Potassium availability for corn from silicate rock residues: Effect of source, particle size, and application method

ABSTRACT

The scarcity of soluble sources of potassium (K) in tropical regions, combined with the need for proper disposal of mining residue from silicate rocks, has increased the interest in using these materials as a source of K for crops. However, owing to their mineralogy, these residues have low solubility, and their efficiency needs to be determined. This study aimed to evaluate the agronomic efficiency of seven silicate rock mining residues as alternative sources of potassium (K) for corn compared with the soluble fertilizer potassium chloride (KCl). Three greenhouse experiments were conducted, varying the residue particle size (ultrafine, filler, and coarse) and application method (surface or incorporated). Potassium sources tested included residues of mica schist (RMS), syenite (RSY), granite (RGR), gold mining residue (RBR), basalts (RB1 and RB2), phonolite (RPH), and KCl, in addition to a control with no K. Two successive corn crops were grown, evaluating the production of shoot dry matter and the accumulation of K, P, Ca, Mg, Na, Zn, Cu, and Mn. The residues were inefficient in supplying K, regardless of the particle size and method of application. The RMS showed a significant plant response, particularly when finely ground and incorporated into the soil. However, because this material contains approximately 10 % Ca- and Mg-carbonates and also supplies other nutrients (Mg, Cu, Zn, and Mn), its effect cannot be unequivocally attributed to potassium release from silicate minerals alone, but rather likely reflects the combined influence of multiple mechanisms. The residues RB1, RB2, RBR, RGR, RSY, and RPH were not efficient under the conditions of this pot experiment. Despite its agronomic potential under controlled conditions, the application of RMS as an alternative K source requires caution regarding its economic viability and extrapolation to field conditions.

Keywords
remineralizer; potassium fertilization; rock powder; silicate agromineral; Zea mays

INTRODUCTION

Brazilian soils are often characterized by low potassium (K) availability, particularly in highly weathered systems, such as Ferralsols and Acrisols (Latossolos and Argissolos), although this condition can vary depending on factors such as parent material, mineralogy, texture, and management practices (Benites et al., 2010; Santos et al., 2025). These soils are generally associated with low K retention capacity and reduced natural fertility.

In Brazil, where cultivation intensity is notably high, K replenishment through fertilization is essential to maintain economically viable production levels (Gomides et al., 2023). Given the scarcity of local sources of soluble potassium, many tropical countries, including Brazil, have directed their efforts toward investigating alternative sources of K. In this context, agrominerals, products from the mining industry intended for use in fertilizer production or direct application to soil (van Straaten, 2022), have gained prominence in the literature. Among them, silicate agrominerals, especially aluminosilicates, stand out because of their wide and diverse distribution across Brazil (Silva et al., 2010). Silicate rock powder, composed mainly of aluminosilicates, is an abundant byproduct of mining activities and may contain considerable amounts of potassium in forms that are not very soluble in water. These materials have been recommended, particularly for alternative agricultural systems, where the use of soluble mineral fertilizers is restricted (Basak et al., 2017, 2023), but also for more intensive crops, such as grain crop rotations, where there is a high export of K from the areas through the harvested product (Soratto et al., 2021a).

However, the scientific literature presents divergent results regarding the agronomic effectiveness of aluminosilicates. Such discrepancies may be attributed to the mineralogical diversity of the evaluated materials and to variations in environmental conditions, especially soil potassium levels. Some studies have reported improvements in the physical and/or chemical properties of soil after their application (Priyono and Gilkes, 2008; Naderizadeh et al., 2010; Mancuso et al., 2014; Manning et al., 2017; Basak, 2019; Maity et al., 2019; Soratto et al., 2021b; Swoboda et al., 2022), although often at high doses and under greenhouse conditions. However, other studies question its effectiveness, highlighting the low solubility of the minerals present and, consequently, the limited release of K and its contribution to plant growth (Hinsinger et al., 1995; Coroneos et al., 1996; Bakken et al., 1997; Wallander and Wickman, 1999; Bolland and Baker, 2000; Schoen et al., 2016; Santana and Dias, 2019; Dias et al., 2025). In a systematic review conducted by Dias et al. (2025), covering 71 scientific articles, the authors concluded that silicate rock powders cannot efficiently replace soluble K sources and do not exhibit residual effects. However, current scientific evidence does not consistently support the commercialization of these residues as reliable sources of K for crops.

In Brazil, studies on ground silicate rocks as sources of nutrients and soil conditioners have intensified, especially after the regulation of the use of remineralizers by Law No. 12.890, of December 10, 2013 (Swoboda et al., 2022). In 2016, the Ministério da Agricultura Pecuária e Abastecimento (MAPA) published Normative Instruction No. 5, which establishes technical and legal criteria for the use of these materials as agricultural inputs.

Despite advances in research, there is still no consensus on the agronomic efficiency of silicate rock powders over the timescale of crop cultivation. Therefore, this study aimed to evaluate the effects of particle size and application method of silicate rock residues on the efficiency of potassium supply for corn.

MATERIALS AND METHODS

Characterization of residues

The silicate mining residues used in this study were collected from three different Brazilian states (Figure 1). Approximately 50 kg of each residue was sent to the Centro de Tecnologia Mineral (CETEM) in Rio de Janeiro (RJ), where it was oven-dried at 60 °C for 24 h. After drying, the materials were homogenized, and subsamples of approximately 2 kg were obtained using the conical quartering method to conduct particle size analysis.

Figure 1
Location of residue and soil sampling sites: mica schist (RMS), syenite (RSY), granite (RGR), basalt (RB1 and RB2), phonolite (RPH), by-product of the comminution of quartz veins occurring in shear zones or hosted in granite-gneiss rocks, originating from gold mining activity (RBR), and soil [LVA: Latossolo Vermelho-Amarelo (Oxisol)].

Granulometric analysis was performed using a Ro-Tap sieve shaker equipped with six sieves with openings of 4,800, 2,800, 2,000, 850, and 300 μm, equivalent to mesh sizes of 4, 7, 10, 20, and 50, respectively. The test was conducted with agitation at 1,200 rpm for 20 min. The masses of the fractions retained on each sieve were determined (Table S1, Supplementary Material).

The mineralogical composition of the residues was characterized by powder X-ray diffraction (XRD). Diffraction patterns were obtained using a Bruker D4 Endeavour diffractometer equipped with Co Kα radiation, operated at 35 kV and 40 mA. Data were collected over a 2θ range of 5–80°, with a step size of 0.02° 2θ and a counting time of 1 s per step. Mineral phases were qualitatively identified by matching the diffraction patterns against reference data available in the Bruker DiffracPlus database. Quantitative phase analysis was subsequently performed by Rietveld refinement using Bruker AXS TOPAS software, version 3.0 (Figure 2).

Figure 2
Minerals present in the samples were obtained using the full multiphase spectrum refinement method (Rietveld Method). Residues of mica schist (RMS), syenite (RSY), granite (RGR), barren rock from gold mining (RBR), basalt (RB1), basalt (RB2), and phonolite (RPH); * K-bearing mineral (different shades of green).

