ABSTRACT
Sulfur (S) deficiency in grain crops has increased in recent decades due to increased average grain yields and reduced anthropogenic and natural S inputs to the soil. This study aimed to evaluate atmospheric S deposition, crop grain yields, and S availability in the soil profile over time following successive applications of agricultural gypsum and elemental S in a Typic Hapludox (Latossolo Bruno Distrófico) under a no-tillage system. The effects of annual applications of 75 kg ha-1 S either as elemental S or agricultural gypsum (CaSO4), as well as a control treatment, were evaluated on the dynamics of available S in the soil profile (down to 0.60 m) and on the grain yield of 27 crops. The crops were distributed across three sites in the Center-South region of Paraná, Brazil, with evaluation times ranging from 4 to 8 years. Soils in the 0.00-0.20 m layer at the three sites showed an available S content of approximately 15 mg dm-3 and organic matter (OM) content higher than 5 % upon setting up the study. To determine atmospheric S deposition, rainwater samples were collected weekly at the three study sites over a one-year period. The mean atmospheric S deposition across the three sites was ≈10 kg ha-1 yr-1, which is enough to replenish approximately 60 % of the S exported through the grain. Successive annual applications of elemental S were more effective than applications of agricultural gypsum in increasing the available S levels in the topsoil layer over time. Regardless of the source used, the greatest increases in available S levels occurred in the subsurface layers (>0.20 m). Only 4 out of 27 (≈15 %) crops evaluated showed increases (9.9 % on average) in grain yield through S fertilization, with no differences between the sources. Mean effect of annual S applications was less than 3 % on cumulative crop yield. Therefore, for Typic Hapludox with high OM content, available S content above the critical threshold, and contribution through atmospheric deposition of ≈10 kg ha-1 yr-1, successive annual S fertilization proves not to be necessary, even in high-yielding systems. Under these conditions, adoption of a fertilization strategy based on replenishing the S exported by crop harvests is more viable.
Keywords
sulfur fertilization; sulfur source; grain yield
INTRODUCTION
Sulfur (S) is a macronutrient that constitutes amino acids and proteins that are vital for plant growth and development (Aula et al., 2019). Despite its importance, S has been the least studied macronutrient because, historically, deficiencies of this nutrient were not of frequent occurrence (Kihara et al., 2017; Zenda et al., 2021). The primary source of S in the soil is organic matter (Narayan et al., 2023), which releases S into the soil solution at the time of its mineralization in the form of sulfate (SO42-). This sulfate can then be taken up by plants, leached, reduced, or reassimilated by the microbiota (Ercoli et al., 2012). Consequently, soils with high organic matter content can potentially release large amounts of SO42- into the soil solution.
Atmospheric S deposition through precipitation is another important S source for the soil (Feinberg et al., 2021), particularly in areas close to large industrial centers and to the ocean (Tiecher et al., 2013; Vieira-Filho et al., 2015). However, due to new industrial technologies (such as renewable energy) and stricter governmental inspection regulations, mean atmospheric S deposition has declined significantly in recent decades worldwide (Divito et al., 2015; Vieira-Filho et al., 2015; Feinberg et al., 2021). Therefore, the probability of positive crop responses to S fertilization has increased, both because of a reduction in atmospheric S deposition and the significant increase in crop yields (Salvagiotti et al., 2017; Aula et al., 2019; Pias et al., 2019).
Application of mineral fertilizers containing S is the main way to increase soil S reserves and thereby supplement plant requirements. Various fertilizers widely used in agriculture to supply other macronutrients contain S in their composition, such as ammonium sulfate (22 % S), single superphosphate (10 % S), potassium sulfate (15 % S), and agricultural gypsum (13 % S). However, in recent decades, due to logistics and the lower unit cost of NPK nutrients, growers have preferred the use of concentrated NPK fertilizers (urea, triple superphosphate, and potassium chloride), which have no S in their composition (Fiorini et al., 2016; Degryse et al., 2020). All the S sources cited above make S available soon after their application already in the form of SO42-, which is the form taken up by plants (Ercoli et al., 2012). A drawback of immediate availability of S as SO42- in the soil is the high probability of nutrient loss through leaching, due to its high mobility in the soil profile (Ercoli et al., 2012; Degryse et al., 2020), particularly in sandy soils and soils with high pH and available phosphorus levels in the topsoil layer (Dalla Nora and Amado, 2013; Tiecher et al., 2019).
Recently, the application of elemental S (95 % S) has gained popularity, which, after application, must be oxidized by soil microorganisms into SO42- (Horowitz and Meurer, 2006). Thus, elemental S has been placed on the market as a slow-release fertilizer to reduce loss through leaching, particularly in environments with high rainfall (Degryse et al., 2018). Previous studies have confirmed that elemental S oxidation occurs in all Brazilian soils; however, oxidation rates vary according to soil properties (such as OM content and acidity) and environmental conditions (such as aeration, moisture, and temperature) (Horowitz and Meurer, 2006; 2007).
Several studies have evaluated the effects of applying different S sources on crop yields under no-tillage systems (NTS), both sources with readily available SO42- and/or elemental S that requires oxidation (Broch et al., 2011; Fiorini et al., 2016; Lopes et al., 2017; Degryse et al., 2020). However, conflicting findings have been reported, making it impossible to define environments where it is more effective to use one source than another, thus requiring supplementary studies (Pias et al., 2019). In addition, yields of all major crops have increased significantly over recent decades (Conab, 2026). Consequently, in high crop yield areas, greater plant nutrient uptake and export occur, requiring higher available S content in these areas than in lower yield areas (Salvagiotti et al., 2017; Pias et al., 2019).
