Evaluation of the Residual Effects of New Methodologies of Correction Surface and Subsurface Acidity on Soil Chemical Attributes in Agropastoral System

1. Introduction

Maintaining the long-term production capacity of soil is essential for increasing crop yields to meet global food demand [1]. There is global interest in sustainable soil management systems such as no-till (NT) [2,3,4] and agropastoral systems.These systems enhance soil quality and, in turn, crop yields by improving the chemical, physical and biological attributes of soil [5] and conserving biodiversity [6]. However, in many areas of the world, increases in crop yields are limited by soil acidity (pH CaCl₂ values below 4.4) [7,8].

In Brazil, most agropastoral systems are located in the Cerrado biome (the Brazilian savanna), where soil acidity is the result of leaching of basic ions promoted by high rainfall over many years. This high degree of weathering produces soils with low calcium (Ca)²⁺, magnesium (Mg)²⁺, potassium (K)⁺, and phosphorus (P) contents [9], low cation exchange capacity (CEC), low base saturation (BS) [10], and high levels of elements that are toxic to plants, such as exchangeable aluminum (Al³⁺) and manganese (Mn) [11,12,13]. The toxicity caused by high Al³⁺ content, P deficiency and low BS hinders root growth, reduces the absorption of water and nutrients by plants [14], and consequently limits crop yields [15,16], particularly in dry periods [17]. In addition, low Ca²⁺ content in deeper soil layers may restrict root exploration to the surface layer.

Applying limestone (LS), i.e., liming, efficiently increases soil pH and BS, supplies Ca²⁺ and Mg²⁺, and decreases exchangeable Al³⁺ and Mn [11,18]. However, the mobility of the products of LS dissolution is low and depends on the leaching of organic and/or inorganic salts throughout the soil profile; consequently, LS reactivity is restricted to the site of soil application [18,19]. This is problematic in systems without soil tillage, such as agropastoral systems of the Cerrado, where soil acidity correction is usually carried out by applying LS on the surface without incorporation into the soil [20]. [8,21,22,23,24] confirmed that surface-applied LS has low mobility, and [25] verified that liming without incorporation corrects acidity only in the 0–0.1 m layer.

Soil amendments like phosphogypsum (PG, CaSO₄) can increase the efficiency of LS application [11,13,26,27,28,29]. PG is more soluble than LS (2.5 g L^-1) [30] and leaches to subsurface layers [31]. As it moves down the soil profile, PG increases the supply of Ca²⁺ and sulfate-S (S-SO₄^2-) and reduces Al³⁺ activity [17,32,33,34]. Once in the soil solution, Ca²⁺ adsorbs to the soil exchange complex and displaces Al³⁺, K⁺ and Mg²⁺ into the soil solution. These cations react with SO₄^2- to form AlSO₄⁺ (which is less toxic to plants than Al³⁺) and the neutral ion pairs K₂SO₄⁰ and MgSO₄⁰, which, along with CaSO₄⁰, have high mobility in the soil profile [35]. During soil acidity correction, the carbonate of LS is consumed and is no longer available to accompany Ca²⁺ to deeper soil layers, whereas the sulfate of PG remains available [36]. By improving the soil profile, PG application promotes the development of the crop root system [34,37,38,39,40] to explore water and nutrients [41], reducing the adverse effects of drought [30], improving crop tolerance to water stress [42], and increasing crop yields [11], especially in the Cerrado biome.

Another input for improving soil fertility is hydrated lime (HL) [20]. HL is a fine powder produced by the hydration of virgin lime by an industrial process and in comprised of Ca(OH)₂ and Mg(OH)₂ [43,44]. However, there is little information on application methodologies and the long-term effects of HL application, particularly in agropastoral systems.

Methodologies that provide long-term amelioration of surface and subsurface acidity and improvements in soil chemical fertility are necessary to decrease production costs and increase the crop yields of farmers in the Cerrado. [20,45] evaluated different methodologies for applying LS, PG, and HL in a Typic Hapludalf in an agropastoral system in Brazil on soil chemical attributes, soybean [Glycine max (L.) Merr.] and maize (Zea mays L.) grain yields, and the dry matter yield of palisade grass (Urochloa brizantha syn. Brachiaria brizantha cultivar Marandu). [20] found that after three seasons, all methodologies corrected surface acidity (0.0–0.2 m) by increasing pH and BS and reducing total acidity, but only subsurface application of HL, combined with both surface application of PG in the 1^st year and surface application of HL in subsequent years increased pH and BS in the subsurface layer (0.4–0.8 m). [20] also determined that applying LS or HL plus PG increased S-SO₄^2- content throughout the soil profile (0.0–0.8 m); in the 0.0–0.2 m layer, applying LS plus PG increased Ca²⁺ content and Ca²⁺/CEC, and applying HL plus PG increased Mg²⁺ content and Mg²⁺/CEC. However, the long-term effects of these methodologies in agropastoral systems are unclear. To address this gap, this study tested the following hypotheses: (a) the improvements in surface and subsurface acidity and soil chemical fertility produced by different methodologies for applying LS, PG, and HL observed by [20,45] persist after two seasons, and (b) further changes in soil chemical attributes occur after two seasons.

2. Materials and Methods

2.1. Description of Study Area

The experiment was carried out at the Advanced Research and Development Division for Rubber Tree and Agroforestry Systems of the Agronomic Institute (IAC) of the São Paulo Agency for Agribusiness Technology (APTA), Votuporanga, São Paulo State, Brazil (20º20’S, 49º58’W and 510 m altitude), in an area with a slope < 5%. The study site is located in the Cerrado biome. Its natural vegetation has peculiar characteristics, such as small trees, thick and twisted trunks, thick bark, with large leaves.