Elemental chemical compositions of the samples were determined by X-Ray Fluorescence (XRF) using a PanAnalytical AXIOS MAX instrument. The measurements were performed using a standardless method (semiquantitative analysis), and the results are presented in figure 3a. Samples were prepared by fusion using a 1:10 dilution (sample:flux) with a borate mixture (Li₂B₄O₇ and LiBO₂) supplied by Maxxifluxi, according to Sieber et al. (2004).

Figure 3
Percentages of major oxides (a) and elemental composition (b) of basalt residues (RB1:0.64 % K₂O; RB2: 1.18 % K₂O), gold mining residue rock (RBR: 0.92 % K₂O), granite (RGR: 1.94 % K₂O), mica schist (RMS: 4.00 % K₂O), phonolite (RPH: 8.10 % K₂O), and syenite (RSY: 12.64 % K₂O).

To determine the elemental composition of the residues, 10 g of each sample was digested with aqua regia, prepared by mixing hydrochloric acid (HCl) and nitric acid (HNO₃) in a 3:1 (v/v) ratio. The resulting solutions were analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES) to quantify K, Cu, Fe, Mn, Zn, P, and S (Figure 3b).

Soil and experimental environment characterization

The experiments were conducted in a greenhouse located in the experimental area of the Soil Department at the Universidade Federal Rural do Rio de Janeiro (UFRRJ) in the municipality of Seropédica, RJ (22° 45’ 33” S, 43° 41’ 50” W), at an altitude of 30 m.

A total of 200 kg of the surface layer (0.00–0.20 m) of a soil classified as Latossolo Vermelho-Amarelo (Oxisol) according to Santos et al. (2025), was collected in the municipality of Piraí, RJ (22° 35’ 55” S, 43° 56’ 25” W). After collection, the soil was air-dried, passed through a 4 mm sieve, and chemically and physically characterized as described by Teixeira et al. (2017). Soil texture corresponded to 481 g kg^-¹ clay, 114 g kg^-¹ silt, and 405 g kg^-¹ sand. Results of the chemical analysis were as follows: pH(H₂O): 4.4; organic matter content: 38 g dm^-³; P (Mehlich-1): 3.7 mg dm^-³; and SO₄²^-: 16 mg dm^-³. The contents of Na (Mehlich-1), K (Mehlich-1), Ca²⁺ (1 mol L^-1 KCl), and Mg²⁺ (1 mol L^-1 KCl) were 2.2, 1.1, 17, and 6 mmol_c dm^-³, respectively. Potential acidity (H+Al) was 59 mmol_c dm^-³; sum of bases (SB) was 26.4 mmol_c dm^-³; cation exchange capacity (CEC) was 85.4 mmol_c dm^-³; base saturation (V) was 31 %; and aluminum saturation (m) was 18.5 %. The micronutrient contents were as follows (mg dm^-³): B: 0.64; Cu: 0.9; Fe: 77; Mn: 18.5; Zn: 3.9; and Mo: 0.1.

Subsequently, the soil was incubated with limestone, following the methodology described by Dias et al. (2022) to raise the base saturation to 60 %. For this purpose, 104 g of dolomitic limestone (PRNT = 80 %) and 50 kg of soil were homogenized with 17 L of water. For 60 days, the mixture was incubated in polypropylene bags, and after pH stabilization, a new chemical characterization was performed. The results indicated a pH(H₂O) of 5.8, organic carbon content of 35 g dm^-³, P(Mehlich-1) of 2.3 mg dm^-³, and SO₄²^- of 25 mg dm^-3. The levels of Na⁺ (Mehlich-1), K (Mehlich-1), Ca²⁺ (1 mol L^-1 KCl), and Mg²⁺ (1 mol L^-1 KCl) were 1.3, 1.1, 30, and 12 mmol_c dm^-³, respectively, while the potential acidity was reduced to 29 mmol_c dm^-³. Sum of bases was 44.6 mmol_c dm^-³, and the total CEC was 73.6 mmol_c dm^-³, with a V% of 61 % and m% of 0 %. Regarding micronutrients, the levels observed after incubation, in mg dm^-3, were as follows: B = 0.64; Cu = 0.7; Fe = 86; Mn = 15; Zn = 3.5; and Mo = 0.1.

As described for the residues, the soil was subjected to X-ray diffraction analysis. After applying the total multiphase spectrum refinement method (Rietveld Method), the following results were obtained: 49.0 % quartz, 44.2 % kaolinite, 3.8 % gibbsite, 0.9 % microcline, 0.8 % anatase, and 0.3 % magnetite.

After characterization of the residues and soil, three simultaneous experiments were conducted to evaluate the efficiency of the residues as sources of K for corn plants grown in plastic pots. The experiments strictly followed the same methodology, varying only the particle size of the residues. The particle sizes were determined in accordance with Normative Instruction No. 5/2016 from the MAPA. The following were evaluated: (i) the coarse particle size allowed by the regulation (coarse or “farelado”, <4,800 µm); (ii) the finest particle size provided for in the legislation (filler, <300 µm); and (iii) a finer particle size than legally specified [<103 µm, based on Santos et al. (2016), referred to as ultrafine] to test whether further reduction in particle size could enhance the availability of K.

Experiment 1 - ultrafine granulometry

The first experiment was conducted with seven residues with particle sizes smaller than 103 µm. For this purpose, 1 kg of each sample, with particle sizes described in Table S1 (Supplementary Material), was micronized in a ball mill until 100 % of the material passed through a 103 µm (140 mesh) sieve. Eight potassium sources were evaluated: seven silicate rock residues (RMS, RSY, RGR, RBR, RB1, RB2, and RPH) and analytical-grade potassium chloride (KCl) used as a soluble reference. The treatments were applied to either the soil surface or incorporated into the soil.

The experiment was conducted in a randomized block design, in an 8 × 2 factorial scheme (sources × placement), with an additional treatment without K application. Three replicates were performed, totaling 17 treatments and 51 experimental units.

The applied K dose was set at 163 mg of K per kg of soil for all sources, equivalent to 196 mg of K₂O per kg of soil, based on total contents. This dose accounted for 90 % of the maximum dry matter yield of corn as a function of K fertilization (as KCl) in a preliminary study conducted under the same soil and environmental conditions (Dias et al., 2022). To supply the fixed K dose, 4.910 g kg^-1 of RMS, 1.554 g kg^-1 of RSY, 10.124 g kg^-1 of RGR, 21.349 g kg^-1 of RBR, 30.690 g kg^-1 of RB1, 16.645 g kg^-1 of RB2, 2.425 g kg^-1 of RPH, and 0.311 g kg^-1 of KCl were applied. Potassium doses were standardized based on total K content, in accordance with current regulatory criteria for material characterization. This approach was intentionally adopted to highlight that total K content alone may not be an adequate indicator of agronomic effectiveness for these materials. However, this approach represents a methodological limitation, as it does not account for differences in mineralogy, dissolution rates, and plant-available K among the evaluated materials, which may disadvantage low-solubility sources under short-term conditions.

