Pot experiments were performed in a greenhouse in Marechal Cândido Rondon, Paraná State, Brazil, (24° 31' S, 54° 01' W and 420 m asl), where the environmental conditions were as follows: minimum and maximum mean air temperature of 18 and 36 °C, respectively and mean air relative humidity of 65%.

Surface samples (0.0–0.20 m) from three representative soils of the western region of Paraná State were selected for Si fertilization studies (Table 1). Physical and chemical properties of the soils were determined by adopting standard procedures, and some properties are listed in Table 2. Soil pH in 0.01 mol L^–1 CaCl₂ solution was determined potentiometrically in a 1:2.5 (soil:solution) suspension by using a combined calomel reference glass electrode and pH meter. Organic matter was quantified by oxidation with potassium dichromate in the presence of sulfuric acid (ROSSET et al., 2014; ROSSET et al., 2016), followed by titration with ammonium Fe(II) sulfate (EMBRAPA, 2009). Available phosphorus (P), exchangeable potassium (K⁺), and cationic micronutrients (Cu⁺, Zn⁺, Fe⁺, and Mn⁺) were extracted using Mehlich-1 solution in a 1:10 (w:v) soil-to-extractant solution ratio (EMBRAPA, 2009, SARTO et al., 2011); P was determined using colorimetry at 725nm wave length and K⁺ and micronutrients were determined using atomic absorption spectrophotometry. Calcium (Ca²⁺) and magnesium (Mg²⁺) were extracted by 1 mol L–1 KCl solution and determined using atomic absorption spectrophotometry. Cation exchange capacity (CEC) was estimated using the summation method (CEC = H + Al⁺ + Ca²⁺ + Mg²⁺ + K⁺). Soluble Si was extracted by 0.5 mol L^–1 acetic acid solution in a 1:10 (w:v) soil-to-extractant solution ratio (KORNDÖRFER et al., 1999) and determined by beta molybdosilicic complex formation with a spectrophotometer at 660 nm wave length. Particle size analysis was performed using the pipette method (EMBRAPA, 2009), according to decantation speed of different soil particles after dispersion in 0.015 mol L^–1 (NaPO3) 6.NaO/1.0 mol L^–1 NaOH by overnight shaking.

Table 1
Brazilian soil classification, approximate equivalence to soil taxonomy and sampling site of the three soils from Paraná State, Brazil.

^† According to Embrapa (2013). ^†† USDA Soil Taxonomy (SOIL SURVEY STAFF, 2010).

Table 2
Some physical and chemical properties of the soils.

^† CEC: cation exchange capacity.

The experimental design was a 3 × 5 factorial in complete randomized blocks, with four replications. Treatments consisted of three soils—Ox1, Ox2, and Ult—and wheat plants grown using 0 (control), 1, 2, 4, and 6 t ha^–1 of CaSiO₃/MgSiO₃. The silicate source used was AgroSilício^® (10.5% Si, 25% Ca, 6% Mg, and 88% effective calcium carbonate equivalent (ECCE). The corrected soils were maintained for 15 days with water content at 60% field capacity. Then, the soils were placed in 8-L plastic pots and fertilized by applying 30 mg kg^–1 of N (urea), 80 mg kg^–1 of P (simple superphosphate), and 60 mg kg^–1 of K (KCl).

Five seeds of the wheat (Triticum aestivum L., cv. BRS Pardela) were sown, and nine days after seedling emergence, they were thinned to three plants per pot. The pots were irrigated daily to maintain soil moisture near at near field capacity. We sowed five wheat seeds per pot (variety: BRS Pardela). After nine days, we carried out the thinning process, leaving three plants per pot. The pots were watered daily in order to keep the soil moisture close to its field capacity. At 30 days after plant emergence, the application of 45 mg kg^–1 N as urea a solution was also performed.

Silicon absorption and nutrition of wheat

At flowering, the plants were separated into leaves and stems, dried in a forced air circulation oven for three days at 65 ± 2°C, ground in a Willye-type mill and then subjected to 3:1 nitric-perchloric acid digestion. The concentrations of K⁺, Ca²⁺, Mg²⁺, S, Cu²⁺, Fe²⁺, Mn²⁺ and Zn²⁺ in leaf were determined by flame atomic absorption spectrometry (Flame-AAS). Phosphorus was determined by colorimetry, Nitrogen by sulfuric acid digestion and steam drag distillation (EMBRAPA, 2009), and Si by digestion with hydrogen peroxide as well as sodium hydroxide and subsequently by colorimetry (KORNDÖRFER et al., 2004).