As described previously by [45], prior to the experiment, the area had been used for grain and seed production under conventional soil tillage. In the 2014/2015 season, castor bean was grown for seed production. During the 2015/2016 growing season, maize was cultivated under no-till in the summer and sunn hemp (Crotalaria juncea) in the winter-spring season. In the 2016/2017 growing season, maize was intercropped with Congo grass (U. ruziziensis syn. B. ruziziensis) in the summer under no-till. The Congo grass was not grazed but was used as a cover crop before the next crop. Details of study area are shown in Figure S1 in the Supplemental Material.

2.2. Description of Soil and Climate

The soil in the experimental area is classified as an Arenic Hapludult [46] and will be referred to in this study as an Ultisol with sandy texture. According to Köppen’s classification, the regional climate type is Aw (tropical with dry winters). The average annual maximum temperature is 31.2°C; the average annual minimum temperature is 17.4°C; and the annual mean temperature is 24°C. The annual average rainfall is 1,328.6 mm. Figure 1 presents monthly rainfall, minimum and maximum relative humidity, and minimum and maximum temperature data for Votuporanga from November 1, 2017, to December 31, 2022.

2.3. Description of Experimental Design

The experimental design was randomized complete blocks with four replications using seven treatments involving three Ca compounds (LS, PG, and HL). Each plot had a size of 0.4 ha.

2.4. Description of Treatments

Seven treatments with different combinations of Ca compounds and application management methodologies, over three growing seasons (2017–2018, 2018–2019, and 2019–2020) were evaluated (Table 1). Three Ca compounds were used: LS, which contained 42% CaO and 7% MgO; PG, which contained 17% CaO and 14% S (S-SO₄^2-); and HL, which contained 40% CaO and 27% MgO. In T1, the control treatment, LS, PG, and LH were not applied, and soil tillage was not performed. In T2, LS and PG were surface-applied (no tillage, NT) at doses of 2,000 kg ha^-1 and 2,500 kg ha^-1, respectively, in the 1^st year; in the 2^nd and 3^rd years, LS was applied to the surface to ensure that Ca⁺² occupied 65% of CEC in the 0.0–0.2 m layer. In the 2^nd and 3^rd years, PG was not applied. In T3 and T4 conventional soil tillage system (CT) was used. In T3, in the 1^st year, LS was applied at a dose of 1,000 kg ha^-1 and incorporated with a heavy-duty hydraulic harrow (28” discs) and a leveling harrow (20” discs, two times). After that, LS was applied at doses of 1,350 kg ha^-1 and incorporated with a moldboard plow up to 0.26 m; and PG was applied at dose of 1,500 kg ha^-1 and incorporated with a leveling harrow. In the 2^nd and 3^rd years, LS was surface-applied to ensure that Ca⁺² occupied 65% of CEC in the 0.0–0.2 m layer, and PG was surface-applied to ensure that Ca⁺² occupied 50% of effective cation exchange capacity (ECEC) in the 0.2–0.4 m layer. In T4, in the 1^st year, LS was applied at a dose of 1,000 kg ha^-1 and incorporated with a heavy-duty hydraulic harrow (28” discs) and a leveling harrow (20” discs, two times). After that, LS was applied at doses of 1,350 kg ha^-1 and incorporated with a moldboard plow up to 0.26 m. In the 2^nd and 3^rd years, LS was surface-applied to ensure that Ca⁺² occupied 65% of CEC in the 0.0–0.2 m layer. In T5, HL and PG were surface-applied (NT) at doses of 725 kg ha^-1 and 2,500 kg ha^-1, respectively, in the 1^st year; in the 2^nd and 3^rd years, HL was surface-applied to ensure that Ca⁺² occupied 65% of CEC in the 0.0–0.2 m layer, and PG was surface-applied to ensure that Ca⁺² occupied 50% of ECEC in the 0.2–0.4 m layer. In T6 and T7 minimum soil tillage system (MT) was used. In T6, in the 1^st year, HL was applied in the 0–0.57 m layer using a 2-shank subsoiler-fertilizer at a dose of 450 kg ha^-1, and PG was surface-applied at a dose of 2,500 kg ha^-1. In the 2^nd and 3^rd years, HL was surface-applied to ensure that Ca⁺² occupied 65% of CEC in the 0.0–0.2 m layer, and PG was not applied. In T7, HL was subsurface applied in the 0–0.57 m layer using a 2-shank subsoiler-fertilizer at a dose of 260 kg ha^-1 in the 1^st year, and PG was surface-applied at a dose of 2,500 kg ha^-1. In the 2^nd and 3^rd years, HL was surface-applied to ensure that Ca⁺² occupied 65% of CEC in the 0.0–0.2 m layer. In the 2^nd and 3^rd years, PG was not applied. Details of the treatments are shown in Figure S2 in the Supplemental Material.

In the 2^nd and 3^rd growing seasons, the dose of LS or HL was calculated based on the results of soil analysis to ensure that Ca occupied 65% of CEC in the 0.0–0.2 m layer according to the methodology of [45]: LS/HL (Mg ha⁻¹) = Ca²⁺ saturation of CEC − exchangeable Ca²⁺ content in cmol_c dm⁻³ over a depth of 0.0–0.2 m. The CEC at a depth of 0.0-0.2 m was calculated as the sum of the contents of exchangeable cations: CEC = Ca²⁺ + Mg²⁺ + potassium (K)⁺ + total acidity (hydrogen (H)⁺ + Al³⁺). The Ca²⁺ saturation (65%) at a depth of 0.0–0.02 m was calculated as 100 × Ca²⁺/CEC.

