2005. The control subplot was the same as in 2005. The treatments thus followed a randomized block design with two replicates. In late April 2004, seeds of the Indica-type IR8, Milyang 23, and Moretsu and the Japonica Akitakomachi were sown in seedling trays and grown for 1 month. The rice seedlings were then transplanted into the flooded subplots in late May by a transplanting machine (PA63, Iseki & Co., Ltd., Ehime, Japan). The site was divided into a grid of 16 × 30 cm cells, and each rice seedling was transplanted into the center of a cell, giving an overall planting density of 20.8 plants m-2. In 2005 and 2006, the Indica-type IR8, Milyang 23, Moretsu, and Chokoukoku were grown and transplanted in the same manner as the rice cultivars in 2004. From 2004 to 2006, basal fertilizer was supplied at a rate of 80 kg N ha-1, 22.5 kg P ha-1, and 33 kg K ha-1 in the form of a mixed fertilizer (Aladdin 484, ZEN-NOH, Tokyo, Japan). Additional fertilizers were supplied in mid-June at rates of 20 kg N ha-1 and 9 kg K ha-1and in late July at 40 kg N ha-1 and 19 kg K ha-1, in the form of a mixed fertilizer (NK-kasei C68, ZEN-NOH). Previous research has shown that the Cd concentration in rice shoots grown under flooded (reducing) soil conditions may be low, because Cd solubility under these conditions is lower than under oxidizing conditions (16). Because shoot Cd uptake by rice plants equals the product of the dry weight (DW) and the Cd concentration of the rice shoots, maximizing shoot Cd uptake requires management practices that enhance both DW and Cd uptake by the rice shoot. The DW of rice shoots grown under flooded soil conditions is higher than that under oxidizing soil conditions during tillering (from transplanting to 30 days before panicle initiation; (17)). Therefore, we maintained the soils of all subplots under flooded conditions during tillering from 2004 to 2006 to maximize the DW of the rice shoots (Table S2). Once the floodwater was drained from the subplots, we maintained the soils under oxidizing conditions until harvesting to maximize Cd accumulation by the rice shoots. After transplanting the rice seedlings, we sampled one rice shoot from each subplot to determine the shoot DW and shoot Cd concentration, and we then multiplied these two values to obtain the shoot Cd uptake. We performed this sampling five times before harvesting in each year. At harvesting, we harvested six rice shoots per cultivar to again determine the DW, Cd concentration, and Cd uptake. To do so, we cut the stems 5 cm above ground level, because that is the lowest height of stem that can be harvested by most common rice harvesters used in Japan. We then harvested all the remaining rice shoots with a rice binder (RA50, Kubota, Osaka, Japan), again cutting the stems approximately 5 cm above ground level. Treatment of the sampled and harvested rice shoots is described in the section “Soil and Plant Analysis”. Effect of Phytoextraction on Grain Cd of Food Rice Subsequently Grown in the Remediated Plots (Experiment 2). To examine the effect of phytoextraction with Indicatype rice cultivars capable of accumulating Cd at high levels on the grain Cd concentrations of a food rice cultivar subsequently grown in the remediated soil, we used the traditional Japanese rice cultivation method to grow the Japonica food rice Yumesayaka in the control plot (in which no plants were grown) and in all the subplots that had