Experiment 2 - filler granulometry

The second experiment followed a methodology similar to the first, except that the residues were ground until all samples had a particle size smaller than 300 µm (50 mesh), except for RPH. The original sample from this source already had a particle size below 103 µm, and no additional coarser material met the criterion for passing through a 300 µm sieve. The experimental design was a randomized block design in a 6 × 2 factorial scheme (six potassium sources and two application methods: surface and incorporated), yielding 12 treatments with three replicates each. The six sources evaluated were RMS, RSY, RGR, RBR, RB1, and RB2, applied in the two aforementioned forms and at the same rates as established in the first experiment.

Experiment 3 – coarse granulometry

The third experiment was conducted using a randomized block design, utilizing residues with particle sizes classified as “farelado”, according to Normative Instruction No. 5/2016 from MAPA (Brasil, 2016). The required particle size specifications were as follows: 100 % passing through a 4 mesh (4.8 mm) sieve, 10 % retained on a 7 mesh (2.8 mm) sieve, 67.5 % retained on a 20 mesh (0.84 mm) sieve, and 13.5 % passing through the latter. To meet these criteria, 1 kg of each residue with the particle size described in Table S1 (Supplementary Material) was sieved, and the coarser fractions were ground in a ball mill until the particle size distribution fully matched the regulatory specifications for each sample. No particles were discarded from any of the samples.

Five potassium sources (RMS, RSY, RGR, RBR, and RB1) were evaluated, applied either incorporated into the soil or applied on the surface, for a total of 10 treatments with three replicates. The RPH and RB2 sources were excluded from this experiment because their initial particle sizes were less than 103 and 300 µm, respectively, making it impossible to adjust them to the coarser particle size range.

Preparation of the experimental units

The experimental units consisted of 5 L plastic pots filled with 3 kg of previously corrected soil, air-dried, sieved through a 4 mm mesh and characterized. Before sowing, each pot received 0.100 L of a solution containing essential elements, excluding K. Through this solution, the following were supplied to the soil: 300 mg dm^-³ of P (as NH₄H₂PO₄), 80 mg dm^-³ of S [(NH₄)₂SO₄], 205 mg dm^-³ of N (NH₄H₂PO₄ and (NH₄)₂SO₄), 4 mg dm^-³ of Zn (ZnSO₄·7H₂O), 3.6 mg dm^-³ of Mn (MnCl₂·4H₂O), 1.5 mg dm^-³ of Fe (FeCl₃·6H₂O), 1.3 mg dm^-³ of Cu (CuSO₄·5H₂O), 0.8 mg dm^-³ of B (H₃BO₃), and 0.15 mg dm^-³ of Mo (Na₂MoO₄·2H₂O).

In treatments incorporating K sources, the soil and the respective residue were transferred to plastic bags, manually homogenized, and then placed in the pots. In treatments without incorporation, K sources were applied evenly to the soil surface after sowing.

Installation, conduction, and experimental collection

Eight seeds of the BM 3066 hybrid corn (Zea mays L.) were sown in each experimental unit. After sowing, 1.00 L of water was applied per pot, increasing the moisture content to approximately 70 % of the field capacity. Thinning was performed eight days after sowing, leaving two plants per pot. Soil moisture was controlled by daily weighing of the pots, with replacement of the water lost through evapotranspiration.

The experiment was conducted over 45 days. After this period, the plants were harvested by cutting the stalks near the soil. The samples were placed in paper bags and dried in a forced-air oven at 65 °C for 72 h, then weighed and ground in a Willey-type mill. Ground samples were digested using the wet method (Jones Jr, 2001). The concentrations of K and Na were determined using flame photometry, P using spectrophotometry, and Ca, Mg, Zn, Cu, and Mn using atomic absorption spectrophotometry (Silva, 2009). A soil sample (20 cm³) from each experimental unit was collected simultaneously with the aerial part sampling to determine the available K in the soil (Mehlich-1), as described by Teixeira et al. (2017).

Thirty days after harvesting the first crop, the second sowing was performed. During these 30 days, the soil moisture was maintained at 70 % of the field capacity. In the second crop, all procedures adopted in the first were repeated, including complete fertilization (without the addition of K or agrominerals), sowing, cultivation, harvesting, and laboratory analyses without reapplication of potassium sources. The shoot dry mass and nutrient accumulation data obtained for both crops were summed to calculate the total values accumulated over the two crop cycles.

Statistical analyses

The data were subjected to analysis of variance (ANOVA) after verifying the statistical assumptions at a 5 % significance level. The normality of the residuals was assessed using the Shapiro-Wilk test, the independence of errors using the Durbin-Watson test, and the homogeneity of variances using the Bartlett test.

ANOVA was conducted separately according to particle size. Initially, all treatments with particle size <103 µm were analyzed as a two-factor factorial, with an additional treatment (control) consisting of eight sources of K (first factor) and two application methods (second factor). When a significant difference was observed between treatments, the means were compared using the Tukey test at a 5 % significance level. In the second approach, each combination of K source, particle size, and application method was considered an individual treatment, including KCl applied by both methods and the control treatment. All ANOVA assumptions were verified, as specified above, to ensure the validity of the analysis. To compare differences between the K sources and the reference treatments (KCl and control), when ANOVA detected a significant difference, the Least Significant Difference (LSD) test was used at the 5 % significance level. Statistical analyses were performed separately for each crop and for the combined total of both crops.

In this study, treatment performance was classified based on statistical comparisons with the reference treatment (KCl) and the control (without K addition), using shoot dry matter (SDM) or potassium accumulation (Kac) as metrics. Materials were classified as “efficient” when their performance did not differ from or was superior to that of KCl. Materials were considered “inefficient” when their performance was statistically lower than KCl but did not differ from or was superior to the control.

To evaluate the similarity between treatments based on multiple variables, a hierarchical heatmap was constructed with simultaneous clustering of rows (treatments) and columns (variables). Means and standard errors were calculated for each treatment, and the data were standardized using Z-score normalization, based on the median and median absolute deviation (MAD). The standardized matrix was used to generate the heatmap, overlaying the actual values (mean ± standard error), and clustering treatments based on Euclidean distance and the complete linkage method. Treatments were grouped into three homogeneous classes, allowing integrated visualization of the relative performance of the potassium sources across the different evaluated variables.

Additionally, a multivariate approach using canonical discriminant analysis (CDA) was adopted to evaluate the separation of treatments based on the different sources and methods of K application, and to assess the correlation between variables across both crops. Biplots were constructed using the scores and centroids of the treatments, as well as the correlation coefficients of the original variables with the canonical axes.