The dose of PG in the 2^nd and 3^rd growing seasons was calculated based on the results of soil analysis to ensure that Ca occupied 50% of ECEC in the 0.2–0.4 m layer, following the methodology of [48]: PG (Mg ha^-1) = (0.5 × Ca²⁺ saturation in ECEC − exchangeable Ca²⁺ content in cmol_c dm⁻³ at a depth of 0.2–0.4 m) × 6.4. The ECEC at a depth of 0.2-0.4 m was calculated as the sum of the contents of exchangeable cations: ECEC = Al³⁺ + Ca²⁺ + Mg²⁺ + K⁺. The Ca²⁺ saturation (50%) at a depth of 0.2–0.4 m was calculated as 100 × Ca²⁺/ECEC.

2.5. Crop Management

The details of soil and crop management in the treatments in the 2017-2018 (1^st), 2018-2019 (2^nd), and 2019-2020 (3^rd) seasons were described previously by [45] and are presented in Tables S1, S2, and S3, respectively, in the Supplemental Material. In brief, the doses of LS, PG, and HL used in the treatments were determined based on the results of soil sampling at the end of the previous season. Seventy-one days after the maize harvest (2018/2019 season), 10 beef cattle with an average weight of 484 kg were introduced into the experimental area. To adjust the grazing height, on August 23, 2019, five more animals with an average weight of 397 kg were introduced into the area. A continuous grazing system was used, and the animals were allowed to move freely about the pasture in all plots. On October 29, 2019, the 15 beef cattle animals were removed from the area because of a drastic reduction in forage supply caused by water deficiency. On November 26, 2019, 10 different newly weaned animals with an average weight of 268 kg were introduced in a continuous grazing system. In the 2020-2021 and 2021-2022 seasons, 10 weaned beef cattle remained in the experimental area. On October 01, 2021, topdressing fertilization of palisade grass was performed using 10-10-10 fertilizer (10% N-urea, 10% P₂O₅- single superphosphate, and 10% K₂O- potassium chloride) at 1,000 kg ha^-1. We used this quantity of 10-10-10 fertilizer because palisade grass demands 13 kg of nitrogen per 1000 kg of forage dry matter yield [49] and the last fertilization of the experimental area was performed on March 1, 2019 (2^nd topdressing fertilization of maize, Table S3).

2.6. Sampling and Analysis

2.6.1. Soil Sampling and Analysis

Before applying the treatments, soil was sampled from the 0.0–0.1, 0.1–0.2, 0.2–0.4, and 0.4–0.8 m layers to determine the initial soil characteristics, fertility [50] and sediment granulometric characteristics (sand, silt, and clay content). The samples were collected at ten random points of each depth increment (0.0–0.1, 0.1–0.2, 0.2–0.4, and 0.4–0.8 m layers). The ten subsamples were homogenized and pooled to form a composite sample of each depth increment. On April 10, 2018, and April 19, 2019, new soil samples were collected for chemical analysis, determination of fertility [50], and calculation of the doses of LS, PG, and HL to be reapplied according to the treatments. The samples were collected at five random points in each plot in the 0.0–0.2 and 0.2–0.4 m layers; thus, five subsamples were collected from each layer of each plot. The five subsamples were homogenized and pooled to form a composite sample of each layer of each plot. Soil was sampled with a metal probe and air-dried before analysis.

The pH (1:2.5 soil/0.01 Μ CaCl₂ suspension) and total acidity pH 7.0 (H⁺ + Al³⁺) were determined by the methodology of [50]. Total acidity was determined by titrimetry and was extracted from the soil using 1 mol L^-1 calcium acetate (C₄H₆CaO₄) pH 7, a buffered solution that removes Al³⁺ and undissociated H from the soil. The S-SO₄^2- content was determined by the calcium phosphate (Ca₃(H₂PO₄)₂) method [49], and the levels of P, K⁺, Ca²⁺, and Mg²⁺ in the soil were determined by extraction with ion-exchange resin [50]. The Ca₃(PO₄)₂ method is based on the extraction of S-SO₄^2- from soil samples by a solution of 0.01 mol L^-1 Ca(H₂PO₄)₂. Quantification is performed by turbidimetry, which detects barium sulfate (BaSO₄) formed by the reaction of barium chloride dihydrate (BaCl₂.2H₂O) with S-SO₄^2- extracted from soil samples. Extraction with ion-exchange resin allows the evaluation of so-called labile phosphorus by gradual dissolution of phosphate compounds from the solid phase of the soil and transfer of orthophosphate ions to the ion-exchange resin. Furthermore, as the extraction is carried out with a mixture of cationic and anionic exchange resins saturated with sodium bicarbonate, the exchangeable cations are also largely transferred from the soil to the resin, especially if the levels are not too high. An advantage of using sodium bicarbonate is that the bicarbonate ions buffer the medium at a pH close to neutrality, which is favorable for the dissolution of phosphates from the soil, while the sodium ions saturate the cationic resin, allowing the removal of exchangeable cations from the soil. The Ca²⁺ and Mg²⁺ were determined by atomic absorption spectrophotometry, and K⁺ was determined by flame photometry. These results were used to calculate the values of BS through the relationship between the content of exchangeable bases in the soil (Ca²⁺, Mg²⁺, and K⁺) and CEC and ECEC in cmolc dm^-3 as well as the percentage ratios of K⁺/CEC, Ca²⁺/CEC, Mg²⁺/CEC, and Ca²⁺/ECEC.

