Introduction

Cadmium (Cd), one of the most toxic heavy metals, is listed in the top 10 of the 2011 priority hazardous substances by the American Agency for Toxic Substance and Disease Registry (ATSDR 2011). High levels of Cd can occur in soils either naturally as a consequence of Cd-rich parent materials or because of anthropogenic activities such as mining, smelting, solid-waste disposal, application of phosphate fertilizers or sewage sludge, and atmospheric deposition (from waste incineration, fossil fuel combustion) (Assche 1998; Boularbah et al. 2006; Sun et al. 2008). Cadmium has relatively high mobility in the environment resulting in higher bioavailability for plant uptake than other metals. Consequently, Cd poses a greater threat than most toxic metals to the food chain (Kabata and Pendias 2001). A variety of studies have reported that food crops are capable of accumulating relatively high levels of Cd from soil (Chen et al. 1999; Cobb et al. 2000; Zhuang et al. 2009), such as amaranth (Chunilall et al. 2005; Fan and Zhou 2009; Li et al. 2012) and Chinese kale (Jinadasa et al. 1997; Moir and Thornton 1989; Tan et al. 2011). People who consume food crops grown in Cd contaminated soil are at risk of an elevated Cd exposure. Therefore, it is important to control the Cd content in crops to ensure food safety.

Different actions can be undertaken to reduce or eliminate the accumulation of Cd by plants. In contaminated farmlands, in situ immobilization of heavy metals using different soil amendments is a cost-effective strategy to alleviate soil heavy metal pollution (Lee et al. 2009; McGowen et al. 2001). Potentially effective amendments include alkaline substances (e.g., lime, fly ash, calcium carbonate, and manganese oxide), phosphates (e.g., diammonium phosphate, phosphate rock, hydroxyapatite), organic material (e.g., green manure, animal excrement, and peat), and other plant fertilizers (Sarwar et al. 2010). However, these amendments may not be effective in all cases. For example, neither lime (Bolan et al. 2003; Maier et al. 1997) nor phosphate (Hong et al. 2008; Tan et al. 2011) has a consistently positive effect on soil Cd immobilization. There is still a need for a range of more efficient and economical approaches for coping with metal toxicity in plants that may occur in large areas.

Several siliceous materials have been applied to in situ stabilization of Cd in soils, including sodium silicate (Feng et al. 2010; Nwugo and Huerta 2008a), potassium silicates (Shi et al. 2010; Zhang et al. 2008), and some silicon-rich materials like steel sludge and furnace slag (Chen et al. 2000; Gu et al. 2011). Numerous studies have demonstrated that Si application can enhance resistance and tolerance to Cd in graminaceous plants such as rice (Nwugo and Huerta 2008b) and maize (Vaculik et al. 2009) that are well known as Si-accumulators. The possible mechanisms for inhibition of Cd transport in Gramineae plants mediated by Si include (1) restrain the apoplasmic transport of Cd by depositing Si on the surface of the cell wall of epidermis and/or endodermis (Shi et al. 2010). Thickening of the Casparian strips and the cell wall of xylem and pericycle may occur after Si deposition in the endodermis (da Cunha amd do Nascimento 2009). Silicon might be attributed to the enhancement of root apoplasmic barrier development by accelerating suberin lamellae deposition and enhancing the tertiary endodermal cell walls formation (Lukacova et al. 2013; Vaculik et al. 2009); (2) precipitation of Si metal complex in the cytoplasm and vacuoles (da Cunha amd do Nascimento 2009). In contrast, less work has been done on the possible role of Si in dicots that are rather poor Si accumulators (less than 1 % of the dry weight) (Neumann and zur Nieden 2001; Treder and Cieslinski 2005).