One of the limitations of studies on S fertilization available in the literature is that these studies are normally of short duration (Pias et al., 2019; Degryse et al., 2020). The present study provides an opportunity to visualize available soil S dynamics and grain yield of 27 crops as a result of annual applications of S from different sources at three sites, with times ranging from 4 to 8 years. We hypothesized that in high crop yield sites, S fertilization, even in soils with available S content above the critical threshold, can increase crop yield. In addition, the use of fertilizer sources with different rates of rendering SO42- available influences the dynamics of S availability in the soil profile. This study aimed to evaluate atmospheric S deposition, crop grain yield, and S dynamics in the soil profile (down to 0.60 m) over time after annual applications of agricultural gypsum and elemental S in Typic Hapludox under NTS in the Center-South region of Paraná.
MATERIALS AND METHODS
Study sites and treatments
The experiments were established in the Center-South region of Paraná, specifically in the municipality of Goioxim in the 2012/2013 crop season, Candói in the 2014 winter crop season, and Reserva do Iguaçu in the 2013/2014 crop season; and the experiments were conducted in the three sites up to the 2018/2019 crop season. A randomized block experimental design was used in the studies, which evaluated the effects of applying 75 kg ha-1 S either as elemental S (95 % S) or agricultural gypsum (15 % S), as well as a control treatment (without S application). Four replications were carried out in Goioxim and Reserva do Iguaçu and three replications in Candói. Sulfur fertilizers were manually applied annually on the soil surface, always during the fall off-season. The plot dimensions were 8 × 10 m (80 m2) in Goioxim and Candói and 9.6 × 7 m (67.2 m2) in Reserva do Iguaçu.
The climate of the study region is Cfb humid subtropical, according to the Köppen classification system, with mean annual precipitation of approximately 2,000 mm and mean temperature of 17 °C. The soil at the three sites is classified as a Typic Hapludox according to Soil Survey Staff (2014) and Latossolo Bruno Distrófico according to Santos et al. (2025). The areas had been managed under a NTS for more than 10 years and followed crop rotation systems with soybean and corn in the summer and wheat and barley in the winter. Soil chemical and physical properties at the time of establishing the experiments are shown in table 1. Soil at the three sites was generally found to have high chemical fertility, acidity levels suitable for crop development, and available S content above critical thresholds, both in the surface (>3 mg dm-3) and subsurface (>9 mg dm-3) layers (Pauletti and Motta, 2019).
Soil chemical and physical properties prior to establishment of the experiments at three sites in the Center-South region of Paraná in Typic Hapludox under a no-tillage system
Crop management and analyzed variables
Across the three study sites, we evaluated the response of 27 crops to S fertilization, specifically, 11 for soybean, 6 for corn, 4 for barley, 4 for wheat, and 2 for common bean. Information on fertilization at sowing and topdressing, row spacing, and plant density used for each crop is shown in table 2. At each site, phytosanitary treatments were performed according to the technical recommendations in effect for each crop, which were identical for all the treatments. Grain yield was determined by harvesting an area of 32 m2 (3.2 × 10 m) for soybean and common bean, 16 m2 for corn (1.6 × 10 m), and 34 m2 (3.4 × 10 m) for the winter cereal crops in each plot. Specialized harvesters were used for harvesting the plots. Grain moisture content was determined in the laboratory, and yield data were adjusted to 130 g kg-1.
Agronomic information for the crops grown to evaluate the effect of sulfur fertilization at three sites in the Center-South region of Paraná in Typic Hapludox under a no-tillage system
For the joint evaluation of the different crops, as recommended by Pearce et al. (2022), the mean relative grain yield (% of maximum yield) of each treatment was determined in relation to the treatment with the maximum yield for each crop, according to equation 1. Subsequently, the mean relative yields (RY) of all the crops were calculated for each site (Goioxim, Reserva do Iguaçu, and Candói), and across all three study sites.
Available S contents in the 0.00–0.10, 0.10–0.20, 0.20–0.40, and 0.40–0.60 m soil layers were assessed annually, always after harvest of the summer crop and before subsequent S application, for four years in Reserva do Iguaçu, five years in Candói, and seven years in Goioxim. Stratified samples were collected manually using a core auger, with collection of nine subsamples per plot to form a composite sample. Available S content in the soil was determined in the laboratory through extraction of S with 0.01 mol L-1 calcium phosphate [Ca(PO4)2] and quantification by ICP (CQFS-RS/SC, 2016). Consequently, the available S content in the soil includes the sulfate desorbed by the extractant and other forms of S that may also be extracted, such as part of the soil organic S.
To determine wet atmospheric S deposition (>90 % of the total deposition), bulk samplers used for rainwater collection and storage were set up at the three study sites. From October 2017 to October 2018, the samplers were checked weekly to determine the precipitation that occurred in the period and to collect rainwater samples (30 mL), which were frozen until processing in the laboratory. To determine the S concentration of the aliquots, they were initially filtered through a 0.22-μm cellulose membrane, with subsequent quantification via ICP.
Statistical analysis
Data obtained were initially subjected to analysis of variance using the F-test at a 5 % significance level, and the mean values of the effects of the S sources were compared using Tukey’s test at the 5 % level. To evaluate the effect of years of successive S applications on the available S content in the soil, polynomial regression equations were fitted at the 5 % level. Statistical analyses were performed using the SISVAR software (Ferreira, 2011). To compare the effect of the treatments on relative crop yield, confidence intervals were determined through the bootstrapping procedure (4999 repetitions) using the Statkey 2.0.1 software (Lock et al., 2026).