undergone 3 years of phytoextraction by the Indica-type rice cultivars IR8, Moretsu, and Milyang 23 or 2 years of phytoextraction by the Indica rice cultivar Chokoukoku. Basal fertilizer was supplied to all subplots at a rate of 50 kg N ha-1, 14 kg P ha-1, and 21 kg K ha-1 in the form of a mixed fertilizer (Aladdin 444, ZEN-NOH). Additional fertilizer was supplied at rates of 20 kg N ha-1 and 9 kg K ha-1 in mid-June and late July in the form of a mixed fertilizer (NK-kasei C68). Seeds of the Japonica Yumesayaka were sown in seedling trays in late April 2007 and grown for 1 month. The rice seedlings were then transplanted into the flooded subplots on May 25 by using the same transplanting machine used in Experiment 1. The planting density of the Yumesayaka rice seedlings was the same as that used in Experiment 1 (20.8 plants m-2). Plots were flooded to a depth of 5 cm, as in Experiment 1 (Table S2). We drained the plots 40 days after transplanting (1 month before panicle initiation) to prevent root rot (caused by reducing soil conditions) and the development of nonproductive tillers. This management technique is called “midseason drainage”. After 1 week of midseason drainage, we flooded the subplots again to a depth of 5 cm, then irrigation was stopped until the floodwater had disappeared from the soil surface (typically after about 3 days), and the soil of the subplots was maintained under oxidizing conditions for 2 days. Flooding and drainage were performed in this manner three times (“intermittent irrigation”). The subplots were then flooded again to a depth of 5 cm for 2 weeks (1 week before and 1 week after panicle initiation). This management technique is called “irrigation for the flowering stage”. Irrigation was then stopped until the floodwaters had disappeared from the soil surface. We performed an additional three cycles of intermittent irrigation, and then we maintained the soils of the plots under oxidizing conditions until harvesting. The dates of the key management steps from transplanting to harvesting are shown in Table S2. In early October 2007, we harvested the unhulled rice grains from six rice shoots per subplot. Treatment of these rice grains is described in the section “Soil and Plant Analysis”. Soil Sampling. We sampled the soils in early May from 2004 to 2007. Traffic by heavy agricultural machines had compacted the soil to about 30 cm, creating a traffic pan. Plowing temporarily loosened the compacted surface soil to a depth of about 15 cm (the plow layer), but increased compaction just below the plow layer (from about 15 to 30 cm), creating a combined traffic pan and plow pan (18). Root growth is seriously impeded by this compaction (19). However, the pans effectively impeded drainage of floodwaters from the paddy fields. Because each rice seedling was transplanted into the center of a 16 × 30 cm cell, we sampled a single 16 × 30 × 15-cm-deep block of soil from each subplot before plowing in each year. The stubble (the rice stem between ground level and 5 cm above the soil) and the root were combined and regarded as the root component of the biomass in our analysis, and this material was included in the soil sample. After sampling of the soil, the root component was carefully removed from each soil block. The soil was then air-dried and passed through a 2-mm stainless-steel sieve before soil analysis. Soil and Plant Analysis. We measured the pH, total carbon (TC), and total nitrogen (TN) in the soil; their mean values in the experimental field and the methods used are shown in Table S1. The values of pH, TC, and TN of soil in each subplot before plowing from 2004 to 2007 are shown in Table S3. To determine which soil Cd fractions were decreased by plant growth, we analyzed the Cd contents of the soils by using two single-extraction methods: 0.1 mol L-1 HCl (20) and Mehlich 3 (21). We also used a sequential-extraction method (22), in which the soil Cd fractions were the exchangeable Cd fraction, the inorganically bound Cd fraction, the organically bound Cd fraction, the oxide VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5879

TABLE 1. Shoot Cd Concentrations from after Transplanting to Harvesting in the Indica-Type Rice Cultivars Capable of Accumulating Cd at High Levels and in the Japonica Food Rice Cultivar Akitakomachi, 2004 to 2006 shoot Cd concentration (mg kg-1)


cultivar 2004 IR8 (I) Moretsu (I) Milyang 23 (IJ) Akitakomachi (J)