To assess the long-term effects of the methodologies used to apply Ca compounds in the 2017-2020 seasons, soil samples were collected on December 12, 2022, for chemical analysis and determination of fertility. Soil was sampled with a metal probe and air-dried before analysis. The samples were collected at five random points of each depth increment (0.0–0.1, 0.1–0.2, 0.2–0.4, and 0.4–0.8 m layers) in each plot. The five subsamples from each layer were homogenized and pooled to form a composite sample of the layer. The chemical attributes of the samples were analyzed using the same methodology described previously [50].

2.6.2. Straw Sampling and Analysis

Straw was sampled in the 1^st and 2^nd growing seasons. Specifically, the straw within an area of 0.5 x 0.5 m was collected at 10 locations in each plot, packed in paper bags, and dried at 65–70ºC in a forced ventilation oven for 72 hours.

2.7. Statistical Analysis

Before analyzing the data to evaluate the residual effect of the treatments applied in the 2017/2018, 2018/2019, and 2019/2020 growing seasons, the normality and homoscedasticity of the data were analyzed by the Shapiro-Wilk and Bartlett tests, respectively, both at 0.05 probability. Differences in soil chemical attributes in each soil layer (0.0–0.1, 0.1–0.2, 0.2–0.4, and 0.4–0.8 m) in the 2022-2023 season were analyzed by ANOVA (F test). Averages were compared by Tukey’s test (P ≤ 0.05) in Assistat [51].

3. Results

The average amount of straw dry matter present in the area was 9,550 kg ha^-1 in the 2017-2018 season and 2,424 kg ha^-1 in the 2018-2019 season. The initial chemical characteristics and the sediment granulometric characteristics of the soil are shown in Table S4 in the Supplemental Material. Table 2, Table 3, Table 4 and Table 5 present the chemical attributes of the soil in the 0.0–0.1, 0.1–0.2, 0.2–0.4, and 0.4–0.8 m layers, respectively, in the 2022-2023 season. The Al³⁺ content was 0.0 cmol_c dm^-3 in all samples from the four layers and thus was excluded from statistical analysis. Figure 2, Figure 3 and Figure 4 illustrate the chemical attributes of the soil in the 0.0–0.1, 0.1–0.2, 0.2–0.4, and 0.4–0.8 m layers in the 2022-2023 season, including P (Figure 2a); S (Figure 2b); organic matter (OM) (Figure 2c); pH (Figure 3a); Ca²⁺ (Figure 3b); Mg²⁺ (Figure 3c); H⁺ + Al³⁺ (Figure 4a); BS (Figure 4b); Ca²⁺/CEC (Figure 4c); and Mg²⁺/CEC (Figure 4d).

3.1. Soil Chemical Properties by Layer

Table 2 presents the characteristics of the surface (0.0–0.1 m) layer of soil. Total acidity decreased by 0.55–0.3 cmol_c dm^-3 in the treatments in which Ca compounds were applied (T2–T7). The differences between treatments in relation to soil S-SO₄^2- content were in maximum of 1.5 mg dm^-3. Treatment T2 increased K⁺/CEC by percentage-point increments of 2.65 over T3, and by 2.29 % over T4. Treatment T5 increased resin extractable P by 21.5 to 26 mg dm^-3 compared with T1 (control), T3 and T4. Treatment T3 increased pH by 1.05, 0.42, 0.55, and 0.2 units over T1, T2, T4, and T7, respectively. Compared with T1, T4, and T7, T3 increased Ca²⁺ content by 2.3, 1.55, and 1.8 cmol_c dm^-3, respectively; BS by percentage-point increments of 23.0, 14.5, and 12.0, respectively; and CEC by 2.13, 1.98, and 1.68 cmol_c dm^-3, respectively. Compared with T1 and T4, T3 increased OM content by 3.0 and 4.0 g dm^-3, respectively, and Mg²⁺ content by 0.6 and 0.55 cmol_c dm^-3, respectively. Compared with T1, T4, and T7, T5 increased Ca²⁺ content by 2.28, 1.53, and 1.78 0.55 cmol_c dm^-3, respectively; BS by percentage-point increments of 22.0, 13.5, and 11.0, respectively; CEC by 2.04, 1.89, and 1.59 cmol_c dm^-3, respectively; and Ca²⁺/CEC by percentage-point increments of 28.0, 11.5, and 20.0, respectively. Compared with T3 and T4, K⁺ content was 0.14 to 0.17 cmol_c dm^-3 higher in T5, T6, and T7. Soil Mg²⁺/CEC increased by percentage-point increments of 8.0 in T6 and T7 compared with T2. Treatments 2 through 7 decreased K⁺/CEC by percentage-point increments of 2.24 to 7.0 relative to T1 (control).

Table 3 presents the characteristics of the 0.1 to 0.2 m soil layer. In the T2 – T7 treatments, pH increased by 0.15–0.8 compared with T1. The total acidity was 0.25–0.5 cmol_c dm^-3 higher in T2 – T7 treatments compared with T1. The BS and Ca²⁺/CEC increased by percentage-point increments of 6.0–17.5 and 1.0–15.75 compared with T1, respectively. Treatment T3 increased Mg²⁺/CEC by percentage-point increments of 2.65 over T1. Treatment T5 increased P content by 11.0–16.25 mg dm^-3 compared with T2, T4, T6, and T7; by 19.75 mg dm^-3 compared with T3; and by 30.25 mg dm^-3 compared with the control. Compared with T4, OM content and Mg²⁺ content were 2.5 g dm^-3 and 0.4 cmol_c dm^-3 higher, respectively, in T3.