In this study, we evaluated the effects of several silicates on Cd uptake by crops in comparison to traditional lime and phosphate amendments. Most studies emphasized the interaction of silicon and phosphorus on plants growth (Ma and Takahashi 1990; Rothbuhr and Scott 1957), but rarely studied their interaction with respect to Cd uptake by plants. In addition, there is lack of understanding of the optimal timing of application of amendments to the soil in relation to preventing excessive Cd uptake by plants (Treder and Cieslinski 2005). Three commonly grown dicotyledonous crops, including grain amaranth (Amaranthus hypochondriacus L. Cv. ‘K112’), red amaranth (Amaranthus tricolor L.) and Chinese kale (Brassica oleracea var. albiflora Kuntze), being the food source with high health risk posed by Cd were selected. The aims of this research are to: (1) investigate the effects of several low-cost silicates on Cd immobilization in Cd-contaminated soil and on reduction of Cd uptake by dicotyledonous crops, (2) identify the possible mechanisms involved in silicate-mediated inhibition of Cd uptake by dicotyledonous crops, (3) examine the effect of placement method (alone or in combination) and timing of application of silicates on Cd transfer from soil to plants, and (4) study Si and P interactions with respect to Cd plant uptake.

Materials and methods

Pot experiments

The soil was collected from the surface layer (0–20 cm) of a vegetable garden, near a waste landfill site in the suburb of Guangzhou, China. The soil was air-dried, crushed, mixed thoroughly, and sieved to 1 cm. Chemical properties of the soil were: soil pH 6.3, organic matter 4.7 %, cation exchange capacity (CEC) 13 cmol kg−1, available Si 82 mg kg−1, available P 122 mg kg−1, and total Cd 6.1 mg kg−1. Pot experiments were set up outdoors in the South China Botanical Garden (Guangzhou, China) beginning in January 2011. The experiment was coincident with the dry season.

  1. Experiment 1:

    Contaminated soil (7.5 kg per pot) was transferred into each of 84 plastic pots (35 cm diameter × 20 cm deep). Basic fertilizers were applied at a rate of 0.2 g kg−1 N and 0.2 g kg−1 K2O soil by adding 2.26 g urea and 3.22 g KNO3 per pot. The amendments included four silicates (talc, calcium silicate, sodium silicate and potassium silicate), phosphate, lime, and control. Non-amended treatment was used as the control. Talc, calcium silicate, and potassium silicate were added at the rate equivalent to the same Si content of sodium silicate. The doses of all soil amendments and timing of application are shown in Table 1. All the amendments were firstly ground into powder. The amendments were separately mixed with the soils to obtain homogeneity and were then equilibrated for 10 days with constant water status (80 % of field capacity). During the incubation period, the soils were thoroughly mixed every 3 days. After soil incubation, 20 seeds of crops (grain amaranth, red amaranth and Chinese kale) were initially sowed to each of the pots and later shinned to six uniform seedlings (2 cm high).

    Table 1 Treatments and material of different amendments used in pot experiment
    Full size table
  2. Experiment 2:

    This experiment was performed to study the effect of combined application of silicate (both sodium and potassium) with phosphate and timing of application of amendments on the inhibition of Cd uptake by plants. As shown in Table 1, the combined amendments were added at two different stages, including before sowing (the same as experiment 1) and after 30 days of growth. Grain amaranth and red amaranth were initially sown in each of the amended pots and later shinned to six uniform seedlings.

All treatments were replicated four times. All pots were kept outdoors and regularly watered to keep soil water holding capacity at a level between 70 and 90 %. The crops were harvested for analysis after a 60-day period of growth.

Sample analysis

Soil pH values were measured using a pH meter (Mettler Toledo FE20) with a water solid ratio of 2.5:1, while available soil Cd was determined using 0.1 M CaCl2 solution (McLaughlin et al. 2000) on the day before sowing.

