Silicon (Si) is an immobile element in the plant, being deposited on leaf blades and sheaths, stems, bark, and roots (Yoshida et al., 1962). This element participates in the biosynthesis of cell wall components that reduce enzymatic degradation. According to Dayanandam et al. (1983), silicon-accumulating plants such as rice tend to have greater resistance to the action of fungi and insects. These benefits may contribute to increased rice yield, as shown by the studies of Tokura et al. (2007), improved seed vigor (Tunes et al., 2014; Oliveira et al., 2016), and decreased use of toxic products in crop management (Vieira et al., 2011), which may assist in practices as integrated pest and disease control.

Some marketed products can be used to supply this element. Caulim® (79% SiO₂) and Sifol® (12% Si), for example, can be used for seed treatment aiming at increased plant yield. Moreover, it is known that the efficiency of agricultural technologies depends on the time of their application (SOSBAI, 2016). It is know that Si is a major constituent in higher plants and the value of Si in crop productivity also has been demonstrated. Nowadays, in many diferent countries, the application of Si fertilizers is very common in crop production systems. Silicon is already recognized as a beneficial element for several crops and may be classified as a quasi-essential element. We need to know if the application in three different moments can also bring benefits to rice plants. These stages of plant development are equally important for rice plants; however, we need to know the total amount of silicon needed and which phases require more of this nutrient. Despite research on the effects of Si on plant growth, much information is still incipient. In this context, further research is needed to address this topic. Thereby, research is needed to study Si application at different times, sources, and doses, exploring all the benefits that this nutrient can bring to rice crop. Given the above, this study evaluated silicon sources and the times for their foliar application in irrigated rice seeds.

The study was carried out under greenhouse conditions. Seeds of rice cultivars (cv) IRGA 424 (medium cycle) and Puitá Inta CL (medium cycle) were used. These cultivars were selected because they are recommended for all rice-growing regions of Rio Grande do Sul and also for showing higher yield among the cultivars. The seeds   were   sown   in   the   first  20 days  of  November  and  the experimental units were harvested on March 23. They were placed in 10 L buckets filled with sieved and homogenized soil collected from the A1 horizon of a Soil Eutrophic Haplic Planosol (EMBRAPA, 2006) belonging to the Pelotas mapping unit, in Rio Grande do Sul State (RS), Brazil. Nitrogen, phosphate, and potassium fertilization were performed according to the recommended doses for irrigated rice (CQFS – Soil fertility and chemistry commission- RS/SC, 2016). Flooding was carried out when plants had four leaves, maintaining a 10 cm water depth during the experiment. Cultural treatments were performed according to the technical recommendations for irrigated rice cultivation in RS (SOSBAI, 2016). The experimental design was bifactorial, consisting of Si application times (Stage I - full tillering, Stage II - booting, and Stage III - full bloom) and its different sources (Caulim®, at 50 Kg ha^-1, and Sifol®, at 3 L ha^-1). Four replicates were used for each treatment. Foliar applications were performed with a CO₂-pressurized back-pack sprayer.

Harvesting was carried out and panicles of each plant packed in brown paper envelopes separately and threshed manually. Harvested seeds were dried at the temperature of 42°C. The following characteristics were analyzed: number of seeds per plant, determined by counting the seeds in each plant; number of panicles per plant; and seed yield per plant (g plant^-1), where seeds were cleaned and moisture was standardized to 13% by drying in an oven with forced air circulation at 32 ± 2°C, followed by weighing on a precision scale. Physiological quality was determined by the germination and first germination count tests, using four replicates of 50 seeds. Seeds were placed between paper sheets and germinated at a temperature of 25 ± 2°C. Counts were performed at 5 and 14 days after sowing (Brasil, 2009). The filter paper method or Blotter Test was used to evaluate seed health. The sample consisted of 200 seeds subdivided into eight subsamples of 25 seeds. The seeds were placed in plastic boxes (gerbox) and incubated at 20°C for seven days (Brasil, 2009). Subsequently, the percentage of fungi-contaminated seeds was determined. Data were subjected to analysis of variance by the F test, and means were compared by the Tukey test at 5% probability, using the Winstat 1.0 software (Machado and Conceição, 2003).