Table 4 presents the characteristics of the 0.2 to 0.4 m soil layer. The differences between treatments in relation to soil S-SO₄^2- content were in maximum of 1.0 mg dm^-3. In the T2 – T7 treatments, OM increased by 1.5–2.5 g dm^-3 compared with T1. Treatment T4 increased Ca²⁺ content by 0.65 cmol_c dm^-3 and BS by percentage-point increments of 15.5 over T1. The Ca²⁺/ECEC was 8.49 and 10.11 percentage-point increments of in treatment T2 compared with T3 and T4 treatments. Treatment T4 increased Mg²⁺ content by 0.52–0.77 cmol_c dm^-3 and CEC by 0.75–1.43 cmol_c dm^-3 compared with the other treatments. In addition, T4 increased Mg²⁺/CEC by 8.59–14.85 percentage points compared with T1, T2, T5, T6, and T7. T6 and T7 increased P content by 12.0–14.75 mg dm^-3 compared with T3 and T4. T2 increased K⁺ content by 0.21 cmol_c dm^-3 compared with the control and Ca²⁺/ECEC by 8.49–10.11 percentage points compared with T3 and T4. No significant differences (P > 0.05) in any soil chemical attributes were observed between the treatments with HL (T5, T6, and T7).

The characteristics of the 0.4 to 0.8 m soil layer are in Table 5. In the T2 – T7 treatments, P increased by 4.5–13.0 mg dm^-3 compared with T1. Treatment T4 increased pH by 0.6 units; decreased total acidity by 0.2 cmol_c dm^-3; and increased BS by percentage-point increments of 9.0 over T1. Compared with the other treatments, T4 increased K⁺ content by 0.13–0.26 cmol_c dm^-3 and K⁺/CEC by percentage-point increments of 4.38–8.68. Treatments T3 and T4 increased K⁺ content by 0.08–0.21 cmol_c dm^-3 compared with T2.

3.2. Trends of Soil Characteristics by Layer and Treatment

Compared with T1, T2–T7 treatments decreased total acidity by 0.3–0.55 cmol_c dm^-3 in the surface layer (0.0–0.1 m) and by 0.25–0.5 cmol_c dm^-3 in the 0.1–0.2 m layer; The P content by 10.5–30.25 mg dm^-3 in the 0.1–0.2 m layer, 10.0–24.75 mg dm^-3 in the 0.2–0.4 m layer, and 4.5–13.0 mg dm^-3 in the 0.4–0.8 m layer; and OM content by 1.5–2.5 g dm^-3 in the 0.2–0.4 m layer and 1.0–2.0 g dm^-3 in the 0.4–0.8 m layer. These results corroborate those of [52], who observed lasting benefits of liming for total acidity, P availability, and OM content for at least three years after application.

On average, compared with T1, T2–T7 treatments increased BS in the 0.0–0.1, 0.1–0.2 and 0.2–0.4 m layers; the pH in the 0.0–0.1 and 0.1–0.2 m layers; the K⁺ content in the 0.2–0.4 and 0.4–0.8 m layers; Ca²⁺ content and Ca²⁺/CEC in the 0.0–0.1 m layer; and K⁺/CEC in the 0.4–0.8 m layer.

In the surface layer, Ca²⁺/CEC was highest in T5; Mg²⁺/CEC was highest in T6 and T7 and lowest in T2; Mg²⁺ content was highest in T3; and Ca²⁺ content, pH, BS, and CEC were highest in T3 and T5. The Ca²⁺/CEC ranged by 29 (T1) to 57% (T5), however none of treatments provided 65% of Ca²⁺ in CEC.

In the 0.1–0.2 m layer, OM content was lowest in T4 and highest in T7; Mg²⁺ content was highest in T3 (as in the surface layer) and T6; BS was highest in T3 and T5 (as in the surface layer); Ca²⁺/CEC was highest in T2; and Mg²⁺/CEC was highest in T3 and lowest in T2 (as in the surface layer). The Ca²⁺/CEC ranged by 33 (T1) to 49% (T5), however, as in the surface layer, none of treatments provided 65% of Ca²⁺ in CEC.

In the 0.2–0.4 m layer, P content was highest in T6 and T7; K²⁺ content was highest in T2; and Ca²⁺ content and BS were highest in T4. Every treatments provided 50% of Ca²⁺ in ECEC. In the 0.4–0.8 m layer, pH, BS, and CEC were highest in T4, whereas total acidity was lowest in this treatment.

4. Discussion

This study confirms that the different methodologies for applying Ca compounds for surface and subsurface acidity correction had residual effects on soil chemical attributes at depths of 0.0 to 0.8 m in this agropastoral system. These findings are consistent with the results of prior studies in tropical no-till systems. Specifically, [52] observed that LS and PG amendment had long-term positive effects on P content in a tropical no-till intercropping system, and [53] observed residual effects of liming in a Brazilian Oxisol, including increased P content and reduced total acidity. In an annual crop system, [33] found a prolonged residual effect of PG that reduced total acidity in the subsurface layers of the soil. Similarly, [54] determined that total acidity in the 0.0–0.3 m layer was reduced two years after liming.

The PG application in the 2017-2018, 2018-2019, and 2019-2020 seasons was responsible for the increases in OM content, Ca²⁺ and Mg²⁺ contents, pH, BS, and CEC in the surface layer and in OM and Mg²⁺ contents in the 0.1–0.2 m layer in T3 compared with T4. Although PG does not directly affect soil pH, it increases P, Ca²⁺, S-SO₄^2-, and CEC availability [52].