Statistical analyses were performed in the R environment (R Development Core Team, 2025), with variance analysis, mean tests, and assumption checks conducted using the AgroR package (Shimizu et al., 2025). The candisc package was used for CDA (Friendly and Fox, 2024), and all graphs were created using the ggplot2 package (Wickham, 2016).

RESULTS

Dry matter production and K accumulation in the aerial part (ultrafine particle size)

Except for RMS, the residues showed low efficiency in sustaining dry matter production in corn plants over the two successive crops. The KCl source produced the highest shoot dry matter (SDM) in the first crop, regardless of the application method (Figure 4a). In the second crop, ultrafine RMS performed similarly to KCl with both application methods. Considering the accumulated SDM across the two crops, it was observed that, when applied to the surface, ultrafine RMS was as effective as KCl (Figure 4a). However, incorporating the RMS reduced its efficiency.

Figure 4
Dry matter (a) and potassium accumulation (b) in the aerial part of corn plants through the application of different potassium sources (KCl: potassium chloride; RB1 and RB2: basalt residues; RBR: residue from gold mining; RGR: granite residue; RMS: mica schist residue; RPH: phonolite residue; RSY: syenite residue), applied on the surface or incorporated into the soil, with residues having a particle size smaller than 103 µm (ultrafine). Identical letters do not differ from each other within the same application method and crop according to Tukey test (p<0.05). * indicates the highest mean within each source when there is a significant difference between application methods according to Tukey test (p<0.05). Vertical bars indicate the standard errors of the mean.

The residues were inefficient in supplying K to corn plants, except for RMS. Incorporation of ultrafine RMS significantly and positively influenced the accumulation of potassium in the aerial parts of corn plants, both in the first crop and in the accumulated values of the two crops (Figure 4b). Under these conditions, the incorporated RMS did not differ significantly from the KCl treatment in terms of K accumulation in either crop. This response may also reflect synergistic effects associated with the distinct chemical composition of RMS, which contains approximately 10 % Ca- and Mg-rich carbonate minerals with relatively higher solubility, as well as measurable concentrations of Zn and Cu, rather than being solely attributed to K release from silicate minerals.

Effect of particle size

When assessing the influence of particle size on the efficiency of the residues, it was observed that residues from silicate rocks performed predominantly worse than KCl in both crops, with responses often similar to the control; that is, most materials behaved as inert (Figure 5).

Figure 5
Shoot dry matter production and potassium accumulation in corn plants grown in soil treated with different sources of potassium, particle sizes (Coarse, Filler, and Ultrafine), and application methods (Incorporated and Surface), evaluated in the first and second crops, as well as in the total accumulated. The coloured bands represent the range of least significant difference (LSD, p<0.05) for comparison with the control treatment (purple band) and KCl (salmon band). Vertical bars indicate the standard error of the mean (n = 3).

Among the residues from the first cultivation, RMS incorporated into the soil as a filler promoted the highest shoot dry matter (SDM) production, although it was still lower than that obtained with KCl application (Figure 5). When incorporated into the soil, reducing the particle size of the filler to ultrafine (<103 µm) resulted in lower RMS efficiency in dry matter production. In this cultivation, the increase in K accumulation due to the application of the residues was mostly negligible, except for the RMS. Reducing the particle size of this residue to ultrafine and incorporating it resulted in K accumulation equivalent to that with KCl.

Syenite (RSY), granite (RGR), basalt (RB1 and RB2), and gold mining (RBR) residues showed similar or slightly superior performance compared with the control treatment in terms of dry matter production, regardless of application method or particle size (Figure 5). However, a slight trend of increased efficiency was observed for granite (RGR) with a reduction in particle size, although the values remained significantly lower than those obtained with KCl.

In the second cultivation, all residues across all particle sizes and application methods showed similar or slightly superior performance compared to the control, except for RMS (Figure 5). Only RMS, especially in the filler form, provided dry matter production and potassium accumulation comparable to those obtained with KCl application. Due to KCl high solubility, the amount of K exported in the first cultivation (average of 445 mg per pot) was close to the initially applied amount (489 mg per pot). Because there was no nutrient replenishment, the residual effect was reduced, especially regarding K accumulation, although it was still higher than that of all residues except RMS.

Considering the total accumulation in both crops, regardless of particle size or application method, the residues performed similarly to the control, except for the RMS (Figure 5). When incorporated into the soil as a filler or in an ultrafine form, RMS resulted in shoot dry matter production and K accumulation equivalent to those obtained with KCl application.

Impact of residues on nutrient accumulation

Hierarchical clustering revealed a clear separation between the treatments with KCl and RMS at the smallest particle sizes, which showed the highest positive Z-scores for most variables evaluated, especially for Ca, Zn, Mg, and K accumulation and dry matter production (SDM) (Figure 6). The RMS treatments, in this group, were particularly characterized by increased accumulation of Ca, Mg, Zn, and Cu, which clustered closely with shoot dry matter (SDM) and K accumulation (Kac), indicating a strong association between these nutrients and plant growth response. This pattern supports that the response associated with RMS is likely driven by a combined multi-nutrient effect rather than by K supply alone. An intermediate group was observed, consisting of RGR at the smallest particle sizes and RMS as a filler. The third group consisted of all other treatments, with the lowest Z-scores, comparable to the control treatment without K application. This pattern indicates low efficiency and limited nutrient release, irrespective of the particle size. The sources RB1, RB2, RBR, and RSY showed responses similar to that of the control for almost all variables, with no consistent evidence of increased absorption of K or other nutrients. These results confirm the greater efficiency of RMS compared to the other residues, although it was still lower than that of the soluble fertilizer (KCl).

Figure 6
Hierarchical heatmap of robust z-scores, representing the influence of different potassium sources, particle sizes, and application methods on dry matter production (SDM) and the accumulation of elements K (Kac), Na (NaAc), Ca (CaAc), Mg (MgAc), Fe (FeAc), Mn (MnAc), Zn (ZnAc), and Cu (CuAc) totalized over two successive corn cropping cycles. The colors of the cells indicate the standardized deviations (robust z-scores) in relation to the overall mean of each variable, with shades of red indicating values above the mean and shades of blue indicating values below the mean. Each cell shows the mean ± standard error of the respective variables. Treatments were hierarchically clustered according to their similarity in response variables. Groups: KCl, ochre; residues, green; and control, burgundy.

The reduction in particle size had a limited, non-systematic effect on the materials' performance. In the case of RGR, a slight tendency toward increased nutrient accumulation was observed as the particle size decreased. For RMS, reducing the particle size from coarse to filler resulted in increased dry matter production and nutrient accumulation.

In general, the pattern observed in figure 6 shows that the evaluated silicate residues performed similarly to the control, not promoting substantial increases in nutrient absorption or corn growth. Only RMS, and to a lesser extent RGR, demonstrated some potential for the release of potassium and elements such as Mg and Cu. The benefit of increased P accumulation resulting from the application of RMS in the filler form was also observed.

