Increasing Levels of Sulfur on Lowland Rice in Different Soil Textures

A viable alternative to increase productivity in rice cultivation areas has been the investment in macronutrient fertilization. Sulfur (S) is an essential nutrient for plants, as it participates in the formation of several organic molecules, being essential to ensure high yields of rice. The aim of this paper is to analyze the rice crop performance in two types of irrigated floodplain soils under different doses of sulfur fertilization. The experiment was carried out on two lowland soils in Formoso do Araguaia-Tocantins, Brazil, under a randomized block design, in a 2 × 8 factorial arrangement, two soil types and eight sulfur doses, using ammonium sulfate as the source of S (0, 10, 20, 30, 40, 50, 60 and 70 kg ha of S). The doses were applied when the plants were in the V3-V4 development stage in both soils. S provided plant a growth up to 40 kg ha, which is the economical feasible dose for plant height in soil 1. With the increase of the dose of S over 60 kg ha in soil 2, there was a decrease in the number of panicles per m. There was an increase in productivity in soil 2 and reduction in soil 1. The highest grain yield was obtained in soil with higher clay content. The dynamics of sulfur fertilization in floodplain soils requires further studies for a better understanding.


Introduction
The production of rice (Oryza sativa L.) in Brazil occurs under dry and irrigated cultivation.These different forms of production must be evidenced, since technological packages provide differentiated levels of production.In the State of Tocantins, rice cultivation occurs mainly under irrigated conditions, where the productive poles are in the Southwest of the State, with the largest producers being the cities of Lagoa da Confusão and Formoso do Araguaia, which adopt the irrigation system by flood during the season crop period.
The lowland soils are very heterogeneous, thus resulting in different behavior when compared to other soils submitted to the productive process (Bastos, Carvalho, Vieira, & Bertoni, 2005).When lowland soils are flooded, several reactions are affected, especially regarding nutrient uptake and availability (Taiz & Zeiger, 2004), which directly affects crop growth.
Among the nutrients that may undergo alterations in the rice production process by flooding, Sulfur (S) can be oxidized and occur in two chemical forms, Sulfides (H 2 S 2 ) and elemental sulfur.However, the plants absorb the element in the aerobic solution of the soil in the inorganic form of anion sulfate (S-SO 4 2-) (Casagrande, Alleoni, Camargo, & Borges, 2003).
The available S-SO 4 2-contents are directly related to soil pH and organic matter, clay and oxides, and the highest probability of responses to sulfate fertilization occurs in soils with high pH, low content of clay and organic matter that receive intensive cultivation associated with the use of concentrated fertilizers without S (Rheinheimer, Rasche, Osório Filho, & Silva, 2007).In addition, the ability of the soil to supply the nutrient demand of the plant is closely related to the organic matter content of the soil and the mineralization of organic S to inorganic forms (Tiecher et al., 2013).However, in the soil solution, sulfate is easily leached, and retained by the mechanisms of adsorption by coordination (Ishiguro & Makino, 2011).Thus, assuming that the pH and amount of organic matter can influence the uptake of S in rice plants which are cultivated in lowland soils, interfering with productivity.Thus, the aim of the study was to analyze the rice crop performance in two soils of the lowland with characteristics different and doses crescent of sulfur fertilization.
Each experimental plot consisted of 4 lines of 5.0 m in length and spaced 0.4 m between rows.Only the two central lines were considered as useful plots for evaluations.The sowing was performed by a mechanized, tractor-seeder set in a conventional soil preparation system.
Planting fertilization was done in the sowing furrow with 250 kg ha -1 of N-P-K (6-20-20).Two applications of nitrogen in coverage were also performed, the first one being at 25 days after emergence at the dose of 54 kg ha -1 of N, and the second 40 days after emergence at the dose of 20 kg ha -1 of N, both the coverage applications were made using NPK formulation 20-0-20.As for the management of weeds, pests and diseases, these were done according to the schedule executed in the property, respecting the needs of the crop.
Based on the useful area of the plot, the evaluated agronomic characteristics were: plant height (AP), which was measured from the soil surface to the apex of the panicle of the central stem (cm); Number of panicles per m2, and grain yield, determined by the collection and weigth of grains of each useful plot, the results being transformed to kilograms per hectare (kg ha -1 ) with grain moisture corrected to 12%.
The experimental data were submitted to individual and joint analysis of variance, with application of the F test.
The joint analysis was performed under conditions of homogeneity of the residual variances.For the comparisons between means of treatments, the regression method was used, opting for the adjustment model with the highest value of R2.Both analyzes were performed using the computational application in genetics and statistics-GENES (Cruz, 2006).