RESULTS
Atmospheric S deposition
Results show little variation in atmospheric S deposition among the studied sites, which is likely due to the short distance between them (<100 km). In Reserva do Iguaçu, the deposition was 8.9 kg ha-1 yr-1 S, while Goioxim showed the highest value of atmospheric S deposition at 10.4 kg ha-1 yr-1 (Figure 1).
Atmospheric sulfur deposition and precipitation from October 2017 to October 2018, at three sites in the Center-South region of Paraná in soils under a no-tillage system.
Available S content along the soil profile
The experiment in Goioxim was conducted for seven years. In the topsoil layer (0.00–0.10 m), S application via agricultural gypsum did not alter the available S content compared with the control; both treatments showed a reduction in S availability in the first three years and then stabilized at values from 7 to 20 mg dm-3 (Figure 2a). In contrast, the application of elemental S was more effective in maintaining stable available S content in the topsoil layer (mean of 35 mg dm-3) throughout the period of conducting the experiment, and S content differed statistically from the other treatments from the fourth year of the study on.
Available S contents in the 0.00–0.10 (a), 0.10–0.20 (b), 0.20–0.40 (c), and 0.40–0.60 m (d) soil layers and along the soil profile after 1 year (e) and 7 years (f) of annual applications of 75 kg ha-1 S using agricultural gypsum and elemental S as sources in Goioxim, Paraná. Vertical black lines indicate the minimum significant difference between the treatments, and means followed by the same letters indicate no difference using Tukey’s test at the 5 % significance level. Red dashed lines represent the soil critical levels (Pauletti and Motta, 2019). ns: non-significant.
In the 0.10–0.20 m soil layer, no significant difference was found in the available S content among the applied sources (Figure 2b); however, a difference was found between the treatments that received S and the control from the fifth year of the study onwards. In the control treatment, the available S content decreased linearly, with a mean content of 5.8 mg dm-3 in the final year of the study (reaching a value one-sixth of that at the beginning). Yet, this value was still above the critical level (3 mg dm-3) (Pauletti and Motta, 2019). Subsurface layers (Figures 2c and 2d) likewise showed no significant difference in available S content between the treatments that received S application; however, both these treatments differed from the control from the fourth and fifth year of the study for the 0.20–0.40 and 0.40–0.60 m layers, respectively. Control treatment exhibited a reduction in the available S content in both subsurface layers, and already from the fourth year on, the content was less than ideal for crop development in the 0.20-0.40 m layer (<9 mg dm-3) (Pauletti and Motta, 2019). Figures 2e and 2f show the available S contents along the soil profile at one and seven years after the beginning of the experiment, respectively. As the years pass, the distance between treatments increases, where elemental S retained statistically higher available S content in the topsoil compared with agricultural gypsum and the control.
In Reserva do Iguaçu, four soil collections were performed over the period of the study. In the topsoil layer (0.00-0.10 m; Figure 3a), the use of elemental S increased the available S content in the soil in relation to the control in the collections made at one, three, and four years after the beginning of the study. In contrast, the available S content in the treatment that received the agricultural gypsum did not differ from the control treatment, characterizing leaching of the nutrient to subsurface layers. In the 0.10–0.20 m layer (Figure 3b), both the use of agricultural gypsum and elemental S effectively increased the available S content in the soil. In this layer, the available S content in the soil in the treatment that received elemental S statistically differed from the treatment with agricultural gypsum only in the fourth year of the study.
Available S contents in the 0.00–0.10 (a), 0.10–0.20 (b), 0.20–0.40 (c), and 0.40–0.60 m (d) soil layers and along the soil profile after 1 year (e) and 4 years (f) of annual applications of 75 kg ha-1 S using agricultural gypsum and elemental S as sources in Reserva do Iguaçu, Paraná. Vertical black lines indicate the minimum significant difference between the treatments, and means followed by the same letters indicate no difference using Tukey’s test at the 5 % significance level. Red dashed lines represent the soil critical levels (Pauletti and Motta, 2019). ns: non-significant.
In the subsurface layers (0.20–0.40 and 0.40–0.60 m), the use of agricultural gypsum significantly increased the available soil S content compared with the control throughout the study period (Figures 3c and 3d). Meanwhile, the available soil S content in the elemental S treatment differed from the control beginning with the second year of evaluation. The control treatment, which had a mean available S content of 22 mg dm-3 in the first year, exhibited reduction in content over the years, reaching 8.7 mg dm-3 in the fourth year, a value below the critical level of 9 mg dm-3 for the 0.20–0.40 m layer (Pauletti and Motta, 2019). Comparison of the available S content along the profile from the first year (Figure 3e) to the fourth year (Figure 3f) shows a sharp reduction in the topsoil layers for both the control treatment and the agricultural gypsum treatment.
In the experiment in Candói, the available S content in the soil topsoil layer (0.00–0.10 m) did not fluctuate significantly over the five years of the study, even with the annual S fertilizer applications (Figure 4a). However, the soil treated with elemental S maintained mean available S content (24.87 mg dm-3) much greater than the other treatments (11.73 for the control treatment and 13.87 mg dm-3 for the agricultural gypsum treatment), with statistical differences among the treatments only in the fourth year of the study. Application of agricultural gypsum did not increase the available S content in the topsoil compared with the control treatment.