June 23 0.2 ( 0.1aa 0.1 ( 0.0a 0.2 ( 0.0a 0.1 ( 0.0a

July 7 0.1 ( 0.0b 0.1 ( 0.0b 0.2 ( 0.0a 0.1 ( 0.0b

Aug 12 2.5 ( 0.1b 3.1 ( 0.1a 2.4 ( 0.2b 1.4 ( 0.1c

2005 IR8 (I) Moretsu (I) Milyang 23 (IJ) Chokoukoku (I)

June 27 0.3 ( 0.0b 0.2 ( 0.0b 0.3 ( 0.0b 0.5 ( 0.0a

July 7 0.2 ( 0.0ab 0.1 ( 0.1b 0.2 ( 0.0b 0.4 ( 0.0a

July 28 2.6 ( 0.4b 2.7 ( 0.3b 2.9 ( 0.0b 4.8 ( 0.4a

2006 IR8 (I) Moretsu (I) Milyang 23 (IJ) Chokoukoku (I)

June 13 0.1 ( 0.0b 0.2 ( 0.0b 0.1 ( 0.0b 0.3 ( 0.0a

June 28 0.4 ( 0.1a 0.5 ( 0.2a 0.5 ( 0.2a 0.5 ( 0.2a

July 21 0.7 ( 0.0a 1.7 ( 0.6a 1.1 ( 0.0a 1.3 ( 0.1a

Sept 1 6.1 ( 0.8b 8.8 ( 0.2a 7.2 ( 0.3ab 2.2 ( 0.2c

Sept 15 16.0 ( 0.3a 16.4 ( 1.8a 10.9 ( 0.7a 1.7 ( 0.4b

Oct 19 23.2 ( 0.1b 37.2 ( 4.5a 22.6 ( 1.4b 1.8 ( 0.3cb

Aug 24 4.6 ( 0.0b 6.4 ( 0.3b 5.2 ( 1.7b 16.6 ( 0.9a

Sept 15 18.8 ( 1.4b 20.1 ( 1.2ab 13.6 ( 1.3b 26.6 ( 0.4a

Oct 18 22.1 ( 2.5bc 32.8 ( 0.4b 16.6 ( 3.0c 70.0 ( 0.9a

Aug 9 3.1 ( 0.6a 3.3 ( 1.0a 2.7 ( 0.0a 5.7 ( 0.6a

Sept 13 13.7 ( 2.8b 17.3 ( 1.5b 16.0 ( 0.4b 28.3 ( 1.5a

Oct 16 32.0 ( 1.1a 28.8 ( 0.0a 33.3 ( 2.7a 33.9 ( 4.5a

Values represent means ( SE (n ) 2). Means for a given year in the same column that are followed by the same letter are not significantly different (P < 0.05). b Date of harvesting of the Japonica food cultivar Akitakomachi was September 29. I, Indica; IJ, Indica-Japonica; J, Japonica. Shoot Zn concentrations from transplanting to harvesting in the Indica-type rice cultivars capable of accumulating Cd at high levels and in the Japonica food rice cultivar Akitakomachi from 2004 to 2006 are shown in Table S6). a

occluded Cd fraction, and the residual Cd fraction. Total soil Cd equaled the sum of these five fractions. These extraction methods are described in more detail in Table S4. The harvested shoots and roots of the Indica-type rice cultivars capable of accumulating Cd at high levels and the harvested shoots of the Japonica food rice cultivar Akitakomachi were washed with tap water to remove soil particles and then rinsed with distilled water. The harvested unhulled grains of the Japonica food rice cultivar Yumesayaka were air-dried for 1 week and then hulled with a FC2K huller (Otake Seisakusyo Co., Ltd., Ama, Aichi, Japan). The washed rice shoots and roots and the hulled rice grains were dried at 65 °C for 2 days and then their DWs were determined. The samples were then ground in a stainless-steel mill (Wonder Blender, Osaka Chemical, Osaka, Japan) for the Cd analysis. A 0.5-g portion of each ground sample was digested with 10 mL of 60% HNO3 in a heating digester (DK 20, VELP Scientifica, Milan, Italy). Plant and soil extracts were filtered through disposable 0.2-µm PTFE syringe filters (DISMIC-25HP, Advantec, Tokyo, Japan). We determined the Cd concentrations in these extracts by means of inductively coupled plasma-optical emission spectroscopy (ICP-OES; Vista-Pro, Varian, Mulgrave, Victoria, Australia). Certified reference materials for plants (NIES CRM No. 1 “Pepper Bush” and NIES CRM No. 10 “Rice Flour - Unpolished”, National Institute for Environmental Studies, Tsukuba, Japan) and for soil (NDG-7, Fujihira Industry Co., Ltd., Tokyo, Japan) were included in the analysis. The recoveries of metals were within the certified limits for the reference materials. Statistical Analysis. Statistical analysis was performed with JMP ver. 7.0.1 software (SAS Institute Japan, Tokyo, Japan). Treatments were compared by Tukey-Kramer’s HSD test. We analyzed the correlations between the Cd concentrations in the seven soil fractions and total in soils from the five treatments (no plants, plus the four Indica-type rice cultivars) and the corresponding grain Cd concentrations in the Japonica food rice cultivar Yumesayaka. To derive regression equations for predicting the grain Cd concentrations of Yumesayaka from the soil Cd concentrations, we performed regression analyses. The performance of each regression equation was evaluated on the basis of the t-values 5880

9


of the regression coefficients, the coefficient of determination (R2), the root-mean-square error (RMSE), and the F-value. The significances of the t- and F-values were determined by the probability of each value (P > F, P > |t|).