The level of silicon absorbed by the plants from the fertilizer was obtained by the difference between the silicon accumulated in the plant shoot and the silicon accumulated by the control treatment. The average recovery of silicon by plants was obtained by the division of the silicon absorbed from the fertilizer by the silicon applied x 100 (KORNDÖRFER et al., 1999).

Original data were analyzed using analysis of variance and regression analysis, and significant equations with the highest coefficients of determination (Tukey test, p ≤ 0.05) were adjusted. All analyses were performed using Saeg (1999) 8.0 software for Windows (Statistical Analysis Software, UFV, Viçosa, MG, BRA).

The application of calcium and magnesium silicate to the soil resulted in a significant increase in Si concentrations within wheat leaves and stems (Figure 1). When the application rate of silicate was increased from 0 to 6 t ha^-1, the silicon content within leaves increased from 27 to 36.4 (Ox1), 21.3 to 49 (Ult), and 22.6 to 33.5 (Ox2) g kg^-1 (Figure 1A). In plants grown in Ult soil, the concentration of Si was higher in leaves (approximately 49 g kg^-1) than in stems (approximately 20 g kg^-1). The increased concentration of Si in the shoots indicates that the silicate we used was reactive and effective in providing Si to the soil and the plant.

Korndörfer et al. (2010) conducted a study evaluating the effects of the surface application of calcium silicate (CaSiO₃; maximum rate: 2 t ha^-1) on Si concentrations in grass shoots. They reported that the Si concentration within the leaves of grass grown in soil treated with calcium silicate was twice that of grass grown in the control soil (no silicate). According to the authors, an increase in the concentration of Si in the shoots would indicate that the calcium silicate is reactive and efficient in providing Si to the soil and the plant.

Lima Filho and Tsai (2007) observed exponential silicon absorption in three cultivars of wheat and two cultivars of oat when adding Si to the nutrient solution at doses of up to 100 mg L^-1. The authors concluded that both grasses (wheat and oat) have high silicon uptake capacities, suggesting that large quantities of silicon can be absorbed when the availability of the element is increased in the substrate.

Figure 1
Effect of calcium and magnesium silicate rates on the content of silicon in the leaf (A) and stem (B) of wheat plants. ** Significant at 1%, * Significant at 5%, and ns: non-significant by the F test.

Sarto et al. (2016) investigated whether the application of CaSiO3 increases the Si content in Brachiaria and interferes with gas exchange and production components. They found that applying CaSiO₃ to the soil increases Si concentration in the leaves. Additionally, they found that Si reduces internal CO₂ concentrations and increases the efficiency of water use and instantaneous carboxylation efficiency. In soils with low pH and high levels of Al⁺³, Si helps alleviate the toxic effects of Al⁺³ by reducing sweating in Brachiaria, which in turn increases water use efficiency (SARTO et al., 2016).

In this study, we observed an increase in the silicon content within stems from 12.5 to 17.5, 8.8 to 20, and 10.6 to 16.5 g kg^-1 in plants grown in Ox1, Ult, and Ox2 soils, respectively (Figure 1B). Silicon deposition in the stem may promote increased resistance to lodging, which is an important feature for the mechanical harvesting of wheat. The selection of varieties with high Si absorption potential can minimize damping-off problems, reducing mechanical harvesting losses (MA; TAKAHASHI, 2002).

We found average silicon recovery rates of 28.20%, 60.61%, and 74.14% in wheat grown in Ox1, Ult, and Ox2, respectively (Table 3). Depending on the silicate application rates, Si recovery rates ranged from 19.68% to 48.38% in Ox1, 47.77% to 78.68% in Ult, and 42.08% to 128.27% in Ox2. Plants grown in Ult and Ox2 soils had higher Si recovery rates (60.61% and 74.14%, respectively). Therefore, more than half of the total Si absorbed by plants grown in Ult and Ox2 came from the silicate fertilizer.

Table 3
Silicon content available in the soil, Si accumulated in the wheat shoots, Si absorbed from the fertilizer and recovery rate, as a function of the calcium and magnesium silicate rates.

* Si absorbed from the fertilizer = Si accumulated in the shoots – Si accumulated by the control treatment** Recovery rate (%) = (Si absorbed from the fertilizer / Si applied) x 100 (KORNDÖRFER et al., 1999).

In a study on the effects of using calcium silicate as a source of Si for cultivating upland rice in four soils representing the Cerrado region (Brazil), Korndörfer et al. (1999) found that Si recovery was increased in soils with higher clay content. Clayey soils contain higher amounts of phyllosilicates (clay minerals that release Si⁴⁺ and Al³⁺) and higher amounts of Si in solution than sandy soils. Although sandy soils are rich in quartz (SiO₂), they are not a good source of Si because the chemical decomposition of quartz is difficult (DEMATTÊ et al., 2011).