Available S contents in the 0.00–0.10 (a), 0.10–0.20 (b), 0.20–0.40 (c), and 0.40–0.60 m (d) soil layers and along the soil profile after one year (e) and 5 years (f) of annual applications of 75 kg ha-1 S using agricultural gypsum and elemental S as sources in Candói, Paraná. Vertical black lines indicate the minimum significant difference between the treatments, and means followed by the same letters indicate no difference using Tukey’s test at the 5 % significance level. Red dashed lines represent the soil critical levels (Pauletti and Motta, 2019). ns: non-significant.
In the 0.10–0.20 m soil layer, the mean available S concentrations in the treatments that received S fertilization did not change over the years of the study and did not differ between the sources used, remaining between 25 and 30 mg dm-3 on average (Figure 4b). However, the available S content in the control treatment decreased linearly, from 22 mg dm-3 in the first year to 9.93 mg dm-3 in the fifth year of the study, which represents a 55 % of reduction. In the subsurface layers (below 0.20 m), the annual applications of both elemental S and agricultural gypsum increased the available S content in the soil, without statistical difference between the sources (Figures 4c and 4d). Nevertheless, both treatments differed from the control beginning with the third year of the study, reaching content greater than 40 mg dm-3 in the fifth year. Regardless of the S source applied, available S accumulated in the layers between 0.20-0.60 m (Figures 4e and 4f).
Grain yield as a function of S application via agricultural gypsum and elemental S
No significant difference was found in crop yield through the use of the different S sources (Table 3). Sulfur application increased grain yield in only 4 out of the 27 crops evaluated in the three sites, and all these 4 cases occurred in the experiment in Goioxim, with a mean increase of 9.9 % in these crops.
Crop grain yield (Mg ha-1) as a function of application of 75 kg ha-1 of sulfur from agricultural gypsum (CaSO4) or elemental S at three sites in the Center-South region of Paraná in Typic Hapludox under a no-tillage system
Combined evaluation of crop response in Goioxim shows significant differences between the treatments that received S and the control (Figure 5a, confidence intervals do not overlap). In Goioxim, the control yielded on average 6 % less than the S fertilization treatments. In Reserva do Iguaçu, the mean relative yield of the treatments was practically equal, without statistical difference (Figure 5b). In the Candói experiment, the relative yield of the elemental S treatment differed from the control, but not from the agricultural gypsum treatment (Figure 5c). In the mean of the 27 crops, the S application treatments exhibited a grain yield increase of less than 3 % (Figure 5d).
Mean relative yield (%) of the crops at the different sites as a function of annual application of 75 kg ha-1 sulfur using agricultural gypsum and elemental S as sources in Goioxim (a), Reserva do Iguaçu (b), Candói (c), and the overall mean of the evaluated sites (d). The center box indicates the range between the upper and lower confidence intervals of the mean value. The same letters among the treatments indicate the absence of a significant difference determined from the confidence intervals. Solid line within the box indicates the mean value. Vertical lines above and below the box indicate the minimum and maximum values, respectively.
Confirming the previous results in which a significant effect of S application was not observed for most of the crops, figure 6 shows the absence of a relationship between the available soil S content in the different layers and the relative crop yield.
Relationship between relative crop yield in Goioxim (Goi.), Reserva do Iguaçu (Res), and Candói (Can.) and the available S content (mg dm-3) in the 0.00–0.10 (a), 0.10–0.20 (b), 0.20–0.40 (c), and 0.40–0.60 m (d) soil layers. Control: Cont; Agricultural gypsum: Gyp; and Elemental S: El-S. Red dashed lines represent the critical levels in the soil (Pauletti and Motta, 2019). N: 81 observations from the 27 crops.
DISCUSSION
The present study monitored wet atmospheric S deposition over a one-year period across the three study sites. The mean results were approximately 10 kg ha-1 yr-1 (Figure 1), and a positive relationship can be observed between the total rainfall volume (≈2000 mm yr-1) and atmospheric S deposition. However, the S concentration in rainwater tends to decline as the rainfall amount increases (dilution effect), as observed by Filippi et al. (2024). These atmospheric S deposition values partially explain the lack of a positive response to S fertilization in most of the crops evaluated in the present study and in several other studies, even when conducted in S-deficient soils (Rheinheimer et al., 2005; Frandoloso et al., 2010; Pias et al., 2019).
Data on atmospheric S deposition in Brazilian agricultural areas are lacking, which compromises S recommendations, particularly prediction of whether or not a positive response to S fertilization will occur (Pias et al., 2019). In São Paulo, for example, a highly industrialized region, annual deposition in 2014 was 14 ha-1 yr-1 of S, while at the same location in 1985, it was 285 ha-1 yr-1 of S (Vieira-Filho et al., 2015). However, in Santa Maria, an inland city in Rio Grande do Sul, the atmospheric deposition was 4.5 kg ha-1 yr-1 of S from 2007 to 2009 (Tiecher et al., 2013). Recently, Filippi et al. (2024) reported atmospheric S deposition of 3.3 ha-1 yr-1 of S in Eldorado do Sul, a city near Porto Alegre, RS. Evaluating S deposition over the year, Filippi et al. (2024) observed that from April to September, atmospheric S deposition was 43 % higher than that from October to March. These studies show high spatial and temporal variability in atmospheric S deposition. In global projections made by Feinberg et al. (2021), atmospheric S deposition in South America is expected to decline between 20-40 % by the end of the century. Although atmospheric S deposition has decreased over the years in Brazil and in the world, it is still significant and important for supplementing the amounts of S supplied by the soil for plant nutrition. Based on the data obtained in the present study, a mean atmospheric S deposition of 10 kg ha-1 yr-1 is sufficient to replenish the S exported through production of 3.5 Mg ha-1 of soybean grain (Filippi et al., 2021) or 10 Mg ha-1 of corn grain (Duarte et al., 2019).