Results and Discussion Phytoextraction. The average shoot DW of each Indica-type rice cultivar except IR8 at harvesting from 2004 to 2006 was nearly 8 Mg ha-1 cultivation-1 (Table S5); this was more than 10 times that of the hyperaccumulator T. caerulescens (0.7 Mg ha-1 cultivation-1; (23)). The paddy field was maintained under flooded conditions for 2 months after transplanting to permit sufficient increase in the DW of the rice shoots (17). Thereafter, the field was drained so that the soil remained under oxidizing conditions until harvesting, because the solubility of soil Cd is higher under these conditions than under reducing conditions (Table S2; (16)). The shoot Cd concentrations of the four Indica-type rice cultivars increased after drainage in late July (Table 1), indicating a successful increase in Cd availability under oxidizing conditions. In contrast, the shoot Cd concentration of the Japonica food rice cultivar Akitakomachi remained low (no more than 2.2 mg kg-1) even after drainage in late July. In 2004, shoot Cd uptake by the Japonica Akitakomachi (23 g ha-1, Figure 1) was very low compared with that by the Indica Moretsu (358 g ha-1). Shoot Cd uptakes by Japonica rice cultivars are lower than those by Indica-type rice cultivars (12), suggesting that Japonica rice cultivars are unsuitable for phytoextraction. In 2005 and 2006, shoot Cd uptake was highest in the Indica Chokoukoku (550 and 333 g ha-1, respectively). The total shoot Cd uptake by the Indica Chokoukoku grown for 2 years (883 g ha-1) was higher than those by the 3-year grown Indicatypes Moretsu (869 g ha-1), Milyang 23 (638 g ha-1), and IR8 (532 g ha-1). This 2-year shoot Cd uptake by Chokoukoku from soil containing 1.63 mg kg-1 of total Cd was higher than the uptake by the hyperaccumulator T. caerulescens (540 g ha-1 after 3 years of cultivation in soil with a total Cd content of 2.8 mg kg-1; (23)), by willow Salix viminalis (170 g ha-1 after 5 years of cultivation in soil with a total Cd content of 2.5 mg kg-1; (24)), and by the poplar (Populus) clone Balsam Spire (57 g ha-1 after 2 years of cultivation in soil with a total Cd content of 0.75 mg kg-1; (25)). In contrast, Cd uptake by


FIGURE 1. Shoot and root Cd uptakes by Indica-type rice cultivars capable of accumulating Cd at high levels, and shoot Cd uptake by the Japonica food rice cultivar Akitakomachi. Means in the same year (shoot or root) labeled with the same letter (a, b, c) do not differ significantly (P < 0.05). J, Japonica; I, Indica; IJ, Indica-Japonica. Shoots were harvested in mid-October from 2004 to 2006. Residual roots were sampled in early May 2007. Shoot dry weights from transplanting to harvesting in the Indica-type rice cultivars capable of accumulating Cd at high levels and in the Japonica food rice cultivar Akitakomachi from 2004 to 2006 are shown in Table S5 in the Supporting Information.

FIGURE 2. Relationships among Cd concentration in the exchangeable fraction (Ex), Mehlich 3-extractable fraction (Me 3), 0.1 mol L-1 HCl-extractable fraction (HCl), and total Cd in subplot soils of the five treatments (no plants, and four Indica-type rice cultivars capable of accumulating Cd at high levels), sampled before plowing in May 2007, and the corresponding grain Cd concentrations in a Japonica food rice cultivar, Yumesayaka. Soil Cd concentrations in the fractions obtained by means of Mehlich 3 extraction and 0.1 mol L-1 HCl extraction for each subplot, sampled before plowing in May from 2004 to 2007, are shown in Table S9 of the Supporting Information). the residual roots of the Indica Chokoukoku was lower than those of the other Indica-type rice cultivars (Figure 1). Cadmium in the residual roots may be released gradually into the soil as the roots are decomposed by soil organisms. Because phytoextraction involves harvesting of plant shoots that have taken up toxic elements from the soil and removing harvestable material from contaminated fields, plants such as the Indica Chokoukoku, with high shoot Cd uptake and

low root Cd uptake, are ideal for phytoextraction. Moreover, the rice plant can be cultivated continuously (26). The shoot DWs of the four Indica-type rice cultivars did not decrease, even after two or three continuous cultivations without irrigation after drainage (Table S5), indicating that growth damage from continuous cultivation and the presence of toxic metals in the soil did not occur. This characteristic of rice is also useful for phytoextraction. VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5881