The crops were harvested and separated into roots, stems, and leaves, and rinsed with distilled water. The samples were oven-dried for 72 h at 70 °C, weighed, and ground to pass a 100-mesh sieve. After digestion of the samples in HNO3-HClO4 (4:1), Cd concentrations in the plant digests were determined using flame atomic absorption spectrometry (FAAS, Hitachi Z-5300). Concentrations of Ca2+, Mg2+, K+, Cu2+, Zn2+, and Mn2+ in the digests were measured by inductively coupled plasma optical emission spectrometer (ICP-OES, Optima 2000). Silicon concentrations in the plants were determined by gravimetric method (Dong 1997). To ensure the precision of analytical procedures, a national standard plant material (poplar leaf GBW07604) was used and blanks were also included in digestion batches.

Statistical analysis

Data from plant and soil samples were statistically analyzed using one-way ANOVA at a significance level of p < 0.05 using SPSS 11.6 software. Duncan’s new multiple range test was used to detect any significant differences between means of different treatments. Simple correlation analysis and linear regression analysis were used to test the relation between soil pH and available Cd.

Results

The effect of single amendments on crop Cd uptake

Plant growth and biomass production

Plants did not show obvious Cd toxicity symptoms such as deformation or yellowing or leaf senescence. The effects of different amendment (talcum powder (TP), calcium silicate (CS), sodium silicate (SS), potassium silicate (PS), quicklime (QL), and potassium dihydrogen phosphate (PDP)) on biomass yields of grain amaranth, red amaranth, and Chinese kale are shown in Fig. 1. The results showed that, in most cases, application of TP, SS, and QL reduced the dry biomass of tested crops. Compared to the control (non-amended plants), addition of SS markedly reduced root and shoot dry weight of grain amaranth by 66 and 56 %, respectively. However, the root and shoot biomass of both red amaranth and Chinese kale showed no statistically significant differences between the SS treatment and the control. Treatment with CS increased biomass of the three crops, having an effect on biomass similar to that of the PDP treatment.

Fig. 1
figure 1

Dry weight of crops under different treatments. Error bars represent ±SE of quadruplicates. Bars with the same letter within a parameter are not significantly different (p > 0.05). CK non-amended treatment, TP talcum powder, CS calcium silicate, SS, sodium silicate, PS potassium silicate, QL quicklime, PDP potassium dihydrogen phosphate

Full size image

Cadmium accumulation in plants

The magnitude of amendment effects on Cd concentration in the three crops followed the order: SS > QL > CS ≥ PDP > PS ≈ CK ≥ TP (Fig. 2). Thus, SS was the most effective of the six amendments in reducing Cd in crops, with the average reductions in root, stem, and leaf concentrations of the three crops being 51, 53, and 72 %, respectively, when compared to the control crops (no amendment treated crops). The effect of QL on Cd uptake in grain amaranth was similar to that of the SS treatment, with the reduction in root, stem, and leaf concentrations being 52, 74, and 74 %, respectively. However, the relative Cd reductions in shoots of red amaranth and Chinese kale in response to the QL treatment were about half that of the SS amendment. The application of CS also decreased Cd concentration in the three tested crops as compared to the control, with the greatest reduction of 45 % in leaves of grain amaranth. Phosphate treatment (PDP) had similar effect to CS treatment on reducing shoot Cd concentrations, while the opposite trend was observed for roots.

Fig. 2
figure 2

Cadmium accumulation in crops under different treatments. Error bars represent ±SE of quadruplicates. CK non-amended treatment, TP talcum powder, CS calcium silicate, SS sodium silicate, PS potassium silicate, QL quicklime, PDP potassium dihydrogen phosphate

Full size image

Cadmium transfer factors in plants

The transfer factors of Cd from root to shoot (TF) of the three crops under different treatments are shown in Table 2. Results showed that all the TF values depended on treatments and crop species. Among the four treatments with silicates, SS and PS treatments greatly reduced TF of all the crops. For red amaranth, the TF was significantly lower (0.91) in the SS treatment than that in the control (2.5). Application of PS reduced the TF value of Chinese kale from 0.98 (control) to 0.44. The CS treatment reduced TF values for grain amaranth only. By contrast, the addition of TP did not significantly change the TF value of any of the tested crops. On the other hand, the effect of PDP was similar to that of PS. QL treatment affected TF differently in the three crops.