Cultivar IRGA 424 showed no effect between phonological stages and foliarly applied silicon sources for seed yield and number of panicles per plant. For number of seeds per plant, however, there was interaction between treatments (Table 1). The use of Caulim^® in the tillering stage differed only from the control, with an average about 20.3% higher. In the booting stage, this source differed from the control and the use of Sifol^®. In the flowering stage, Sifol^® showed better results for number of seeds per plant, differing from the control and the treatment with Caulim^® (Table 1). For cultivar Puitá Inta CL, the application times and silicon sources tested did not affect the number of panicles. Notwithstanding, there was a simple effect of silicon sources for seed yield and number of seeds per plant in the lot with application during the tillering stage, and the use of Sifol^® was superior to Caulim^® for both variables analyzed (Table 1).

Silicon participates in the biosynthesis of cell wall components. Silicon-accumulating plants tend to have a more rigid leaf architecture and thicker cuticle, which enables  greater  use of solar energy and helps to reduce plant water loss. These characteristics later contribute to a more efficient  production  of  photo  assimilates,  which are important in seed formation (Taiz and Zeiger, 2015). Silicon  supply  in  the  different stages of rice growth and development can contribute to increased seed yield and to the formation of more vigorous and healthy seeds.

In the evaluations regarding physiological quality, there was a simple effect for silicon sources, with statistical differences for both IRGA 424 and Puitá Inta CL. For the first germination count of cultivar IRGA 424, Sifol^® differed statistically from Caulim^®, showing the highest average in the harvested lots with application during the vegetative and flowering stages (Table 2). For the first germination count of cultivar Puitá Inta CL, Sifol^® showed better results for the lot with application at booting stage. For the lot with Si application at flowering stage there were no statistical differences between Sifol^® and Caulim^® (Table 2). Regarding the germination percentage, there were no differences between treatments for both cultivars. Germination percentage was above 90% in all cases (Table 2).

Studying seeds treated with different silicon sources and doses, Oliveira et al. (2016) observed an increased vigor in the treatment with Caulim^® at 30, 60, 90, and 120 g 100 kg seeds^-1. In addition, studying the same cultivars but with aluminum silicate and ground rice husk ash as silicon sources, Tunes et al. (2014) observed an increased seed vigor through the results of root length and field emergence. Study using silicon sources for seed coating have shown that the expression of some important enzymes differs in the germination process (Tunes et al., 2014), which suggests that silicon influences this process. These results corroborate those found in the present study, where silicon treatments promoted increased physiological and sanitary quality of seeds of both cultivars.

Overall averages show that the seeds treated with silicon sources had lower incidence of fungi after harvest compared to the control (Table 3).

Studying the effect of calcium silicate and rice husk ash on fungal spots in irrigated rice seeds, Roma-Almeida et al. (2016) did not observe the reduction of this fungus by applying Si sources. However, Datnoff et al. (2007) provided   Si   to   plants   via   soil   and   leaf,  observing satisfactory results in the control of various diseases, both in mono- and dicotyledonous plants. According to Figueiredo and Rodrigues (2007), silicon is linked to the induction of a series of metabolic reactions in plants, resulting in the formation of compounds such as phytoalexins and lignins, which confers increased resistance to attack by phytopathogens and pests.

With increasing doses of silicon in rice, Berni and Pradhu (2003) observed a significant decrease in the severity of blast disease caused by the fungus Pyricularia oryzae. This fungus was not found in the present study. However, other pathogens found in both treated and untreated seed samples cause spots in grains and seeds (Phoma sp., Curvularia lunata, Nigroppora oryzae, Alternaria sp., Fusarium sp.). The high incidence of these pathogens has been of concern to producers, as they lead to sterility and decreased seed weight (Malavolta et al., 2007). The use of Sifol^® and Caulim^®, regardless of application time, proved to be important in this study, minimizing fungal levels in the seeds produced compared to the control average.

Foliar application of silicon through Caulim^® and Sifol^® proved to be beneficial, contributing to crop seed production. In general, foliar applications of silicon at tillering and boosting stages were efficient to increase the number of seeds produced in cultivar IRGA 424. The use of silicon through Sifol^® and Caulim^® improved seed quality in both cultivars, which was observed through the first germination count. Moreover, their application minimized fungal levels in harvested seeds. For cultivar IRGA 424, Caulim^® can be applied at boosting stage, and Sifol^® can be applied at tillering and flowering stages.

The authors have not declared any conflict of interests.