TABLE 2. Soil Cd Concentrations (mg kg-1) in the Fractions Obtained by Means of Sequential Extraction, and Their Totals for Each Subplot, Sampled before Plowing in 2007 sequential-extractable fraction plot

exchangeable

inorganically bound

organically bound

oxide occluded

residual

total

no plants IR8 Moretsu Milyang 23 Chokoukoku

0.62 ( 0.02a 0.52 ( 0.04ab 0.43 ( 0.03bc 0.50 ( 0.00abc 0.37 ( 0.01c

0.44 ( 0.03a 0.32 ( 0.02b 0.30 ( 0.01b 0.32 ( 0.00b 0.25 ( 0.01b

0.39 ( 0.00a 0.27 ( 0.04b 0.24 ( 0.01b 0.27 ( 0.02b 0.23 ( 0.00b

0.11 ( 0.01a 0.08 ( 0.01a 0.09 ( 0.00a 0.10 ( 0.01a 0.08 ( 0.01a

0.07 ( 0.01a 0.07 ( 0.01a 0.05 ( 0.00a 0.06 ( 0.01a 0.05 ( 0.00a

1.63 ( 0.01a 1.26 ( 0.04b 1.11 ( 0.03cd 1.24 ( 0.02bc 0.99 ( 0.02d

a


a Values represent means ( SE (n ) 2). Means in the same column that are followed by the same letter are not significantly different (P < 0.05). Soil Cd concentrations in the fractions obtained by means of sequential extraction and in the total for each subplot, sampled before plowing in May from 2004 to 2006, are shown in Table S7 of the Supporting Information.

TABLE 3. Grain Yields and Grain Cd Concentrations of Yumesayaka, a Japonica Food Cultivar Grown after Phytoextraction Using Indica-Type Rice Cultivars Capable of Accumulating Cd at High Levels plot

grain yield of Yumesayaka (Mg ha-1)

grain Cd concentration of Yumesayaka (mg kg-1)

no plants IR8 Moretsu Milyang 23 Chokoukoku

4.9 ( 0.3aa 5.4 ( 0.1a 4.9 ( 0.2a 4.9 ( 0.1a 5.4 ( 0.4a

1.02 ( 0.07a 0.89 ( 0.00ab 0.74 ( 0.01bc 0.86 ( 0.05ab 0.54 ( 0.05c

a Values represent means ( SE (n ) 2). Means in the same column that are followed by the same letter are not significantly different (P < 0.05).

Soil Cd Decrease after Phytoextraction. The exchangeable, inorganically bound, organically bound, and total soil Cd concentrations were lowest in the Indica Chokoukoku subplot, despite the fact that this cultivar was grown for only 2 years (Table 2). This suggests that this cultivar can take up Cd more efficiently than the other Indica-type rice cultivars from the more resistant (inorganically and organically bound) fractions, as well as from the more bioavailable (exchangeable) fraction. This uptake capability equaled that of the hyperaccumulator T. caerulescens when pot-grown (27). The Cd uptake by the residual roots of the Indica Chokoukoku (29.5 g ha-1, Figure 1) corresponded to only 0.02 mg kg-1 of soil Cd. Even allowing for the return of this root Cd to the soil by microbial decomposition, the total soil Cd concentration in the Chokoukoku subplot was 38% less than the mean value in the subplots with no plants (a reduction from 1.63 to 1.01 mg kg-1, Table 2). This reduction in total soil Cd concentration by the 2-year grown Indica Chokoukoku was higher than the reduction by 3-year grown hyperaccumulator T. caerulescens (by 15% of total soil Cd, assuming that this plant took up Cd from soil to a depth of 15 cm and with a bulk density of 0.85 Mg m-3; (23)). Grain Cd Decrease in Food Rice Grown after Phytoextraction. The Japonica food rice cultivar Yumesayaka grown after phytoextraction by the four Indica-type rice cultivars and in the subplots without phytoextraction showed normal growth. The average of the grain yields of Yumesayaka grown in the four subplots after phytoextraction and in the no plant subplot (5.1 Mg ha-1, Table 3) was similar to that of Japonica food rice cultivars in Japan in 2007 (5.2 Mg ha-1; (6)). The grain Cd concentrations of Yumesayaka grown after 2 years of phytoextraction with the Indica Chokoukoku were reduced by 47% (to 0.54 mg kg-1) of that of the same rice cultivar grown without phytoextraction (1.02 mg kg-1; Table 3). Prediction of Grain Cd from Soil Cd. For the various soil fractions the order of the significant coefficients of correlation 5882