Canonical discriminant analysis differentiated between KCl and RMS at smaller particle sizes compared with the other residues (Figure 7). The rocks, except for RMS, showed considerable overlap among themselves, with little distinction between the application methods, indicating relatively similar behavior. It can be observed that, although KCl was associated with higher SDM and K accumulation in the first crop, RMS at the smaller particle sizes was mainly associated with higher accumulations of P, Mg, Zn, and Cu. This evidence suggests that the positive effect on SDM observed with RMS may be related to the supply of other nutrients or to a corrective effect that increased P availability.

Figure 7
Canonical discriminant analysis (CDA) applied to growth variables and nutrient accumulation in corn plants grown in the first and second crops, as a function of different potassium sources and application methods. Dispersion of K sources (KCl and silicate rock residues) according to the first two canonical axes (Can1 and Can2), applied on the surface (circles) or incorporated (triangles), and in different particle sizes (G1: ultrafine; G2: filler; and G3: coarse). Correlation of variables height, diameter, leaf area, shoot dry matter (SDM), and accumulation of nutrients such as K (Kac), Na (NaAc), Ca (CaAc), Mg (MgAc), Fe (FeAc), Mn (MnAc), Zn (ZnAc), and Cu (CuAc), with the canonical axes.

In the second cultivation, KCl and RMS at the smallest particle sizes stood out markedly from the other residues, with both associated with higher SDM values and K accumulation, whereas RMS was also mainly associated with higher Cu accumulation. The main distinction between the sources was related to shoot dry matter production and K accumulation in both cultivations.

Potassium available in the soil

Available K levels were very low across all treatments, and soil K was depleted during the first cultivation period. Under these conditions, the available soil K after the first cultivation was significantly higher with the application of RMS in the filler and ultrafine particle sizes, particularly when applied to the surface (Figure S1, Supplementary Material). At the end of the second cultivation, it was found that RMS applied on the surface, including the finer particle sizes, resulted in higher available soil K values than the other sources applied superficially (Figure S1, Supplementary Material). However, when potassium sources were incorporated, there were no significant differences among the treatments. The application of residue in coarse form did not affect the available soil K content in either cultivation (Figure S1, Supplementary Material).

DISCUSSION

With the exception of RMS, the residues exhibited inefficiency in supplying potassium to corn plants grown in pots. The comparison among materials based on equivalent total K inputs highlights an important limitation of using total K₂O content as an indicator of agronomic effectiveness. The results of this study clearly demonstrate that total K₂O is not a reliable predictor of plant-available K, as the material mineralogical composition strongly influences it. For instance, the syenite residue (RSY), which presented the highest total K₂O content among the evaluated sources, showed agronomic performance similar to the control treatment, indicating very limited K availability to plants under the tested conditions. This discrepancy reflects the low solubility of K-bearing minerals, such as K-feldspar, compared with more reactive mineral phases. Therefore, the effectiveness of these materials as K sources is governed not only by their total K content but primarily by their mineralogy and dissolution behavior, particularly in short-term assays.

Potassium associated with K-feldspar, known as alkali feldspar, has low bioavailability, even when finely ground (Bakken et al., 2000). Feldspathoids, such as nepheline, which constitutes approximately 14 % of the phonolite residue sample (RPH), are chemically similar to feldspars but present greater solubility due to structural differences, as the energy required to hydrolyze Al-O-Si bonds is lower than that required for Si-O-Si bonds (Lange et al., 2021). This provides a plausible explanation for the slightly higher SDM observed with RPH compared to RSY. However, even in this case, the plant growth was far inferior to that obtained with KCl.

Basalt powders, which are widely studied as multi-nutrient rock amendments but are typically poor in K-bearing minerals (Figure 2) and exhibit low solubility, have limited capacity to supply plant-available K under short-term conditions, contributing to their generally low agronomic effectiveness. In an evaluation of nutrient accumulation in corn grown successively in pure basalt powder as a substrate, Krahl et al. (2020) observed that the concentration of K in the dry matter of corn ranged from 2.6 to 15 g kg^-1 across seven crops. The authors noted that this concentration range is below the ideal range for corn growth (17.5 to 22.5 g kg^-1). In contrast, RMS was the only residue that demonstrated agronomic efficiency. The greatest accumulation of K derived from residues was observed in the smallest particle size fraction of RMS when incorporated into the soil. This behavior can be explained by its mineralogical composition, particularly the presence of biotite, which is more reactive and capable of releasing K under rhizosphere conditions (Basak et al., 2017; Krahl et al., 2022). In the rhizosphere, increased proton activity, organic acid exudation, and microbial processes promote mineral dissolution and the replacement of interlayer K by hydrogen (Basak, 2019).

These findings are consistent with previous studies. Mohammed et al. (2014) observed that biotite was more effective than nepheline syenite and microcline in supplying K, while microcline showed no difference from the control. The results obtained in the present study further highlight differences in agronomic efficiency among the evaluated sources.

Despite the improvement in agronomic efficiency provided by the comminution and soil incorporation of RMS, the average increase in dry matter compared with the control treatment, per gram of K source applied, was markedly higher for KCl than for the residues. Relative to the control, the incorporation of KCl into the soil increased SDM by 62 g for each gram of fertilizer applied per pot. Among the residues, the highest value obtained was only 4 g g^-1, corresponding to the incorporation of RMS as a filler. Thus, the KCl dose applied was approximately 16 times lower than that of the best-performing residue (RMS).

While mineralogical composition largely explains the intrinsic reactivity of the residues, physical factors also play a key role in controlling their behavior in soil. In this context, particle size, as well as mineralogical characteristics, played a critical role in residue reactivity. Smaller particles have larger specific surface areas, favoring biogeochemical reactions, particularly when residues are incorporated into the soil, increasing contact with the rhizosphere solution. It is expected that more reactive minerals release larger amounts of K when reduced to smaller particle sizes, whereas highly stable minerals such as potassium feldspars show only slight increases in reactivity as a result of comminution (Zhang et al., 2023), as observed with RSY. However, reducing particle size from filler to ultrafine did not improve corn dry matter production with RMS, indicating that increased reactivity does not necessarily translate into better plant performance.

Particle size played a dual role by controlling the specific surface area available for mineral dissolution and the degree of contact with the rhizosphere, thereby influencing K release. While finer particles enhanced reactivity, particularly for RMS, this effect was limited for minerals with inherently low solubility and did not consistently translate into greater plant growth, indicating that granulometry alone is insufficient to overcome mineralogical constraints.

The RMS increased dry matter production when applied to filler particles, achieving performance equivalent to that of KCl. However, K accumulation was significantly lower than that obtained with KCl application, suggesting that factors beyond K supply contributed to plant growth. The RMS contains approximately 10 % carbonates (calcite and dolomite) and significant levels of Mg, Cu, Zn, and Mn. Canonical discriminant analysis revealed increased accumulation of these nutrients and higher P accumulation (Figure 7). Carbonates may raise soil pH, reducing P adsorption and precipitation with Al and Fe, thereby enhancing P availability (Penn and Camberato, 2019). These interpretations are consistent with the observed responses but should be considered as plausible hypotheses, since no direct measurements were performed.