Data from the three sites show that the soils contained high available S content along the soil profile at the beginning of the experiment (Table 1). In the absence of replenishment of the S exported via crop grains and of S losses to the environment, the concentrations reduced across all layers of the soil profile. These values even dropped below optimal in the 0.20-0.40 m layer over a four-year period in Goioxim (Figure 2c) and Reserva do Iguaçu (Figure 3c). This result was expected due to leaching of the nutrient to deeper layers and to depletion from the soil through S export without replenishment.
The time required for the soil to express S deficiency can vary greatly depending on management conditions, soil properties (clay and organic matter content), and environmental conditions (precipitation) (Filippi et al., 2024). Thus, monitoring available S content is indispensable to prevent S deficiency in crops. In the experiment in Reserva do Iguaçu, a reduction in available S content in the soil was reported in the first two years of the study, regardless of the treatment (Figure 3). This result may be related to the management practices that were carried out before the experiment began, with high application rates of single superphosphate for phosphorus correction, as the fertilizer contains 10 % S in its composition.
In the topsoil (0.00-0.20 m layer) at the three study sites, the control treatment, which received no S application over the years, maintained available S content above the critical level (3 mg dm-3) (Pauletti and Motta, 2019). This result is related to the S supply capacity of the organic matter, which is the primary source of this nutrient in the soil, as well as atmospheric S deposition. To demonstrate the importance of soil organic matter on soil S availability, Filippi et al. (2024) observed an increase of 0.85 mg dm-3 in available S for every 1 g kg-1 increase in soil organic matter content. Elemental S was more effective than agricultural gypsum at maintaining higher available S content in the topsoil (Figures 2a, 3a, and 4a), which is mainly due to its lower leaching potential (Ercoli et al., 2012) and lower immobilization by microorganisms (Degryse et al., 2020). This is because S must first be oxidized to SO42-, which has greater mobility in the soil profile (Ercoli et al., 2012). Oxidation rate of elemental S depends on a series of abiotic factors that influence microbiological activity (Horowitz and Meurer, 2007; Degryse et al., 2020). In a study conducted by Degryse et al. (2020) using the S34 isotope, greater efficiency of fertilization was reported with elemental S compared with S-SO42-; the mean amount recovered by plants was 9 % for S-SO42- and 16 % for elemental S in relation to the total amount applied (78 % greater efficiency).
Annual S applications showed that the soil accumulates S more readily in subsurface layers than in the topsoil. This result is associated with the higher adsorption capacity of the nutrient into functional groups (e.g., Fe and Al oxides and kaolinite), due to lower pH levels and reduced competition with phosphate for adsorption sites (Rheinheimer et al., 2005). According to Alvarez et al. (2007), in addition to the formation of inner-sphere complexes, S in the form of SO4-2, which is the primary inorganic form in the soil, may also be retained electrostatically by soil positive charges, which are also more prevalent in the subsurface than in topsoil layers. Moreover, S in the topsoil is more susceptible to incorporation into the microbial biomass (immobilization), since microbiological activity is greater in the topsoil layers (Degryse et al., 2020).
As a significant part of available soil S is below the 0.20 m level, evaluating the subsurface layer for fertilizer recommendation purposes is essential [which is already being done in the Cerrado (Rein and Sousa, 2004) and in Paraná (Pauletti and Motta, 2019)], as well as managing the soil to prevent chemical and/or physical barriers that limit the depth of crop root systems under NTS (Dalla Nora and Amado, 2013). In a study conducted by Aula et al. (2019) evaluating the overall efficiency of S fertilizer use on cereals, results showed low efficiency, an average of only 18 %. Thus, improvements in fertilizer management are necessary for greater sustainability of the system. These management improvements necessarily involve choosing the right S source, applying the appropriate rate, and selecting the time of greatest plant demand, always striving for greater synchrony between soil S availability and plant requirements. The use of conservationist management practices that provide increased organic matter content, combined with a crop rotation system that includes legumes, is a promising practice to maintain high levels of available soil S (Filippi et al., 2024).
Sulfur fertilization, regardless of the source, had little impact on crop yield. Only 15 % of the evaluated crops exhibited yield increases, and the mean increase in these crops was 9.9 % (Table 3). In a systematic review of crop response to S fertilization in Brazil, Pias et al. (2019) observed an average increase of 19 % in the crops that showed a positive response. In Argentina, the mean increase was 13 % (Salvagiotti et al., 2017), and in Africa, the mean increase reached 35 % (Kihara et al., 2017) – values well above those observed in this study. Joint evaluation of the crop yields shows an increase from S application of less than 3 % (Figure 5d). An explanation for the low effect of S application on yield is that although this was a medium- to long-term study (four to seven years), the available S content remained above the critical threshold (Pauletti and Motta, 2019) in all the plots for the topsoil and in most of the plots for the 0.20-0.40 m layer. This did not lead to a positive correlation between the available soil S content and crop yield (Figure 6). The only site where the crops responded positively to S application was Goioxim, and this occurred in 4 out of 13 crops. From evaluation of the data in figure 6c, which shows the lack of correlation between available S content in the 0.20-0.40 m layer and relative yield, it can be seen that only 5 treatments exhibited mean S values below the critical threshold (9 mg dm-3), and all of them belonged to the control treatment of the Goioxim experiment. The lack of correlation between the available S content and crop yield is due to the high initial available S and organic matter content of the studied soils, as already reported by Pias et al. (2019) in a systematic review for soils throughout Brazil. Although the subsurface layer became S deficient over the course of the study, in the topsoil layer, the available S content remained above the critical threshold, thereby adequately nourishing the plants. This result shows that, in most cases, other factors limited crop yield, not soil S availability. If the study were conducted for a longer time, with greater reduction in available S content in the soil, a positive relationship between S fertilization and crop yield might be observed.