9


between soil Cd concentration and the corresponding grain Cd concentration in the Japonica food rice Yumesayaka was as follows: exchangeable > Mehlich 3-extractable > total > 0.1 mol L-1 HCl-extractable > inorganically bound (Table S8). Soil Cd concentrations in the more resistant fractions (organically bound, oxide occluded, and residual) had little effect on the grain Cd concentrations of the Japonica food rice. The scatter plot of grain Cd against the top four soil-Cd variables gave a nonlinear curve with a downward concavity (Figure 2). Linear, logarithmic, square root, reciprocal, and exponential functions were thus employed. The estimated parameters of the logarithmic and reciprocal functions were statistically significant for all four variables of soil Cd (tvalues in Table S10). Among 20 functions, the reciprocal function of the exchangeable soil Cd fraction had the highest R2 value (0.994) and F-value (520.41) and the smallest RMSE value (0.016; Table S10). The results of these analyses suggest that the reciprocal function of the exchangeable soil Cd fraction is the best function for predicting the grain Cd concentration of Japonica food rice. Phytoextraction by High-Cd-Accumulating Rice. Recently, phytoextraction has been criticized by several researchers because of the long period required for restoration, the difficulty of producing a high-biomass crop of the desired species, and the lack of knowledge of agronomic practices and management (28, 29). The results of our research should help to dispel these criticisms. The DW of, and Cd uptake by, the Indica rice Chokoukoku were higher than those in the hyperaccumulator T. caerulescens (Figure 1; (23)). Paddy rice can be cultivated continuously (26), and its cultivation system is well-integrated and highly mechanized. The Indica rice Chokoukoku was managed by agricultural techniques familiar to farmers who grow Japonica food rice; it is therefore well suited to planting on a wide scale. The 2-year phytoextraction using Chokoukoku without irrigation after drainage reduced the total soil Cd concentration by 38%, and it reduced the Cd concentration in the grain of subsequently grown Japonica food rice by 47% without decreasing yield. However, the grain Cd concentration of the Japonica food rice was still above the Codex Alimentarius Commission’s international standard for the Cd content of rice grain (0.4 mg kg-1; (4)). Although our study showed the shoot Cd uptake by the Indica IR8 was lower than that by the Indica Chokoukoku (Figure 1), 3-year phytoextraction by Indica IR8 on a paddy field reduced the Cd concentration in soil Cd extracted with 0.1 mol L-1 HCl from 0.48 to 0.33 mg kg-1 and the Cd concentration in the grain of subsequently grown Japonica food rice to 0.11 mg kg-1 (Honma et al., in press). Even if the rate of reduction of soil Cd concentration by phytoextraction with Chokoukoku were to become half of that in the first 2 years, an additional 2 years of phytoextraction by Chokoukoku would reduce the grain Cd concentration of Japonica food rice Yumesayaka to below 0.4 mg kg-1.


These results suggest that phytoextraction with the Indica rice cultivar Chokoukoku can remove Cd from paddy fields polluted with Cd at low to moderate levels and can reduce the grain Cd concentration of the Japonica food rice cultivar Yumesayaka to below the Codex standard within a reasonable time frame. However, a potential hazard is inadvertent use of the phytoextractor grain as a food for humans and domestic animals. Large numbers of people were poisoned in Iraq in the early 1970s when mercury-treated grain meant for seed was eaten in homemade bread; poisoning also occurred in the United States in 1969 when treated grain was fed to hogs whose meat was subsequently eaten (30, 31). Phytoextraction by Indica rice cultivars capable of accumulating Cd at high levels will be applicable to the remediation of paddy fields in Monsoon Asia that have low to moderate levels of Cd contamination, provided that careful attention is paid to disposal of the high-Cd rice. Use of the phytoextraction techniques described here will help reduce the risk of Cd contamination of rice from paddy fields.