Table 2 The shoot/root Cd concentration ratios (TF) of Cd in plants
Full size table

Silicon accumulation in crop leaves

Grain amaranth and red amaranth accumulated much more Si in leaves than Chinese kale (Fig. 3). Silicon concentration in the crops was influenced by the type of amendments. Except for impacts of QL and CS treatment on Si in leaves of Chinese kale, application of the amendments increased Si concentration in the leaves of the crops. Addition of SS resulted in the highest Si concentration in the leaves of grain amaranth, which was 1.8-fold higher than that of the control. In Chinese kale, the highest Si concentration was observed with the PS amendment, which was 4.4-fold higher than that of the control. Treatment of PDP ranked first (in red amaranth) or second (in grain amaranth and Chinese kale) in increasing Si accumulation.

Fig. 3
figure 3

Silicon accumulation in leaves of crops under different treatments. Error bars represent ±SE of quadruplicates. CK non-amended treatment, TP talcum powder, CS calcium silicate, SS sodium silicate, PS potassium silicate, QL quicklime, PDP potassium dihydrogen phosphate

Full size image

Bioavailability of mineral nutrients and ratios of bioavailable nutrient concentrations to Cd

Table 3 shows the concentrations of six mineral nutrients (Ca, Mg, K, Cu, Zn, Mn) in leaves of grain amaranth and Chinese kale. Result showed that application of the amendments affected the bioavailability of mineral nutrients. However, macronutrients (Ca, Mg, K) were less affected by the amendments than micronutrients (Cu, Zn, Mn), except for the SS treatment. In most cases, added CS, SS, and QL into the soil significantly decreased Cu, Zn, and Mn uptake by crops. Meanwhile, Cu, Zn, and Mn concentrations were increased by PS treatment in Chinese kale but decreased in grain amaranth.

Table 3 Nutrient concentrations (Ca, Mg, K Cu, Zn, Mn) in leaves of crops
Full size table

Applications of amendments affected the concentration ratios of nutrients (Ca, Mg, K, Cu, Zn, Mn)/Cd in leaves of the crops (Fig. 4). Compared to the control, application of SS and QL amendments significantly increased Ca/Cd, Mg/Cd, K/Cd, and Cu/Cd ratios and decreased Zn/Cd and Mn/Cd ratios in leaves of the crops. Similar trends were obtained in the crops under CS amendment, but with smaller changes in Ca/Cd, Mg/Cd, K/Cd, Cu/Cd, and Zn/Cd ratios. On the other hand, both the applications of PS amendment and PDP amendment significantly increased the K/Cd and Mn/Cd ratios, while the TP treatment did not affect any of the concentration ratios of nutrients/Cd.

Fig. 4
figure 4

Concentration ratios of nutrients to Cd in leaves of grain amaranth and Chinese kale. CK non-amended treatment; TP talcum powder, CS calcium silicate, SS sodium silicate, PS potassium silicate, QL quicklime, PDP potassium dihydrogen phosphate

Full size image

The effect of placement method and timing of application

Plant growth and biomass

Figure 5 shows the effects of combined application of SS + PDP and PS + PDP on the biomass of grain amaranth and red amaranth. Soil amendment with SS + PDP applied before sowing had no significant effect on dry biomass of either crops, whereas PS + PDP significantly increased root and shoot dry weight of red amaranth and grain amaranth by 60–67 and 30–36 %, respectively, as compared to the control. After 30 days of growth, SS + PDP (30 days) applied to the soil decreased biomass of both crops, whereas PS + PDP (30 days) performed differently, increasing biomass of grain amaranth and decreasing biomass of red amaranth.