In the present study, the mean yield of the crops at the three studied sites was high, particularly for corn (13.6 Mg ha-1) and soybean (4.7 Mg ha-1), which exceeded the mean yields of the crops in the state of Paraná by 113 and 57 %, respectively (2018/2019 crop season, Conab, 2026). Therefore, the initial hypothesis – that S fertilization would increase yields under high-yielding conditions even if available S content in the soil was above the critical threshold – was rejected. This result corroborates the findings of Pias et al. (2019), who analyzed the response of 57 crops to S fertilization throughout Brazil and observed no positive response in any crop when the available S content was above the critical threshold.
Few studies in Brazil have evaluated S sources (Pias et al., 2019). Among the available studies, some showed no significant difference among the sources (Fiorini et al., 2016), corroborating the findings of the present study (Table 3). However, other studies observed greater efficiency for specific fertilizer sources, depending on the characteristics of the crop environment (Broch et al., 2011). As elemental S did not differ from agricultural gypsum regarding grain yield in the present study, it can be inferred that its use, in combination with concentrated fertilizers (NPK), may be an effective strategy for reducing transportation and application costs (Horowitz and Meurer, 2007). Degryse et al. (2020) confirmed greater efficiency in the use of elemental S compared with an S-SO42- source in a tropical climate, particularly due to its greater residual effect.
According to Järvan et al. (2008), improvements in grain quality of cereal crops are more commonly observed than yield responses following the application of S fertilizers. Supplying S to wheat, for example, increases the grain gluten content and the biological quality of the proteins, due to an increase in the proportions of cysteine and methionine, essential amino acids for flour quality in the breadmaking process (Järvan et al., 2008). Therefore, even in the absence of yield increases in most of the evaluated crops, enhancements may have occurred in grain quality, which warrants future investigations on this subject.
The lack of effect from S fertilization on most of the crops is directly related to the high OM content, which exceeded 5.5 % in the topsoil layers at the three sites when the experiments were established (Table 1). In soils with high clay and OM contents, the probability of obtaining a response to S fertilization is reduced (Rheinheimer et al., 2005; Filippi et al., 2024), as 90–95 % of total soil S is in organic forms (Sutar et al., 2017). Degryse et al. (2020) highlighted the difficulty of predicting positive responses to S application, as the mineralization rate of the organic matter over the crop cycle depends on many factors, and a precise estimate is not possible. In addition to OM, this study shows the importance of atmospheric S deposition, which also assists in replenishing the S exported by crops in the control treatment. Over the crop years, atmospheric S deposition alone replenished around 49, 59, and 64 % of the S exported through grains harvested in Reserva do Iguaçu (73 kg ha-1 of S exported), Candói (78 kg ha-1 of S exported), and Goioxim (113 kg ha-1 of S exported), respectively. Consequently, efforts should be made to map atmospheric S deposition in agricultural areas and thereby contribute to more assertive management.
CONCLUSIONS
Successive annual applications of elemental S were more effective than applications of agricultural gypsum at increasing the available S content in the topsoil layer over time, which is due to the need of be first oxidized in the soil to render the SO42- available. Annual S applications using both agricultural gypsum and elemental S increased S availability in the soil subsurface layers (0.20-0.60 m). Even in soils with high clay and organic matter content, not applying S over multiple crop seasons (4 years) reduced available S across all soil layers, making the soil subsurface layer S deficient in two out of the three studied sites. Although the soil became S deficient in the subsurface, no correlation was found between available S content and crop grain yield.
Increased grain yield from S fertilization was found in only 4 out of 27 (≈15 %) of the evaluated crops, and the mean increase in those specific crops was 9.9 %. The agricultural gypsum and elemental S sources did not lead to differences in regard to crop grain yield. The average effect of annual S applications on cumulative crop yield was less than 3 %. Therefore, in soils with available S above the critical threshold, recommendation of successive annual S fertilization is not economically viable, even in high-yield environments, where part of S demand (≈59 %) is met by atmospheric S deposition (≈10 kg ha-1 yr-1). In this scenario, S fertilization management should aim at replenishing the S exported in harvested grains and continuously monitor available S content in the soil to ensure its capacity to supply S to crops.
How to cite:
DATA AVAILABILITY
The data will be provided upon request.
REFERENCES
- Alvarez VVH, Roscoe R, Kurihara CH, Pereira NF. Enxofre. In: Novais RF, Alvarez VVH, Barros NF, Fontes RLF, Cantarutti RB, Neves JCL, editors. Fertilidade do solo. Viçosa, MG: Sociedade Brasileira de Ciência do Solo; 2007. p. 596-644.
-
Aula L, Dhillon JS, Omara P, Wehmeyer GB, Freeman KW, Raun WR. World sulfur use efficiency for cereal crops. Agron J. 2019;111:2485-92. https://doi.org/10.2134/agronj2019.02.0095
» https://doi.org/10.2134/agronj2019.02.0095 -
Broch DL, Pavinato PS, Possentti JC, Martin TN, Quiqui EMD. Produtividade da soja no cerrado influenciada pelas fontes de enxofre. Rev Cienc Agron. 2011;42:791-6. https://doi.org/10.1590/S1806-66902011000300027
» https://doi.org/10.1590/S1806-66902011000300027 - Comissão de Química e Fertilidade do Solo - CQFS-RS/SC. Manual de calagem e adubação para os Estados do Rio Grande do Sul e de Santa Catarina. 11. ed. Porto Alegre: Sociedade Brasileira de Ciência do Solo - Núcleo Regional Sul; 2016.