The increases in Mg²⁺ content, CEC, and Mg²⁺/CEC in the 0.2–0.4 m layer and K⁺ content and K⁺/CEC in the 0.4–0.8 m layer in T4 compared with the other treatments were due to the slow effect of LS. After LS application, there is a period of maximum reactivity in the soil, after which the residual effect gradually decreases [47]. The factors that influence the reaction time and residual effect of LS in the soil include the soil buffer power, degree of homogenization during LS incorporation [49], and LS particle size [47]. [56] observed that the reactivities of the particle size fractions assigned to LS by Brazilian law were only reached approximately 18 months after application. [57] observed increased Mg²⁺ content two years after the application of LS without PG.

4.1. Effects of the Treatments on pH and Total Acidity

Consistent with the results of this study, [54,58] observed residual effects of the surface application of LS and PG under no-till on pH, with increases in the 0.0–0.1 and 0.1–0.2 m layers. The combined application of LS and/or HL with PG provided better conditions for LS and/or HL to act on the soil solution and increase the pH, corroborating the results of [59]. Increased pH is the result of physical downward movement of fine LS particles through the continuous porosity in the soil profile formed after the decomposition of dead roots or by soil organisms; the formation of ionic pairs between NO₃^- or SO₄^2- (released from fertilizer or mineralized OM) and Ca and Mg from the amendment; or the formation of water-soluble CaL⁰ or CaL^- type complexes, which involving the incorporation of hydrophilic groups onto the ligand (L) such as sulfonated or charged substituents and then coordinating the metal ion, between plant residues on the soil surface (carboxylic and phenolic radicals) and Ca²⁺ or Mg²⁺ [54,60,61,62,63,64]. In addition, PG can indirectly increase soil pH in deeper layers as the S-SO₄^2- from the PG displaces OH^- from soil colloid surfaces into the soil solution [16]. In agropastoral systems under no tillage, low-molecular-weight organic acids, released during the decomposition of animal waste, mainly feces, or exuded by pasture residues [65], can increase the effects of surface application of LS and/or HL at greater depths by favoring the downward movement of Ca²⁺ and Mg²⁺ in the soil profile.

Although the incorporation of LS with plowing and harrowing (T3 and T4, conventional soil tillage) and the subsurface application of HL with a shank subsoiler-fertilizer (T6 and T7, minimum tillage) corrected total acidity in the 0.4–0.8 m layer in 2020 [20], residual effects of these treatments were not observed. Decreases in total acidity were observed only in the 0.0–0.1 and 0.1–0.2 m layers in T2–T7. The 10-10-10 fertilizer used in topdressing fertilization of palisade grass contained urea as the N source. Urea contains N-NH₃ [25], which rapidly oxidizes to nitrate in the soil, releasing H⁺ [66]. This may be one reason for the lack of a decrease in total acidity (H⁺ + Al³⁺) in the 0.2–0.4 and 0.4–0.8 m layers in T2–T7. It is also possible that the decomposition of organic residues originating from soybean, maize and palisade grass acidified the soil, decreasing pH and increasing total acidity [67,68].

In T2, residual effects were not observed in the 0.4–0.8 m layer because the action of LS without incorporation is restricted to the surface layer in acidic soils (such as those in the Cerrado biome). In this treatment, PG was applied only in the first season (2017-2018 season), thus restricting the benefits of the combined application of LS with PG to the 2017-2020 seasons. [62,70,71,72,73,74] also found more pronounced positive effects of surface liming in the surface layers of the soil. The basic anions from the dissolution of LS (OH^- and HCO₃^-) move by mass flow to the deeper layers of the soil, where they react with acidic cations (H⁺, Fe²⁺, Al³⁺, and Mn²⁺), preventing further alkalization reactions [75].

4.2. Effects of the Treatments on Base Saturation

T2–T7 had residual positive effects on BS in the 0.0–0.1, 0.1–0.2, and 0.2–0.4 m layers. The LS, PG, and HL applied in the 2017-2018, 2018-2019, and 2019-2020 seasons contained 42% CaO and 7% MgO; 17% CaO and 14% S (S-SO₄^2-); and 40% CaO and 27% MgO, respectively. Reaction of Ca²⁺ in the soil solution with the soil exchange complex displaces K⁺ and Mg²⁺ into the soil solution, where these cations react with SO₄^2- to form the neutral ionic pairs K₂SO₄⁰ and MgSO₄⁰, which, along with CaSO₄⁰, have great mobility in the soil profile [35]. Ca²⁺, K⁺, and Mg²⁺ compete for some of the same adsorption sites in the soil [76].

Therefore, increasing Ca²⁺, K⁺, and Mg²⁺ contents in the soil profile also increases BS. This is very important in production systems under no tillage because increasing base saturation can reduce chemical impediments to root development and increase resistance to water deficits, which are common for maize and soybean crops [42].

4.3. Effects of the Treatments on P Content

[77] observed residual positive effects of PG application on P content and attributed these effects to the presence of P residue in PG itself. These residual effects can also be attributed to the formation of compounds between Ca and P due to the dissociation of compounds containing Ca and S in the presence of water, which releases Ca²⁺ and SO₄^2- ions. Ca²⁺ subsequently reacts with P, reducing the solubility of P [78]. [16,79,80] also observed increases in P content in the surface layer after the application of high doses of PG.