In a study conducted in pots with tea plants (Camellia sinensis L.), Pramanik and Kalita (2019) found that applying 0.5 g kg^-1 of mica residue to soil with approximately 70 mg kg^-1 of available K resulted in equivalent K accumulation and higher dry matter production compared with KCl applied at the same equivalent K dose. However, soil-available K was significantly higher with KCl application. According to the authors, the acidic soil and humid conditions of the tea-growing region favored the release of potassium from mica.

Dias et al. (2025) found that although rock powders are generally inefficient at supplying K to plants, they perform better in pot and greenhouse experiments than in field experiments. The authors further reported that pot experiments tend to overestimate potassium accumulation, showing significantly higher results than those obtained in field experiments when doses above the equivalent of 5 Mg ha^-1 are evaluated. Therefore, the efficiency observed with the RMS in the present study should be interpreted with caution.

Although the pot results indicate the agronomic efficiency of finely ground RMS, the economic implications of its large-scale use require careful attention. Studies have indicated that one of the main additional costs associated with rock powders is the transportation between the mineral source and agricultural areas (Kuhn et al., 2025). Distances of 100-500 km are already considered logistical barriers that can make competitive costs unfavorable compared with soluble fertilizers, in addition to their low efficiency.

In this context, regionally designed field studies that account for local soil and climatic conditions are essential before recommending the use of silicate rock residues as an alternative to soluble fertilizers. Such studies should address cost–benefit analysis, return on investment, K recovery under field conditions, as well as the costs of grinding, transport, and incorporation, and the persistence of effects over successive harvests.

Interpretation of the present results should be considered in light of several methodological limitations. This study was conducted under greenhouse conditions using a pot-based experimental system, which may not fully represent field environments in terms of soil heterogeneity, root development, and water dynamics. Factors such as soil spatial variability, climate, water availability, root density, and uniformity of application over large areas can greatly alter the observed efficiency (Desmalles et al., 2025).

In addition, only a single soil type and a single crop species were evaluated, limiting the generalizability of the findings across different pedoclimatic conditions. Although two successive crops were grown, the evaluation period was relatively short, potentially underestimating the long-term contribution of low-solubility materials. The number of replicates was also limited (n = 3), which may reduce the statistical power to detect subtle treatment effects. Furthermore, potassium doses were standardized based on total K content, following regulatory criteria, which does not necessarily reflect plant-available K and may disadvantage low-solubility materials in short-term assays.

The interpretation of the observed response for RMS is further limited by its complex chemical composition, including carbonate minerals and other nutrients, making it difficult to isolate the specific contribution of K supply. Finally, no direct measurements were performed to separate the effects of K release from potential liming effects or multi-nutrient interactions. Field studies with appropriate experimental designs are needed to evaluate aspects such as cost-benefit relationships, return on investment, K recovery under field conditions, the costs associated with grinding, transport, and incorporation, and the persistence of effects over successive cropping cycles.

CONCLUSIONS

The application of silicate rock residues as alternative sources of potassium were, for the most part, ineffective in supplying the nutrient to corn under greenhouse conditions. Reducing particle size and incorporating the material into the soil generally did not improve their performance.

Among the evaluated materials, only mica schist residue (RMS), containing approximately 10 % carbonates, provided a consistent agronomic response, particularly when finely ground and incorporated into the soil. However, this response cannot be unequivocally attributed solely to potassium release from silicate minerals.

The RMS dose was approximately 16 times higher than that of KCl, indicating the need for higher doses to achieve equivalent responses in terms of K accumulation and biomass production. The other residues (RB1, RB2, RBR, RGR, RSY, and RPH) are not efficient K sources under the conditions of this study, with responses generally similar to those of the control treatment.

Despite its agronomic potential under controlled conditions, the use of RMS as an alternative K source requires caution regarding economic viability and extrapolation to field conditions. Future studies should be designed to isolate the effects of K supply from other chemical contributions in order to assess its potential as a potassium source more conclusively.

ACKNOWLEDGEMENTS

The authors thank the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brazil (CAPES), the Embrapa, the Centro de Tecnologia Mineral (CETEM), the Programa de Pós-Graduação em Agronomia – Ciência do Solo of the Universidade Federal Rural do Rio de Janeiro (UFRRJ), and FNDCT/Finep/Rede FertBrasil for financial and logistical support.

How to cite:

Dias RC, Zonta E, Benites VM, Netto-Ferreira JB, Monte MBM, Veneu DM, Silva RSA, Teixeira PC. Potassium availability for corn from silicate rock residues: Effect of source, particle size, and application method. Rev Bras Cienc Solo. 2026;50:e0260022. https://doi.org/10.36783/18069657rbcs20260022

FUNDING

This research was funded by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brazil (CAPES) (Finance Code 001), Empresa Brasileira de Pesquisa Agropecuária (Embrapa) (Project SEG 20.20.06.001.00.00), and Fundo Nacional de Desenvolvimento Científico e Tecnológico (FNDCT), through the Financiadora de Estudos e Projetos (Finep), under the Rede FertBrasil (Agreement No. 01.22.0080.00, Ref. 1219/21).

SUPPLEMENTARY MATERIALS

Supplementary data to this article can be found online at https://www.rbcsjournal.org/wp-content/uploads/articles_xml/1806-9657-rbcs-50-e0260022/1806-9657-rbcs-50-e0260022-suppl01.pdf

SUPPLEMENTARY MATERIALS

DATA AVAILABILITY

The data will be provided upon request.