Fig. 5
figure 5

Dry weight of grain amaranth and red amaranth. Error bars represent ±SE of quadruplicates. CK non-amended treatment, SS + PDP SS and PDP composition added before sowing, PS + PDP PS and PDP composition added before sowing, SS + PDP (30 days) SS + PDP added in 30 days after sowing, PS + PDP (30 days) PS + PDP added in 30 days after sowing

Full size image

Cadmium concentration in grain amaranth and red amaranth

As shown in Fig. 6, the two types of combined amendments significantly reduced Cd concentration in the shoots of grain amaranth and red amaranth. Compared to the control, application of SS + PDP was more efficient in decreasing Cd concentration in the crops than PS + PDP, with average Cd reductions of 52 % for roots, 65 % for stems and 68 % for leaves. The application of SS + PDP (30 days) decreased Cd uptake by the crops, but the decrease was less than for the same amendments applied at the beginning. Application of PS + PDP (30 days) decreased Cd accumulation in shoots of the crops; again, the decrease was less than when the amendments were applied at the beginning.

Fig. 6
figure 6

Cadmium accumulation in grain amaranth and red amaranth. Error bars represent ±SE of quadruplicates. CK non-amended treatment; SS + PDP SS and PDP composition added before sowing, PS + PDP PS and PDP composition added before sowing, SS + PDP (30 day) SS + PDP added at 30 days after sowing, PS + PDP (30 days) PS + PDP added at 30 days after sowing

Full size image

Effects of amendments on soil pH and available Cd

The pH values and the concentrations of available Cd of the soils are shown in Table 4. Soil pH increased from 6.3 in the non-amended soil to 7.7 in QL-treated soil, while available Cd in the soil decreased from 1.7 for non-amended control to 0.31 for the QL treatment. Application of amendments significantly increased soil pH and decreased the concentration of available Cd, with the exception of PDP. One week after the amendments were added, soil pH values followed the order: QL ≈ SS > SS + PDP > CS ≥ PS ≈ PS + PDP > TP ≈ CK ≈ PDP. The soil concentration of available Cd followed the order: QL < SS < SS + PDP < CS ≤ PS ≈ TP < PS + PDP < CK.

Table 4 Soil pH, concentration of available Cd in the soil treated with different amendments after 1 week incubation without plants
Full size table

Correlation analysis indicated significant effects of soil pH on soil available Cd (r = −0.929, P < 0.001). The correlation between pH and available Cd can be described by the following regression equation (Fig. 7).

Fig. 7
figure 7

Correlation between the concentration of available Cd and soil pH

Full size image
$$ \left[\mathrm{Cd}\right]\mathrm{extractable}\kern0.5em =\kern0.5em 6.25-0.737\kern0.5em \mathrm{pH}\kern.2em \left({R}^2\kern0.5em =\kern0.5em 0.858,p\kern0.5em <\kern0.5em 0.001\right) $$
(1)

Discussion

It has been suggested that Si has a positive effect on growth and biomass of Si accumulators such as rice (Zhang et al. 2008) and maize (Liang et al. 2005) grown in both soil and hydroponic media. However, the results of the present study showed that Si application (except for CS (calcium silicate)) did not increase the biomass of any of the three tested dicotyledons species. This agrees with other studies showing no growth promoting effects in leafy vegetables (Chinese cabbage and lettuce) (Wang et al. 2012) and sweet basil (Putwattana et al. 2010) grown in contaminated soil. Moreover, the effect of amendments on crop yield differed with plant genus. Amaranthaceae crops, especially red amaranth, were more sensitive to the effects of the different amendments than Chinese kale (Fig. 1). Great differences in response to the application of Si have been observed not only between species but also within species. For example, Kulikova and Lux (2010) studied the effect of Si on five Zea mays L. hybrids, and found that growth promoting effects on shoots and roots in the Si + Cd treatment was only achieved for one hybrid. Thus, Si-mediated enhancement of plant biomass production is not a universal phenomenon in either Si accumulators or non-Si accumulators including dicotyledonous plants.