As expected, P content in the 0.0–0.1 and 0.1–0.2 m layers increased under no-till (T5 and T2) compared with conventional soil tillage (T3 and T4) because the lack of soil turnover promotes nutrient accumulation in the surface layers [16,81]. [82,83] observed similar results and concluded that P adsorption capacity decreases as the soil concentration of P increases. In a study of soil management and residual effects of PG application on soil physical and chemical attributes in Brazil, [84] also determined that P content was higher under no-till than under conventional tillage, especially in the surface layer, and attributed this difference to the immobility of P in the soil. [85] reported greater enrichment of nutrients in the 0.0–0.05 m layer under no-till and found that this increase was greatest for P content, which was 4 –7 times higher than under conventional tillage.

The P content in surface layers is higher under no-till than under conventional tillage. Because P is not very mobile, concentrations of P are highest in the superficial soil layer. However, under no-till, the constant supply of organic material increased organic P in deeper layers, which increased the fertility of the soil after mineralization. [86,87,88,89,90] showed that a few years after the establishment of no-till, the chemical, physical and biological properties of the surface layer differ from those in an intensive tillage system that includes regular moldboard plowing with secondary tillage.

4.4. Effects of the Treatments on K⁺ Content and K⁺/CEC

The increases in K⁺ content and K⁺/CEC in the surface layer in the control were the result of K fertilization in the 2017-2018, 2018-2019, and 2019-2020 seasons. K⁺ was probably adsorbed by soil colloids in the surface layer, as observed by [25]. By contrast, the dry matter yield of palisade grass forage in T2 in the 2019-2020 of 4,967 kg ha^-1 higher than T1 (as shown in 45), provided the increase in K⁺ content by 0.21 cmol_c dm^-3 in the 0.2–0.4 m layer. This occurred due K is the most abundant cation in plant tissues, being absorbed from the soil solution in large quantities by the roots in the form of the K⁺ ion [91]. In addition, K is required in large quantities by crops, equaling the amounts of N required, and being three or four times more accumulated in residues than P [92]. As shown in [45], the leaf contents of K, N, and P in the forage sample were 22.46; 14.07; and 3.18 g kg^-1 in T2 and 21.01; 12.92; and 2.63 in T1. As cited by [93], forage grasses, as palisade grass, have a very extensive and constantly renewed root system, which, combined with their high dry matter production potential, are capable of altering soil OM and nutrient levels in a short period of time. [94], in an sandy Oxisol in Votuporanga, São Paulo State, Brazil, verify an average increase in K⁺ content by 0.23 cmol_c dm^-3 with Congo grass used as cover crop after two seasons. Cover crops, as palisade grass, can restore considerable amounts of nutrients to crops, since these plants absorb nutrients from the subsurface layers of the soil and subsequently release them into the surface layer through the decomposition of their residues [95].

In T2–T7, Ca compounds were applied in these three seasons, resulting in downward movement of K⁺ from the surface layer to the 0.2–0.4 and 0.4–0.8 m layers. Consistent with these observations, [54] observed downward movement of LS and increased BS at depths of 0.3 m for two years after LS application on the soil surface. In addition, due to the thermodynamics of ion exchange and the properties of Ca²⁺, PG can potentially increase leaching losses of Mg²⁺ and K⁺ to deep layers [31,96]. The CaSO₄ in PG reduces Al³⁺ activity, decreasing subsurface acidity effects and redistributing basic cations such as Mg²⁺ and K⁺ from the surface to subsurface layers [37,39,40]. This decrease in subsurface acidity and redistribution of basic cations facilitate expansion of the root system, improving access to water and nutrients in deeper layers of the soil [78]. In sandy soil, as used in this study, these effects are very important, especially in years with dry periods.

The increase in K⁺/CEC in the surface layer in T2 compared to T3 and T4 may reflect greater K⁺ cycling due to the higher dry matter yield of palisade grass forage in T2 in the 2019-2020 season. Forage dry matter yield was 2,222 and 5,087 kg ha^-1 higher in T2 than in T3 and T4, respectively (as shown in [45]). K⁺ cycling in T2 may also have been enhanced by the lack of soil tillage (no-till) and the high average straw dry matter of maize and Congo grass present in the surface in the 2017-2018 season (9,550 kg ha^-1).

4.5. Effects of the Treatments on Ca²⁺ Content and Ca²⁺/CEC

The residual positive effects of Ca compounds on Ca²⁺ content and Ca²⁺/CEC were restricted to the surface layer, consistent with the results of a previous study of the long-term effects of PG and LS application under no-till [58]. This restriction is due to the greater ionic radius of Ca (compared with Mg and K) and its lower mobility in the soil profile [18]. There is competition between Mg²⁺ and Ca²⁺ for negative charges in the soil, with preference for Ca²⁺ at exchange sites [97].

4.6. Effects of the Treatments on Mg²⁺ Content and Mg²⁺/CEC

The application of Ca compounds increased Mg²⁺ content in the subsurface layer. A high concentration of Ca²⁺ in the soil favors the displacement of Mg²⁺ from exchange sites. The displaced Mg²⁺ can form an ion pair with SO₄^2- or be leached in the form of Mg²⁺ ions, which is the preferred form of displacement in the profile [98]. [30] observed similar effects and concluded that PG application favors cation exchange after adjustment of the acidity of the surface layers with LS. PG affects the distribution of not only S-SO₄^2- and Ca²⁺ but also other nutrients in the soil, especially when applied at high doses, promoting greater movement of Mg²⁺ to subsurface layers [99]. The supply of Ca²⁺ by PG solubilization promotes the substitution of Mg²⁺ from the exchange complex to the soil solution and the formation of the ionic pair MgSO₄⁰. This ionic pair is more easily leached in the soil by water infiltration, promoting the movement of Mg²⁺ in the soil profile [4,30,31].