In this study, we observed higher average Si recovery rates in one of the clayey soils (74.14% in Ox2) than in the sandy soil (60.61% in Ult), but lower Si recovery rates in the other clayey soil (28.20% in Ox1). The relatively high Si recovery rate in the sandy Ult soil may be explained by the conditions in which the experiment was conducted. The plants were watered daily, resulting in increased interaction between silicate particles and the solid phase of the soil. This, in turn, led to highly reactive silicate soils. According to Souza et al. (2008), soils with low levels of clay, organic matter, and water retention capacity show decreased interaction between silicate particles and the solid phase of the soil.

The concentrations of the nutrients N, P, S, Cu²⁺, and Fe²⁺ were not affected by the application of silicate to the soil (Table 4). These results were similar to those reported by Prado et al. (2002a), who studied the effects of applying limestone and steel slag at different rates on the absorption of nutrients in sugarcane plants. The authors found no effects on the concentrations of N, P, K⁺, Ca⁺², Mg⁺², and S in the plants. In this experiment, the concentrations of most nutrients measured within the leaves are in the range classified as medium and are suitable for the crop (EMBRAPA, 2009; RAIJ, 2011). However, N and S concentrations were below the adequate range, and Fe²⁺ and Mn²⁺ concentrations were above the adequate range.

Our results were also similar to those reported by Sarto et al. (2014a), who investigated the nutrition of wheat as a function of silicate fertilization. They observed that the concentrations of N, P, Mg⁺², S, Cu²⁺, and Fe²⁺ in wheat flag leaves were not affected by the application of calcium silicate to the soil.

Table 4
Average values of macro and micronutrients in the wheat leaf as a function of the increasing rates of calcium and magnesium silicate.

* Significant at 5% ** Significant at 1% by the F test.ns: non-significant at 5% by the F test.L: Linear.

The adequate range for N concentrations in plants is between 20 and 34 g kg^-1 (MALAVOLTA et al., 1997). In this study, plants grown in Ox1, Ult, and Ox2 soils showed average levels of 18.18, 18.34 and 23.12 mg kg^-1, respectively. The adequate range for S in plants is between 1.5 and 3.0 g kg^-1 (MALAVOLTA et al., 1997). In our experiment, plants showed mean values of 1.19, 1.11 and 1.15 g kg^-1 in Ox1, Ult, and Ox2 soils, respectively.

A linear model revealed that the content of K⁺ in leaves of plants grown in Ox2 soil decreased significantly with increasing rates of silicate application (Figure 2). According to Malavolta et al. (1997), cations K⁺ and Ca²⁺ cations compete for the same absorption sites within plant roots. The reduction in K⁺ content in the shoots of wheat grown in Ox2 soil may be due to K⁺ competing with calcium connected to silicate within the calcium and magnesium silicate (CaSiO₃/MgSiO₃) that we applied to the soil. Similarly, Sarto et al. (2014a) and Soratto et al. (2012) found that the application of silicon to wheat leaves increased the K⁺ concentration within flag leaves.

Figure 2
Effect of calcium and magnesium silicate rates on the K content of wheat leaf. * Significant at 5% and ** significant at 1% by the F test. ^ns non-significant at 5% by the F test.

Prado et al. (2003) evaluated the residual effects of steel slag on a yellow Oxisol cultivated with sugarcane and observed an increase in Ca²⁺ and Mg²⁺ concentrations at a depth of 0­–0.2 m. The increase in Ca²⁺ and Mg²⁺ observed in our study (Figure 3) may have resulted from the chemical composition of the fertilizer that we used. The greater availability of these nutrients for the plants from the calcium and magnesium silicate may explain the increased concentration of calcium and magnesium in wheat leaves (SARTO et al., 2015). Prado et al. (2002b) also observed an increase in the absorption of calcium and magnesium in the shoots of lettuce plants as a result of the application of steel slag, magnesium, and dolomitic limestone to an Oxisol. The authors reported that the correctors they used are a source of Ca²⁺ and Mg²⁺, elevating the Ca²⁺ and Mg²⁺ concentrations in the soil. This, in turn, influences the levels of these nutrients within plant shoot.

Figure 3
Effect of the rates of calcium and magnesium silicate on the content of Ca (A) and Mg (B) in the wheat leaf. * Significant at 5% and ** significant at 1% by the F test. ^ns non-significant at 5% by the F test.