Different silicates used in this experiment showed various effects in reducing Cd accumulation by crops (Fig. 2). Among the four tested silicates, SS (sodium silicate) was the most effective amendment in decreasing Cd concentrations in plants, which agrees well with studies finding a decrease in both root and shoot Cd accumulation in rice using Na2Si3O7 treatment under hydroponic condition (Nwugo and Huerta 2008b) and in maize by adding Na2SiO3·9H2O to soil (Liang et al. 2005). Liming is a well-known and proven practice for controlling uptake of Cd by plants (Bolan and Duraisamy 2003). However, in this experiment, liming was not as effective as the SS treatment, with the decrease in Cd concentrations of red amaranth and Chinese kale by liming being about half that achieved by the SS treatment (Fig. 2). Calcium silicate had a somewhat smaller effect (16–43 %) than the Cd reduction (38–60 %) in grain and straw of rice reported by Li et al. (2008). Cadmium concentrations in the PS (potassium silicate) treated crops were consistent with reports that application of PS increased Cd retention in roots (endodermis and epidermis) while reducing Cd translocation to shoots in strawberry grown in soil (Treder and Cieslinski 2005) and pakchoi (Brassica chinensis L.) cultivated in hydroponics (Song et al. 2009).

An important finding of this study is that higher Si in crops did not always result in lower crops Cd concentration, indicating that greater Si accumulation in plants did not directly contribute to lower Cd accumulation. For example, the PS amendment resulted in the highest Si concentrations in leaves of Chinese kale, but did not decrease the Cd concentration in leaves and even increased the Cd concentrations in roots by two fold, as compared to the control (Fig. 3). We suggest that Si in the dicotyledonous crops played a greater role in alleviating metal stress than in restricting Cd transport from soil to plants. da Cunha and do Nascimento (2009) also found not only deposition of both silica and Cd in the cell wall of the epidermis, endodermis, pericycle, and xylem of roots and in mesophyll cell wall of leaves in maize but also higher Cd accumulation after Si addition to soil, suggesting that Si played a major physiological role in alleviating Cd stress and achieving cell detoxification in maize by co-precipitation of a Si metal complex.

Several studies have suggested that Si could restrict the transport of Cd from roots to shoots by retaining Cd in the root, a mechanism which may function not only in monocotyledon species such as rice (Zhang et al. 2008) and wheat (Rizwan et al. 2012) but also in dicotyledonous species such as peanut (Shi et al. 2010) and pak choi (Song et al. 2009). We found not only higher Si concentrations, but also lower TF values in PS or SS treated crops than in the control (Table 2). It is claimed that Si bound to the cell walls exhibits a high affinity for Cd (Wang et al. 2000). We suggest that Cd silicate precipitation may be promoted in root cells of the silicate-treated plants due to the relatively high Cd and silicate concentrations in the root solution. Furthermore, application of silicate may increase phosphate uptake by plants, as the addition of silicates to soil can increase phosphate desorption by competition (Lee et al. 2004; Roy et al. 1971) due to the similar chemical structures and properties of orthosilicic acid and orthophosphoric acid (Obihara and Russell 1972). This interaction of silicate and phosphate has occurred in our study, as increased Si concentrations were observed in leaves of the crops receiving PDP (potassium dihydrogen phosphate) treatments. High phosphate supply induces plants to form inositol phosphates which is then stored in the roots. Inositol phosphates complex strongly with heavy metals such as Cd, Cu, and Zn due to their anion charge (Persson et al. 1998; Turner et al. 2002), and thus may reduce the transport of Cd from root to plant top.

We thus conclude that the inhibitive effect of silicate on Cd uptake by crops is substantially indirect, that is, dependent on their initial effect on soil, including increased soil pH and introduction of relatively high concentrations of competing cations into soil solution. These changes in soil chemistry induce Cd adsorption and reduce Cd competitiveness for plant uptake.