REFERENCES

• Bakken AK, Gautneb H, Myhr K. Plant available potassium in rocks and mine tailings with biotite, nepheline and K-feldspar as K-bearing minerals. Acta Agric Scand B Soil Plant Sci. 1997;47:129-34. https://doi.org/10.1080/09064719709362452
  » https://doi.org/10.1080/09064719709362452
• Bakken AK, Gautneb H, Sveistrup T, Myhr K. Crushed rocks and mine tailings applied as K fertilizers on grassland. Nutr Cycl Agroecosyst. 2000;56:53-7. https://doi.org/10.1023/A:1009709914578
  » https://doi.org/10.1023/A:1009709914578
• Basak BB, Sarkar B, Biswas DR, Sarkar S, Sanderson P, Naidu R. Bio-intervention of naturally occurring silicate minerals for alternative source of potassium: Challenges and opportunities. Adv Agron. 2017;141:115-45. https://doi.org/10.1016/bs.agron.2016.10.016
  » https://doi.org/10.1016/bs.agron.2016.10.016
• Basak BB, Sarkar B, Maity A, Chari MS, Banerjee A, Biswas DR. Low-grade silicate minerals as value-added natural potash fertilizer in deeply weathered tropical soil. Geoderma. 2023;433:116433. https://doi.org/10.1016/j.geoderma.2023.116433
  » https://doi.org/10.1016/j.geoderma.2023.116433
• Basak BB. Waste mica as alternative source of plant-available potassium: Evaluation of agronomic potential through chemical and biological methods. Nat Resour Res. 2019;28:953-65. https://doi.org/10.1007/s11053-018-9430-3
  » https://doi.org/10.1007/s11053-018-9430-3
• Benites VM, Carvalho MCS, Resende AV, Polidoro JC, Bernardi ACC, Oliveira FA. Potássio, cálcio e magnésio. In: Prochnow LI, Casarin V, Stipp SR, editors. Boas práticas para uso eficiente de fertilizantes. Vol. 2. Piracicaba: IPNI; 2010. p. 133-204.
• Bolland MDA, Baker MJ. Powdered granite is not an effective fertilizer for clover and wheat in sandy soils from Western Australia. Nutr Cycl Agroecosyst. 2000;56:59-68. https://doi.org/10.1023/A:1009757525421
  » https://doi.org/10.1023/A:1009757525421
• Brasil. Ministério da Agricultura, Pecuária e Abastecimento. Instrução Normativa nº 5, de 10 de março de 2016. Brasília, DF: Diário Oficial da União; 2016 [cited 2025 Sep 17]. Available from: https://www.gov.br/agricultura/pt-br/assuntos/insumos-agropecuarios/insumos-agricolas/fertilizantes/legislacao/in-5-de-10-3-16-remineralizadores-e-substratos-para-plantas.pdf
  » https://www.gov.br/agricultura/pt-br/assuntos/insumos-agropecuarios/insumos-agricolas/fertilizantes/legislacao/in-5-de-10-3-16-remineralizadores-e-substratos-para-plantas.pdf
• Coroneos C, Hinsinger P, Gilkes RJ. Granite powder as a source of potassium for plants: A glasshouse bioassay comparing two pasture species. Fertil Res. 1996;45:143-52. https://doi.org/10.1007/BF00790664
  » https://doi.org/10.1007/BF00790664
• Desmalles C, Jordan-Meille L, Hernandez J, Thomas CL, Dunham S, Deng F, McGrath SP, Haefele SM. Impact of basalt rock powder on ryegrass growth and nutrition on sandy and loamy acid soils. Agronomy. 2025;15:1791. https://doi.org/10.3390/agronomy15081791
  » https://doi.org/10.3390/agronomy15081791
• Dias RC, Teixeira PC, Pereira RN, Loiola JAD, Pereira MG. Curvas de resposta do milho à adubação potássica obtidas através de experimento fatorial. Revista Agrária Acadêmica. 2022;5:235-49.
• Dias RC, Zonta E, Teixeira PC, Alves BJR, Castro C, Veneu DM, Araújo ES, Mattiello EM, Bona FD, Oliveira FA, Cantarella H, Souza HA, Zilli JE, Gomes Junior JD, Polidoro JC, Guilherme LRG, Carvalho MCS, Monte MBM, Moraes MF, Pavinato PS, Nunes RS, Cantarutti RB, Almeida REM, Straliotto R, Urquiaga S, Scivittaro WB, Benites VM. Agrominerais silicáticos são fontes eficientes de potássio para plantas cultivada? Uma meta-análise. In: Brunetto G, Rozane DE, Loss A, Souza HA, editors. Tópicos em ciência do solo. Viçosa, MG: Sociedade Brasileira de Ciência do Solo; 2025. v. 13. p. 392-450.
• Friendly M, Fox J. candisc: Visualizing generalized canonical discriminant and canonical correlation analysis. R package version 0.9.0. 2024. https://doi.org/10.32614/CRAN.package.candisc
  » https://doi.org/10.32614/CRAN.package.candisc
• Gomides JFFB, Aranibar CGM, Gutiérrez NL, Zuffo AM, Aguilera JG, Gonzales HL, López LMS, García WEV, Leite MS, Ratke RF, Steiner F, Morales-Aranibar LF. Identification of modern high-yield soybean genotypes for potassium-use efficiency in sandy soil of the Brazilian Cerrado. Agronomy. 2023;13:2639. https://doi.org/10.3390/agronomy13102639
  » https://doi.org/10.3390/agronomy13102639
• Hinsinger P, Bolland MDA, Gilkes RJ. Silicate rock powder: Effect on selected chemical properties of a range of soils from Western Australia and on plant growth as assessed in a glasshouse experiment. Fertil Res. 1995;45:69-79. https://doi.org/10.1007/BF00749883
  » https://doi.org/10.1007/BF00749883
• Jones Jr JB. Laboratory guide for conducting soil tests and plant analysis. Boca Raton: CRC Press; 2001. https://doi.org/10.1201/9781420025293
  » https://doi.org/10.1201/9781420025293
• Krahl LL, Paz SPA, Angelica RS, Valadares LF, Sousa-Silva JC, Marchi G, Martins ES. Successive off-take of elements by maize grown in pure basalt powder. Afr J Agric Res. 2020;15:229-39. https://doi.org/10.5897/ajar2019.14546
  » https://doi.org/10.5897/ajar2019.14546
• Krahl LL, Valadares LF, Sousa-Silva JC, Marchi G, Martins ES. Dissolution of silicate minerals and nutrient availability for corn grown successively. Pesq Agropec Bras. 2022;57:e01467. https://doi.org/10.1590/S1678-3921.pab2022.v57.01467
  » https://doi.org/10.1590/S1678-3921.pab2022.v57.01467
• Kuhn CES, Sousa JVL, Menezes BC, Gomes ACF. Spatial analysis of the feasibility of using rock powder as fertilizer in agriculture in Mato Grosso in Brazil. Sustainability. 2025;17:7668. https://doi.org/10.3390/su17177668
  » https://doi.org/10.3390/su17177668
• Lange I, Toro M, Arvidson RS, Kurganskaya I, Lüttge A. The role of crystal heterogeneity in alkali feldspar dissolution kinetics. Geochim Cosmochim Acta. 2021;309:329-51. https://doi.org/10.1016/j.gca.2021.06.032
  » https://doi.org/10.1016/j.gca.2021.06.032
• Maity A, Sharma J, Pal RK. Novel potassium-solubilizing bio-formulation improves nutrient availability, fruit yield and quality of pomegranate (Punica granatum L.) in semi-arid ecosystem. Sci Hortic. 2019;255:14-20. https://doi.org/10.1016/j.scienta.2019.05.009