4.7. Effects of the Treatments on Organic Matter Content

The increases in OM content in the subsurface layer in T2-T7 were due to the residual positive effects of Ca compounds on fertility throughout the soil profile, which included increases in P and K⁺ contents, BS, and K⁺/CEC. These improvements were reflected in the dry matter yield of palisade grass forage, which increased by 867–4967 kg ha^-1 in the 2019-2020 season (as shown in [45]). Similarly, [100] reported long-term positive effects of PG on black oat biomass 44 and 55 months after the application of the last third of the PG rate. [99] also found a residual positive effect of PG on soil chemical properties and the root dry mass density of sugarcane throughout the soil profile (0.0–2.0 m). The oldest stems and stalks of palisade grass forage have high C/N ratios and high levels of lignin and polyphenols, which results in slow decomposition [101] and favors increases in OM throughout the soil profile [102]. The root system of palisade grass forage also contributed to the increase in OM content in the subsurface layers. The effects of PG in the soil solution occur through the hydrolysis of CaSO₄ (calcium sulfate) to Ca²⁺ + SO₄^2- + CaSO₄⁰ [77]. The Ca²⁺ ions react with the soil exchange complex and displace Al³⁺, K⁺ and Mg²⁺ to the soil solution, where they react with sulfate (SO₄^2-) to form the AlSO₄⁺ complex (a form of Al that is less toxic to plants) and the neutral ionic pairs K₂SO₄⁰, MgSO₄⁰ and CaSO₄⁰, which have high mobility [103]. These ionic pairs move to deeper layers, improving the soil profile, crop root system development, water absorption through plant roots [38,104], and, ultimately, crop yields [105].

In a study of annual crops under no-till, [58] observed reduced Al³⁺ saturation and increased Ca²⁺ content 36 and 72 months after PG application, along with increases in maize and soybean yields. In a clayey, kaolinitic, thermic Rhodic Hapludox, [79] observed increases in maize grain yields of 7% 7–10 years after LS and PG were applied to the surface or 8% 7–10 years after LS was incorporated into the topsoil and PG was applied to the surface. These increases reflected the residual positive effects of Ca compounds in OM content and fertility throughout the soil profile. [52] also observed lasting improvements in soil chemical attributes at least three years after liming, including increases in P, Ca²⁺, Mg²⁺, and OM contents and consequently CEC and BS. Eight years after three applications of soil amendments, [29] determined that surface liming increased OM content by promoting aboveground and root biomass. [54] observed increases in Ca²⁺ and Mg²⁺ contents, pH, and BS and a decrease in total acidity in the 0.0–0.3 m layer two years after liming. The residual effects of liming observed by [53] included increases in Ca²⁺ and Mg²⁺ contents, pH, BS, and CEC and a decrease in total acidity.

4.8. Considerations

Several studies have reported limitations of surface application of LS in systems without soil tillage (no-tillage) [18,19,20]. The low mobility of the products of LS dissolution restricts the efficiency of this soil liming agent in reducing acidity in subsurface layers of the soil and results in variable loads, also known as pH-dependent loads and characteristics of highly weathered soils common in tropical regions such as Brazil, that depend on the leaching of organic and/or inorganic ion pairs deeper in the soil [19]. To contribute to methods for ameliorating surface and subsurface acidity and improving soil chemical fertility in agropastoral systems under no-till, we established this study of different application methodologies (surface, incorporation by soil tillage, or subsurface) for Ca compounds (LS, PG, and HL) during the 2017-2020 seasons. In contrast to our expectations, the improvements in soil chemical attributes throughout the soil profile (0.0–0.8 m) after the incorporation of LS and subsurface application of HL in the 2017-2018 season did not persist in subsequent seasons. The surface application of LS and HL reduces energy (fuel) costs compared with application under tillage [106]. Because the positive effects of applying LS plus PG and applying HL plus PG on soil fertility in the 0.0–0.2 m layer were maintained, T2 and T5 are good options for correcting soil acidity in agropastoral system. These two treatments also increased P content in the 0.1–0.8 m layer and OM content in the 0.2–0.8 m layer, which may promote crop root system development and plant water and nutrient absorption, thereby reducing the adverse effects of drought, improving crop tolerance to water stress, and increasing crop yields, especially in the Cerrado biome. Evaluations of the long-term effects of different methodologies for mitigating soil acidity are therefore encouraged.

5. Conclusions

The present study evaluated the residual effects of different methodologies for applying Ca compounds in an agropastoral system in an Ultisol with sandy texture two seasons after the last application. The correction of surface acidity (0.0–0.2 m) through increased pH and base saturation and reduced total acidity persisted, but the previously observed improvement of acidity in the 0.4–0.8 m layer after the incorporation of LS and subsurface application of HL in the 2017-2018 season did not. In addition, in the 0.0–0.2 m layer, the positive effects of applying LS plus PG on Ca²⁺ content and Ca²⁺/CEC and applying HL plus PG on Mg²⁺ content and Mg²⁺/CEC were maintained, but the positive effects of these amendments on S-SO₄ content throughout the soil profile (0.0–0.8 m) were not. Finally, residual positive effects of Ca compound application on P content in the 0.1–0.8 m layer and OM content in the 0.2–0.8 m layer were observed. These results demonstrate that Ca compound application can promote improvements in root development, especially in deep layers, and crop development. Because increasing the soil contents of Ca and P enables greater root exploration in deeper layers, the present findings have implications for the adoption of practices that minimize water stress in plants.