Results and Discussion
According to results presented in Table 2, for the source of variation doses, significance was verified for the variables plant height, number of panicles and productivity (p ≤ 0.01).The fact that the S doses are significant is mainly related to the rate of replacement of this nutrient to the soil, in the form of a concentrated fertilizer, in addition to the fact that it has a high content of organic matter, which reduces immobilization.
Thus, even if S is partially supplied by the atmosphere and soil organic matter, it can not be used as a concentrated fertilizer (Soares, Bardiviesso, Barbosa, & Barcelos, 2017), because symptoms of S deficiency in cultivated plants growing in the last decades due to factors such as the export of nutrients by intensive harvests, leaching, erosion, and mainly due to the absence of the macronutrient in the NPK formulations, ie, the use of only S-free concentrated fertilizers (Schimitt, Vendrusculo, Gatibonil, Santin, Wendling, Dall'Orsoletta, & Coldbella, 2018).
Concerning the effects of soil types, no significance (p ≤ 0.05 and 0.01) was found for any of the characteristics evaluated, which indicates that the soil alone does not interfere on the analyzed variables.However, there was a significant interaction (p ≤ 0.05) between the soil and the S doses applied for plant height and grain yield (Table 2).Note.** are significant (P ≤ 0.01); * Significant (P ≤ 0.05); ns Not significant by F test.
The effect of S on plant height (Figure 1) is best described by the quadratic regression model for soil 1. Increasing doses of S were observed to increase plant growth up to 40 kg ha -1 , with a maximum height of 89 cm.
In contrast, at doses above 50 kg ha -1 of S, the plants did not have a good response to the growth stimulus of the apical part, with reduction in plant height.
This fact may have occurred because soil 1 has lower levels of organic matter; so that low levels of organic matter associated with compacted soil layers or sub-surface aluminum toxicity make the sulfur sulfate not accessible to the root system of plants and may even be deficient in the most demanding crops (Chao, Harward, & Fang, 1962).
The soil flooding for rice cultivation triggers a series of electrochemical changes that interfere with the availability of nutrients, causing the reduction of sulfate to sulfide in the nutritional dynamics of S (Assis, 2000), which may cause S deficiency, due to losses by H 2 S volatilization, and consequently affecting plant growth (Carmona, Conte, & Fraga, 2009).
The volatilization losses are more accentuated in clayey soils, which when used for irrigated rice cultivation, even if superficially drained, can maintain a subsurface layer in the anaerobic state, causing volatilization and loss of S (Jordan & Ensminger, 1958).With lower nutrient contents in the deeper layers, can explain the lower growth of the plants in soil 1 compared in soil 2.
In addition, the sufficiency level of S currently considered is 10 mg dm -3 of S SO 4 2-for demanding crops, and of 5mg dm -3 of S-SO 4 2-for the other crop groups (CQFS, 2004).However, there are few reports of experiments with sulfated fertilization in the State of Tocantins, which requires studies that can establish a critical range for the soils of the region, directing the producers to increase their yields.
For soil 2, there was an increasing response at plant height with increasing S dose, with plants varying from 91 to 94 cm in height at 0 and 60 kg ha -1 , respectively.These results confirm that when the soil presents a higher content of clay and organic matter, sulfate fertilization tends to be adsorbed by the colloids of the clay, increasing the availability of the nutrient to the plants (Ensminger, 1954;Peak, Ford, & Sparks, 1999).It is expected that, heavy sulphate fertilizations on soils under these conditions will not cause a reduction in the height of rice plants grown in flooded soils.The non-reduction in plant height and the low amplitude of oscillation of the values found are interesting factors for the producers who perform the mechanized harvest, since the uniformity in the height of plants can reduce losses caused in the harvester's winch.

Table 1 .
Chemical and physical characteristics of the soils under study

Table 2 .
Summary of analysis of variance, characteristics; plant height; number of panicles and grain yield of rice in Formoso do Araguaia-TO F.V. D.O.F.