Acknowledgments We thank Ms. M. Tozawa, Ms. A. Onuki, Mr. N. Imai, Mr. M. Setagawa, and Ms. K. Mori for their help in the greenhouse and the laboratory at NIAES. This work was supported by a Grant-in-Aid (Hazardous Chemicals) from the Ministry of Agriculture, Forestry, and Fisheries of Japan (HC-04, -05, -06 and -07-1160).

(10)

(11) (12) (13)

(14) (15) (16) (17) (18)

(19)

Note Added after ASAP Publication

(20)

Figure 1 and the Supporting Information were replaced in the version of this paper published ASAP July 2, 2009; the corrected version published ASAP July 8, 2009.

(21)

Supporting Information Available

(22)

Tables S1 to S10. This information is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Singh, B. R.; McLaughlin, M. J. Cadmium in soils and plants: summary and research perspective. In Cadmium in Soils and Plants; McLaughlin, M. J., Singh, B. R., Eds.; Kluwer Academic Publishing: Dordrecht, 1999; pp 257-267. (2) Kobayashi, J. Pollution by cadmium and the itai-itai disease in Japan. In Toxicity of Heavy Metals in the Environment; Oeheme, F. W., Ed.; Marcel Dekker: New York, 1978; pp 199-260. (3) Ministry of Agriculture Forestry and Fisheries of Japan. Website; Measures taken to improve agricultural land soil pollution; http://www.maff.go.jp/j/syouan/nouan/kome/k_cd/taisaku/ horitu.html (In Japanese). (4) Report of the 29th Session of the Codex Alimentarius Commission; ALINORM 06/29/41; Codex Alimentarius Commission: Rome, 2006. (5) Ministry of Agriculture Forestry and Fisheries of Japan. Website; Cadmium in Japanese crops; http://www.maff.go.jp/j/syouan/ nouan/kome/k_cd/kaisetu/gaiyo2/index.html (In Japanese). (6) Monthly Statistics of Agriculture, Forestry and Fisheries; Ministry of Agriculture Forestry and Fisheries of Japan: Tokyo, 2008. (7) Vangronsveld, J.; Cunningham, S. D. Introduction to the concepts. In Metal-Contaminated Soils; Vangronsveld, J., Cunningham, S. D., Eds.; Springer: Berlin, 1998; pp 1-15. (8) Brown, S. L.; Chaney, R. L.; Angle, J. S.; Baker, A. J. M. Zinc and cadmium uptake by hyperaccumulator Thlaspi caerulescens and metal tolerant Silene vulgaris grown on sludge amended soils. Environ. Sci. Technol. 1995, 29, 1581–1585. (9) Robinson, B. H.; Leblanc, M.; Petit, D.; Brooks, R. R.; Kirkman, J. H.; Gregg, P. E. H. The potential of Thlaspi caerulescens for

(23) (24) (25)

(26) (27) (28)

(29)

(30)

(31)