Soil pH is considered a critical factor controlling the mobility of Cd in soils (Bolton and Evans 1996; Eriksson 1989; Li et al. 2008) and affecting plant Cd uptake (Singh and Myhr 1998). Results from this study showed that SS was the most effective in reducing Cd bioavailability compared to other silicates; this is likely due to a substantial soil pH increase (from 6.3 to 7.7) after the application of SS (Table 4). The correlation analysis between soil available Cd and soil pH (Equation 1) also showed that soil pH increase was a major reason for the reduction in soil available Cd, a result consistent with the reports of Chen et al. (2000) and Liang et al. (2005).

Soil pH changes could strongly affect available Cd and absorption of cations in the soil. Different metal ions have adsorption curves uniquely dependent on soil pH due to their different chemical properties (Gomes et al. 2001; Li 2001). Thus, modifying soil pH could also change the adsorption capacities of different metal ions and their concentrations in soil solution. Results showed that the application of amendments not only affected the bioavailable Cd, but also affected the bioavailability of mineral nutrients (Table 3). Plant nutrients are not only required for better plant growth and development, but also helpful to alleviate heavy metal stress. Significant change in ratios of bioavailable mineral nutrient concentrations to Cd (Ca/Cd, Mg/Cd, K/Cd, Cu/Cd, Zn/Cd, Mn/Cd) resulted from the soil amendments (Fig. 4). We suggest that changes in the bioavailable mineral nutrient status in the soil induced by modification of soil pH and exogenous addition of nutrients is another factor that directly affects Cd uptake by plants. Increasing the bioavailable nutrient/Cd ratios in soil (e.g., Ca/Cd, Mg/Cd, K/Cd and Cu/Cd) could result in lower Cd uptake by plants due to ion competition. Song et al. suggested that the main reason that Si reduced uptake and transport of Cd in maize and rice could be that Si enhanced uptake of Ca ions into plants (Song et al. 2009). In our study, much higher bioavailable Ca/Cd, Mg/Cd, K/Cd, and Cu/Cd ratios under SS or CS treatment were found, leading to suppression of Cd uptake by more cations competing for exchange sites with Cd ions at the root surface (Bolan et al. 2003). Although the bioavailable Zn/Cd and Mn/Cd ratios decreased in the silicate-amended soils, competition of these micronutrients with Cd is probably much weaker than that of the macronutrients.

Our results are consistent with other research showing that the combined application of silicate and phosphate to soil could promote plant growth as a consequence of the interaction of silicate and phosphate (Ma and Takahashi 1990). The combined application of SS and PDP amendment in this study was not only as effective as the sole application of SS amendment in decreasing Cd uptake by crops, but also solved the inhibitory effect on crop yield caused by the sole addition of SS amendment. The application of PS + PDP treatment also showed a significant effect in decreasing shoot Cd concentration and increasing biomass of crops: the combined treatment was more effective than the sole application of PS or PDP. The combined application of silicates and phosphate might be beneficial to nutrient balance by both having an inhibitory effect on Cd uptake and improving the fertilizer effect on crop growth. However, the application of silicates with phosphate (30 days after sowing) was not a more effective practice to reduce Cd uptake than application before sowing (Fig. 6). This phenomenon is probably caused by the facts that the more sensibility of crops to the effect of the amendments added after sowing (for 30 days) and full penetration and absorption among soil, crops, and amendments when Si were added 10 days before sowing. These results indicated that timing of application was one of the important factors influencing the effect on Cd uptake by plants.

Conclusion

The maximum reduction of Cd accumulation in plants was observed with the SS treatment. The application of SS combined with PDP before sowing not only decreased Cd uptake by crops but also overcame the inhibitory effect of SS on crop yield. Thus, optimum placement and timing of amendments can enhance Cd retention in soil and improve plant growth. Higher Si in crops did not resulted in lower Cd concentrations, however, application of SS and PS restricted Cd transport from root to shoot, suggesting that the beneficial role of Si may be related more to alleviation of metal stress than to the inhibition of Cd uptake by plants. The effect of the specific silicates on soil pH and concentrations of competitive cations in the soil solution are important in governing soil Cd availability in soil.