There were significant differences between Cu²⁺, Zn²⁺, Fe²⁺, and Mn²⁺ concentrations in the leaves (Table 4). The Zn²⁺ and Mn²⁺ values decreased with increasing silicate application rates (Figure 4). These results are similar to those obtained by Moraes et al. (2009), when assessing the effects of calcium silicate and copper sulfate on nutrient contents in beans, the authors observed that increasing calcium silicate rates had no significant effect on the concentration of Cu²⁺, Fe²⁺, and Mn²⁺ within shoots. However, increasing calcium silicate rates led to a reduction in Zn²⁺. Increasing calcium silicate rates may lead to lower concentrations of Zn²⁺ and Mn²⁺ in wheat leaves because silicate use reduces the soil pH (SARTO et al., 2015). A reduced pH, in turn, reduces the bioavailability of these micronutrients in the soil.

Figure 4
Effect of the rates of calcium and magnesium silicate on the content of Zn (A) and Mn (B) in wheat leaf. * Significant at 5% and ** significant at 1% by the F test. ns non-significant at 5% by the F test.

The concentrations of Fe²⁺ and Mn²⁺ were above the adequate range for wheat crops. According to Embrapa (2009) and Raij (2011), the adequate range for Fe²⁺ is between 10 and 300 mg kg^-1. In our study, plants grown in Ox1, Ult, and Ox2 soils showed average Fe concentrations of 691.5, 466.7 and 438.37 mg kg^-1, respectively. The adequate range for Mn²⁺ is between 25 and 150 mg kg^-1 (EMBRAPA, 2009; RAIJ, 2011). In our study, plants grown in Ox1 and Ox2 soils showed average Fe concentrations of 196.11 (Ox1) and 403.06 (Ox2) mg kg^-1.

The calcium and magnesium silicate provided a better balance in the absorption of Mn²⁺ by wheat, reducing the amount of excessive Mn²⁺ absorbed by the plant (Table 4). According to Pozza et al. (2009), fertilizing rice plants with Si leads to increased oxidation of Fe²⁺ and Mn²⁺ on the root surface. Consequently, these nutrients precipitate and are not absorbed by the plant. The deposition of Si in plant leaves helps to improve the distribution of Mn²⁺ and prevent the toxic effects of Mn²⁺ within tissues (MA; TAKAHASHI, 2002).

The concentrations of Zn²⁺ within leaves were reduced from 37.5 to 11.17 (Ox1), 34.8 to 17.85 (Ult), and 48.12 to 22.07 (Ox2) mg kg^-1 when increasing calcium silicate rates from 0 to 6 t ha^-1 (Figure 4A). This reduction in leaf Zn²⁺ concentrations may have occurred because this nutrient is reduced in the presence of calcium silicate, which leads to competitive inhibition for the absorption of nutrients (MALAVOLTA et al., 1997). The concentrations of Mn²⁺ within leaves were reduced from 248.1 to 120.17 (Ox1) and 562.12 to 232.44 (Ox2) mg kg^-1 when increasing calcium silicate rates from 0 to 6 t ha^-1 (Figure 4B). Lima Filho and Tsai (2007) also observed decreases in Mn²⁺ and Zn²⁺ concentrations in wheat when supplying a nutrient solution containing Si. They attributed these decreases to the fact that dry matter accumulation increased faster than nutrient accumulation, resulting in a dilution effect for most nutrients studied.

Sarto et al. (2014a) investigated the effects of the application of calcium silicate on the nutrition and yield of wheat. They also observed a reduction in leaf Zn²⁺ and Mn²⁺ concentrations (reducing concentrations by 29% and 68%, respectively) when increasing calcium silicate rates from 0 (control) to 9.6 mg ha^–1. The decrease in leaf Zn²⁺ and Mn²⁺ concentrations with increasing calcium silicate rates may have occurred because the silicate they used reduced the soil pH (SARTO et al., 2015), thereby reducing the bioavailability of these micronutrients in the soil.

Zanão Júnior et al. (2010) studied rice grown in nutrient solutions with different rates of Mn²⁺ and Si. They observed an increase in the dry matter production of roots, petioles, and leaves with increasing Si. The authors also showed a reduction in the amount of Mn²⁺ translocated to the leaves when adding Si to the nutrient solution. Thus, adding Si to the nutrient solution helped reduce the negative effects of excess Mn²⁺ in the plants. Similarly, Horst and Marschner (1978) observed that the supplementation of 0.75 mg kg^-1 soluble silicon (Aerosil) leads to a reduction of Mn²⁺ levels in bean plants.

marcos__sarto@hotmail.com