  » https://doi.org/10.1016/j.scienta.2019.05.009
• Mancuso MAC, Soratto RP, Crusciol CAC, Castro GSA. Effect of potassium sources and rates on arabica coffee yield, nutrition, and macronutrient export. Rev Bras Cienc Solo. 2014;38:1448-56. https://doi.org/10.1590/S0100-06832014000500010
  » https://doi.org/10.1590/S0100-06832014000500010
• Manning DAC, Baptista J, Sanchez Limon M, Brandt K. Testing the ability of plants to access potassium from framework silicate minerals. Sci Total Environ. 2017;574:476-81. https://doi.org/10.1016/j.scitotenv.2016.09.086
  » https://doi.org/10.1016/j.scitotenv.2016.09.086
• Mohammed SMO, Brandt K, Gray ND, White ML, Manning DAC. Comparison of silicate minerals as sources of potassium for plant nutrition in sandy soil. Eur J Soil Sci. 2014;65:653-62. https://doi.org/10.1111/ejss.12172
  » https://doi.org/10.1111/ejss.12172
• Naderizadeh Z, Khademi H, Arocena JM. Organic matter induced mineralogical changes in clay-sized phlogopite and muscovite in alfalfa rhizosphere. Geoderma. 2010;159:296-303. https://doi.org/10.1016/j.geoderma.2010.08.003
  » https://doi.org/10.1016/j.geoderma.2010.08.003
• Penn CJ, Camberato JJ. A critical review on soil chemical processes that control how soil pH affects phosphorus availability to plants. Agriculture. 2019;9:120. https://doi.org/10.3390/agriculture9060120
  » https://doi.org/10.3390/agriculture9060120
• Pramanik P, Kalita C. Evaluated potential of mica waste to release sufficient potassium to support the growth of plants in acidic tea-growing soil in northeast India. Commun Soil Sci Plant Anal. 2019;50:1258-66. https://doi.org/10.1080/00103624.2019.1614603
  » https://doi.org/10.1080/00103624.2019.1614603
• Priyono J, Gilkes RJ. High-energy milling improves the effectiveness of silicate rock fertilizers: A glasshouse assessment. Commun Soil Sci Plant Anal. 2008;39:358-69. https://doi.org/10.1080/00103620701826498
  » https://doi.org/10.1080/00103620701826498
• R Core Team. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2025.
• Santana EB, Dias JCT. Phosphate solubilizing activity of native soil microorganisms from the rhizosphere of Jatropha curcas and from phosphate-solubilizing bacteria inoculum. Genet Mol Res. 2019;18:gmr18271. https://doi.org/10.4238/gmr18271
  » https://doi.org/10.4238/gmr18271
• Santos HG, Jacomine PKT, Anjos LHC, Oliveira VA, Lumbreras JF, Coelho MR, Almeida JA, Araújo Filho JC, Lima HN, Marques FA, Oliveira JB, Cunha TJF. Sistema brasileiro de classificação de solos. 6. ed. rev. ampl. Brasília, DF: Embrapa; 2025.
• Santos WO, Costa RF, Vergutz L, Mattiello EM. Production and evaluation of potassium fertilizers from silicate rock. J Plant Nutr Soil Sci. 2016;179:547-56. https://doi.org/10.1002/jpln.201500484
  » https://doi.org/10.1002/jpln.201500484
• Schoen C, Aumond JJ, Stürmer SL. Efficiency of the on-farm mycorrhizal inoculant and phonolite rock on growth and nutrition of Schinus terebinthifolius and Eucalyptus saligna Rev Bras Cienc Solo. 2016;40:e0150440. https://doi.org/10.1590/18069657rbcs20150440
  » https://doi.org/10.1590/18069657rbcs20150440
• Shimizu GD, Marubayashi RYP, Gonçalves LSA. AgroR: Experimental statistics and graphics for agricultural sciences. R package version 1.3.7. CRAN: Package AgroR; 2025. https://doi.org/10.32614/CRAN.package.AgroR.
  » https://doi.org/10.32614/CRAN.package.AgroR
• Sieber JR, Yu LL, Marlow AF, Butler TA. Uncertainty and traceability in alloy analysis by borate fusion and XRF. X-Ray Spectrom. 2005;34:153-9. https://doi.org/10.1002/xrs.789
  » https://doi.org/10.1002/xrs.789
• Silva FC. Manual de análises químicas de solos, plantas e fertilizantes. 2. ed. rev ampl. Brasília, DF: Embrapa Informação Tecnológica; 2009.
• Silva MB, Fernandes FRC, Luz AB, Castilhos ZC, Castilhos C, Rego F, Fernandes C, Benvindo DA, Zuleica L. Agrominerais para o Brasil. Rio de Janeiro: CETEM/MCT; 2010.
• Soratto RP, Crusciol CAC, Campos MD, Gilabel AP, Costa CHMD, Castro GSA, Ferrari Neto J. Efficiency and residual effect of alternative potassium sources in grain crops. Pesq Agropec Bras. 2021a;56:e02686. https://doi.org/10.1590/S1678-3921.pab2021.v56.02686
  » https://doi.org/10.1590/S1678-3921.pab2021.v56.02686
• Soratto RP, Crusciol CAC, Campos M, Costa CHM, Gilabel AP, Castro GSA, Ferrari Neto J. Silicate rocks as an alternative potassium fertilizer for upland rice and common bean crops. Pesq Agropec Bras. 2021b;56:e01411. https://doi.org/10.1590/S1678-3921.pab2021.v56.01411
  » https://doi.org/10.1590/S1678-3921.pab2021.v56.01411
• Swoboda P, Döring TF, Hamer M. Remineralizing soils? The agricultural usage of silicate rock powders: A review. Sci Total Environ. 2022;807:150976. https://doi.org/10.1016/j.scitotenv.2021.150976
  » https://doi.org/10.1016/j.scitotenv.2021.150976
• Teixeira PC, Donagemma GK, Fontana A, Teixeira WG. Manual de métodos de análise de solo. 3. ed. rev e ampl. Brasília, DF: Embrapa; 2017.
• van Straaten P. Distribution of agromineral resources in space and time - a global geological perspective. Pesq Agropec Bras. 2022;57:e01453. https://doi.org/10.1590/S1678-3921.PAB2022.V57.01453
  » https://doi.org/10.1590/S1678-3921.PAB2022.V57.01453
• Wallander H, Wickman T. Biotite and microcline as potassium sources in ectomycorrhizal and non-mycorrhizal Pinus sylvestris seedlings. Mycorrhiza. 1999;9:25-32. https://doi.org/10.1007/s005720050259
  » https://doi.org/10.1007/s005720050259
• Wickham H. ggplot2: Elegant graphics for data analysis. New York: Springer; 2016.
• Zhang X, He Z, Jia W, Meng F, Zhang W, Lu C, Hao X, Gai G, Huang Z, Xu M, Wang K, Yun S. Mechanism of potassium release from feldspar by mechanical activation in presence of additives at ordinary temperatures. Materials. 2024;17:144. https://doi.org/10.3390/ma17010144
  » https://doi.org/10.3390/ma17010144

*

Corresponding author: e-mail: [e-mail]

Editor:

José Miguel Reichert https://orcid.org/0000-0001-9943-2898 and Clístenes Willians Araújo do Nascimento https://orcid.org/0000-0002-5103-5524

Publication Dates

• Publication in this collection
  02 Nov 2026
• Date of issue
  2026

History

• Received
  14 Feb 2026
• Accepted
  03 May 2026