-
Companhia Nacional de Abastecimento - Conab. Séries históricas de safras agrícolas. Brasília, DF: Conab; 2026 [cited 2026 Feb 15]. Available from: https://www.gov.br/conab/pt-br/atuacao/informacoes-agropecuarias/safras/series-historicas-das-safras
» https://www.gov.br/conab/pt-br/atuacao/informacoes-agropecuarias/safras/series-historicas-das-safras -
Dalla Nora D, Amado TJC. Improvement in chemical attributes of oxisol subsoil and crop yields under no-till. Agron J. 2013;105:1393-403. https://doi.org/10.2134/agronj2013.0031
» https://doi.org/10.2134/agronj2013.0031 -
Degryse F, Baird R, Silva RC, Holzapfel CB, Kappes C, Tysko M, McLaughlin MJ. Sulfur uptake from fertilizer fortified with sulfate and elemental S in three contrasting climatic zones. Agronomy. 2020;10:1035. https://doi.org/10.3390/agronomy10071035
» https://doi.org/10.3390/agronomy10071035 -
Degryse F, Silva RC, Baird R, Beyrer T, Below F, McLaughlin MJ. Uptake of elemental or sulfate-S from fall- or spring-applied co-granulated fertilizer by corn—A stable isotope and modeling study. Field Crop Res. 2018;221:322-32. https://doi.org/10.1016/j.fcr.2017.07.015
» https://doi.org/10.1016/j.fcr.2017.07.015 -
Divito GA, Echeverría HE, Andrade FH, Sadras VO. Diagnosis of S deficiency in soybean crops: Performance of S and N:S determinations in leaf, shoot and seed. Field Crop Res. 2015;180:167-75. https://doi.org/10.1016/j.fcr.2015.06.006
» https://doi.org/10.1016/j.fcr.2015.06.006 -
Duarte AP, Abreu MF, Francisco EAB, Gitti DC, Barth G, Kappes C. Reference values of grain nutrient content and removal for corn. Rev Bras Cienc Solo. 2019;43:e0180102. https://doi.org/10.1590/18069657rbcs20180102
» https://doi.org/10.1590/18069657rbcs20180102 -
Ercoli L, Arduini I, Mariotti M, Lulli L, Masoni A. Management of sulphur fertiliser to improve durum wheat production and minimise S leaching. Eur J Agron. 2012;38:74-82. https://doi.org/10.1016/j.eja.2011.12.004
» https://doi.org/10.1016/j.eja.2011.12.004 -
Feinberg A, Stenke A, Peter T, Hinckley ELS, Driscoll CT, Winke LHE. Reductions in the deposition of sulfur and selenium to agricultural soils pose risk of future nutrient deficiencies. Commun Earth Environ. 2021;2:101. https://doi.org/10.1038/s43247-021-00172-0
» https://doi.org/10.1038/s43247-021-00172-0 -
Ferreira DF. Sisvar: a computer statistical analysis system. Cienc Agrotec. 2011;35:1039-42. https://doi.org/10.1590/S1413-70542011000600001
» https://doi.org/10.1590/S1413-70542011000600001 -
Filippi D, Alves LA, Flores JPM, Ambrosini VG, Bayer C, Bissani CA, Tiecher TL, Gianello C, Tiecher T. Legume cover crops combined with no-tillage increase extractable sulfur in a subtropical Acrisol. Eur J Agron. 2024;153:127049. https://doi.org/10.1016/j.eja.2023.127049
» https://doi.org/10.1016/j.eja.2023.127049 -
Filippi D, Denardin LGO, Ambrosini VG, Alves LA, Flores JPM, Martins AP, Pias OHC, Tiecher T. Concentration and removal of macronutrients by soybean seeds over 45 years in Brazil: A meta-analysis. Rev Bras Cienc Solo. 2021;45:e0200186. https://doi.org/10.36783/18069657rbcs20200186
» https://doi.org/10.36783/18069657rbcs20200186 -
Fiorini IVA, Pinho RGV, Pires LPM, Santos AO, Fiorini FVA, Cancellier LL, Resende EL. Avaliação de fontes de enxofre e das formas de micronutrientes revestindo o NPK na cultura do milho. Rev Bras Milho Sorgo. 2016;15:20-9. https://doi.org/10.18512/1980-6477/rbms.v15n1p20-29
» https://doi.org/10.18512/1980-6477/rbms.v15n1p20-29 -
Frandoloso JF, Lana MC, Fontaniva S, Czycza RV. Eficiência de adubos fosfatados associados ao enxofre elementar na cultura do milho. Rev Ceres. 2010;57:686-94. https://doi.org/10.1590/S0034-737X2010000500019
» https://doi.org/10.1590/S0034-737X2010000500019 -
Horowitz N, Meurer EJ. Oxidação do enxofre elementar em solos tropicais. Cienc Rural. 2006;36:822-8. https://doi.org/10.1590/S0103-84782006000300015
» https://doi.org/10.1590/S0103-84782006000300015 -
Horowitz N, Meurer EJ. Relação entre atributos de solos e oxidação de enxofre elementar em quarenta e duas amostras de solos do Brasil. Rev Bras Cienc Solo. 2007;31:455-63. https://doi.org/10.1590/S0100-06832007000300005
» https://doi.org/10.1590/S0100-06832007000300005 - Järvan M, Edesi L, Adamson A, Lukme L, Akk A. The effect of sulphur fertilization on yield, quality of protein and baking properties of winter wheat. Agron Res. 2008;6:459-69.