phytoremediation of contaminated soils. Plant Soil 1998, 203, 47–56. McGrath, S. P.; Dunham, S. J.; Correll, R. L. Potential for phytoextraction of zinc and cadmium from soils using hyperaccumulator plants. In Phytoremediation of Contaminated Soil and Water; Terry, N., Banuelos, G., Eds.; Lewis Publishers: Boca Raton, FL, 2000; pp 109-128. Murakami, M.; Ae, N.; Ishikawa, S. Phytoextraction of cadmium by rice (Oryza sativa L.), soybean (Glycine max (L.) Merr.), and maize (Zea mays L.). Environ. Pollut. 2007, 145, 96–103. Arao, T.; Ae, N. Genotypic variations in cadmium levels of rice grain. Soil Sci. Plant Nutr. 2003, 49, 473–479. Murakami, M.; Ae, N.; Ishikawa, S.; Ibaraki, T.; Ito, M. Phytoextraction by a high-Cd-accumulating rice: Reduction of Cd content of soybean seeds. Environ. Sci. Technol. 2008, 42, 6167– 6172. Ministry of the Environment of Japan. Website: Environmental Quality Standards for Water Pollution; http://www.env.go.jp/ en/water/wq/wp.pdf. Asami, T.; Kubota, M.; Minamisawa, K. Natural abundance of cadmium, antimony, bismuth and some other heavy metals in Japanese soils. Jpn. J. Soil Sci. Plant Nutr. 1988, 59, 197–199. Kabata-Pendias, A.; Pendias, H. Trace Elements in Soils and Plants; CRC Press: Boca Raton, FL, 2001. Takahashi, E. Effects of soil moisture on the uptake of silica by rice seedlings. J. Sci. Soil Manure, Jpn. 1974, 45, 591–596 (In Japanese, with English title). Brady, N. C.; Weil, R. R. Soil architecture and physical properties. In The Nature and Properties of Soils, 12th Edition; Brady, N. C., Weil, R. R., Eds.; Prentice Hall: Upper Saddle River, NJ, 1999; pp 117-170. Morita, S.; Suga, T.; Nemoto, K.; Yamazaki, K. Analysis on root system morphology using a root length density model. 2. Examples of the analysis. Jpn. J. Crop Sci. 1987, 56, 48–49. Amacher, M. C. Nickel, cadmium, and lead. In Methods of Soil Analysis. Part 3, Chemical Methods; Sparks, D. L., Ed.; SSSA and ASA: Madison, WI, 1996; pp 739-768. Mehlich, A. Mehlich 3 soil test extractant: A modification of Mehlich 2 extractant. Commun. Soil Sci. Plant Anal. 1984, 15, 1409–1416. Sadamoto, H.; Iimura, K.; Honna, T.; Yamamoto, S. Examination of fractionation of heavy metals in soils. Jpn. J. Soil Sci. Plant Nutr. 1994, 65, 645–653 (In Japanese, with English abstract). Hammer, D.; Keller, C. Phytoextraction of Cd and Zn with Thlaspi caerulescens in field trials. Soil Use Manage. 2003, 19, 144–149. Hammer, D.; Kayser, A.; Keller, C. Phytoextraction of Cd and Zn with Salix viminalis in field trials. Soil Use Manage. 2003, 19, 187–192. Laureysens, I.; De Temmerman, L.; Hastir, T.; Van Gysel, M.; Ceulemans, R. Clonal variation in heavy metal accumulation and biomass production in a poplar coppice culture. II. Vertical distribution and phytoextraction potential. Environ. Pollut. 2005, 133, 541–551. De Datta, S. K. Principles and Practices of Rice Production; Wiley: Singapore, 1981. Hammer, D.; Keller, C. Changes in the rhizosphere of metalaccumulating plants evidenced by chemical extractants. J. Environ. Qual. 2002, 31, 1561–1569. McGrath, S. P.; Lombi, E.; Gray, C. W.; Caille, N.; Dunham, S. J.; Zhao, F. J. Field evaluation of Cd and Zn phytoextraction potential by the hyperaccumulators Thlaspi caerulescens and Arabidopsis halleri. Environ. Pollut. 2006, 141, 115–125. Robinson, B.; Schulin, R.; Nowack, B.; Roulier, S.; Menon, M.; Clothier, B.; Green, S.; Mills, T. Phytoremediation for the management of metal flux in contaminated sites. For. Snow Landsc. Res. 2006, 80, 221–234. Bakir, F.; Damluji, S. F.; Aminzaki, L.; Murtadha, M.; Khalidi, A.; Alrawi, N. Y.; Tikriti, S.; Dhahir, H. I.; Clarkson, T. W.; Smith, J. C.; Doherty, R. A. Methylmercury poisoning in Iraq. Science 1973, 181, 230–241. Curley, A.; Sedlak, V. A.; Girling, E. F.; Hawk, R. E.; Barthel, W. F.; Pierce, P. E.; Likosky, W. H. Organic mercury identified as cause of poisoning in humans and hogs. Science 1971, 172, 65–67.

ES8036687

VOL. 43, NO. 15, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5883

Phytoextraction by Rice Capable of Accumulating Cd at High Levels

Recommend Documents