-
Kihara J, Sileshi GW, Nziguheba G, Kinyua M, Zingore S, Sommer R. Application of secondary nutrients and micronutrients increases crop yields in sub-Saharan Africa. Agron Sustain Dev. 2017;37:25. https://doi.org/10.1007/s13593-017-0431-0
» https://doi.org/10.1007/s13593-017-0431-0 - Lock RH, Lock PF, Morgan KL, Lock EF, Lock DF. Statistics: Unlocking the power of data. 4th ed. Hoboken, NJ: Wiley; 2026.
-
Lopes LO, Moura MCS, Matias SSR, Lopes RS, Teixeira MPR. Effect of sources and doses of sulfate on soybean crop. Pesq Agropec Trop. 2017;47:296-302. https://doi.org/10.1590/1983-40632016v4745394
» https://doi.org/10.1590/1983-40632016v4745394 -
Narayan OP, Kumar P, Yadav B, Dua M, Johri AK. Sulfur nutrition and its role in plant growth and development. Plant Signal Behav. 2023;18:2030082. https://doi.org/10.1080/15592324.2022.2030082
» https://doi.org/10.1080/15592324.2022.2030082 -
Pearce AW, Slaton NA, Lyons SE, Bolster CH, Bruulsema TW, Grove JH, Jones JD, McGrath JM, Miguez FE, Nelson NO, Osmond DL, Parvej MR, Pena-Yewtukhiw EM, Spargo JT. Defining relative yield for soil test correlation and calibration trials in the Fertilizer Recommendation Support Tool. Soil Sci Soc Am J. 2022;86:1338-53. https://doi.org/10.1002/saj2.20450
» https://doi.org/10.1002/saj2.20450 -
Pias OHC, Tiecher T, Cherubin MR, Mazurana M, Bayer C. Crop yield responses to sulfur fertilization in Brazilian no-till soils: A systematic review. Rev Bras Cienc Solo. 2019;43:e0180078. https://doi.org/10.1590/18069657rbcs20180078
» https://doi.org/10.1590/18069657rbcs20180078 - Rein TA, Sousa DMG. Adubação com Enxofre. In: Sousa DMG, Lobato E, editors. Cerrado: Correção do solo e adubação. Planaltina, DF: Embrapa Cerrados, 2004. p. 227–242.
-
Rheinheimer DS, Alvarez JWA, Osorio Filho BD, Silva LS, Bortoluzzi EC. Resposta de culturas à aplicação de enxofre e a teores de sulfato num solo de textura arenosa sob plantio direto. Cienc Rural. 2005;35:562-9. https://doi.org/10.1590/S0103-84782005000300011
» https://doi.org/10.1590/S0103-84782005000300011 -
Salvagiotti F, Prystupa P, Ferraris G, Couretot L, Magnano L, Dignani D, Gutiérrez-Boem FH. N:P:S stoichiometry in grains and physiological attributes associated with grain yield in maize as affected by phosphorus and sulfur nutrition. Field Crop Res. 2017;203:128-38. https://doi.org/10.1016/j.fcr.2016.12.019
» https://doi.org/10.1016/j.fcr.2016.12.019 - Santos HG, et al. Sistema brasileiro de classificação de solos. 6. ed. Brasília, DF: Embrapa Solos, 2025.
- Soil Survey Staff. Keys to soil taxonomy. 12th ed. Washington, DC: United States Department of Agriculture, Natural Resources Conservation Service; 2014.
-
Sutar RK, Pujar AM, Kumar BNA, Hebsur NS. Sulphur nutrition in maize – A critical review. Int J Pure App Biosci. 2017;5:1582-96. https://doi.org/10.18782/2320-7051.6092
» https://doi.org/10.18782/2320-7051.6092 -
Tiecher T, Denardin LGO, Martins AP, Pias OHC, Nunes LR, Tiecher TL, Alvarez JWR, Mallmann FJK, Brunetto G, Santos DR. Dynamics of sulfate and basic cations in soil solution as affected by gypsum fertilization in an Ultisol of Southern Brazil. Arch Agron Soil Sci. 2019;65:1998-2012. https://doi.org/10.1080/03650340.2019.1587748
» https://doi.org/10.1080/03650340.2019.1587748 -
Tiecher T, Santos DR, Alvarez JWR, Mallmann FJK, Piccin R, Brunetto G. Respostas de culturas à adubação sulfatada e deposição de enxofre atmosférico. Rev Ceres. 2013;60:420-7. https://doi.org/10.1590/S0034-737X2013000300016
» https://doi.org/10.1590/S0034-737X2013000300016 -
Vieira-Filho MS, Lehmann C, Fornaro A. Influence of local sources and topography on air quality and rainwater composition in Cubatão and São Paulo, Brazil. Atmos Environ. 2015;101:200-8. https://doi.org/10.1016/j.atmosenv.2014.11.025
» https://doi.org/10.1016/j.atmosenv.2014.11.025 -
Zenda T, Liu S, Dong A, Duan H. Revisiting sulphur—The once neglected nutrient: Its roles in plant growth, metabolism, stress tolerance and crop production. Agriculture. 2021;11:626. https://doi.org/10.3390/agriculture11070626
» https://doi.org/10.3390/agriculture11070626












