Something aggravating is seen in the Southern Cone of Rondônia (SCRO), where spatial irregularities in rainfall distribution caused by changes in land use and cover (Khanna et al., 2017) have been altering the agricultural scenario of the main annual crops in this region (Andrea et al., 2018), including rice.

The municipality of Vilhena, an exponent of rice production in the SCRO, is located in a transition region between the Savanna and the Brazilian Amazon. The area is characterized by a clayey texture, with 64% clay, 16% silt, and 20% sand, although the soil has a low cation exchange capacity (CEC). Tropical regions, predominantly in Brazil, have more weathered soils, with a predominance of low activity clays and medium levels of organic matter. In this region the Latosols are highly weathered, with limitations in food production due to their low natural fertility (Lopes, 1996; Hunke et al., 2015), as they are acid soils with low availability of nitrogen, phosphorus, potassium, calcium, magnesium, zinc, boron and copper, with high aluminum saturation and high P fixation.

The farmer must pay attention to the characteristics of the sown cultivar, the amount of inputs, and the management techniques used in order for the rice crop to perform well and yield a high economic return (Xavier et al., 2021). Furthermore, it is still necessary to analyze its phenological and/or genotypic characteristics, as well as the soil edaphoclimate where it will be sown (Quevedo-Amaya et al., 2020; Sandhu et al., 2019).  

There has been a decrease in the number of rice farmers in Brazil in recent years. However, it began to exhibit a high bias in its price in the second half of 2019 and with the entry into the core of the off-season, as there is a significant deficit between the country's supply and demand for grain (IPEA, 2020). This can be explained by the fact that a portion of farmers invest in soybean production rather than rice production due to the lower cost (Conceição et al., 2017; Júnior, 2019). In addition, weed infestations resistant to current methods of control (Rubin et al., 2014) and blast disease caused by the fungus Magnaporthe oryzae (Barr) Couch [anamorphic Pyricularia oryzae (Cav.)] (Prabhu et al., 2009) contribute to the difficulty of producers to remain in business.

Camargo et al. (2008), Dario et al. (2012), Marchezan et al. (2001), and Wei et al. (2012) investigated the possibility of  increasing  tillage  productivity  by  including micronutrients in rice, but found no productivity increases. On the other hand, Mahmoodi et al. (2020), Nadeem and Farooq (2019), Phattarakul et al. (2012), and Prom-u-thai et al. (2020) obtained yield increments with the application of Zn. These different results are associated with the soil type and rice cultivation system in lowlands (flooded areas) and uplands (rainfed areas). According to Lahijani et al. (2020), deficiency of micronutrients, especially Fe and Mn, is a determining factor in reducing the productivity and quality of agricultural crops. Foliar sprays, according to these authors, are viable alternatives to rice culture for increasing yield rates due to the role of micronutrients in plant nutrition and enzyme activation.

In acidic soils, Zn deficiency is common in upland rice (Fageria and Nascente, 2014). Plants lacking Zn during the early stages of development may have their development impaired and will hardly be able to reach their maximum genetic yield potential. It affects the maintenance of some enzymatic activities and the tryptophan synthetase enzyme, causing a decrease in cell volume and lower apical growth (Epstein and Bloom, 2006).

Molybdenum is not readily available in soils with low pH, and it may be one of the limiting factors to the biological fixation of atmospheric nitrogen concerning legumes and affecting the nitrate reductase enzyme activity in deficient leaves (Sundim et al., 2002; Guimarães et al., 2007). This can be proven in the works of Das Gupta and Basuchaudhuri (1977) and Fageria and Baligar (1997), who identified that foliar application of molybdenum with the addition of mineral nitrogen has a positive effect on the accumulation of dry mass in rice cultivar.  Manganese is an element of vital importance for the development and growth of plants. Furthermore, it is involved with enzymes activated by cations and in the photosynthetic evolution of oxygen (Taiz and Zaiger, 2004).

Given the scarcity of research that verifies the effects of foliar fertilization based on the micronutrients Zn, Mo and Mn in rice cultivation in highlands, the present study aimed to evaluate the performance of some yield components and the productivity of four rice cultivars in Vilhena-RO, through the application of foliar fertilizer from three commercial products in the vegetative stage V5. 

The experiment was carried out in the experimental area of ??Faculdade Marechal Rondon – FARON (12°46'02”S 60°05'49”W and 588 m altitude), located in the municipality of Vilhena-RO, on the banks of the BR-364 road, in the CSRO (Figure 1), from November 2020 to March 2021. The soil is the Latossolo Vermelho-Amarelo distrófico type of clayey textural class (Godinho et al., 2009). According to Alvares et al. (2013), the climate of the region is tropical rainy (Am) with a well-defined dry season. The climatic data referring to the experiment period were obtained from the micrometeorological  station  located  at  FARON,  adjacent   to  the experimental area where the present study was carried out.

The previous occupation was degraded pasture, and the soil was prepared with two passes of a leveling harrow before the installation of the experimental plots. Because it was the first agricultural year, the soil was chemically analyzed before rice sowing by collecting deformed samples (soil samples taken by modifying their natural structure) with augers at two depths (0-10 and 10-20 cm) (Table 1). Magnesian limestone (PRNT 86%) was applied 3.2 t ha^-1 to raise the pH of the area, as well as to raise the soil base saturation to 60%. Limestone was applied manually in a homogeneous manner throughout the experimental area, and incorporation was done by harrowing, 30 days before rice sowing. The experimental design was randomized blocks, in a 4 × 2 factorial scheme, with factor 1 consisting of 4 certified cultivars (ANa 9005 CL, ANa Cambará, ANa 8001 and BRS A502) recommended and widely sown in the region, and factor 2 consisting of foliar fertilizer application and no foliar fertilizer application. Each experimental plot included five lines of 5 m in length, spaced at 0.45 m, and a seed density of 100 kg ha^-1 (67 seeds per linear meter). The three central lines were used as a useful area to evaluate the characteristics of the cultivars, disregarding 0.5 m from each  extremity  and  the  two  lateral  lines, obtaining 5.4 m² of useful area.

Fertilization was performed in the furrow, using 240 kg ha^-1 of the fertilizer 04-24-12, following soil analysis and the recommendation protocol for rice cultivation under dryland conditions and clay soil of Rical - Rack Ind. e Com. de Arroz Ltda. In all cultivars, cover crop fertilization was performed with the formulation 20-00-20, 160 kg ha^-1, divided into two stages: the first done at the effective tillering, about 15 days after emergence (DAE), and the second applied at maximum tillering of the crop (vegetative stages V5 to V6), approximately 32 DAE.

Foliar fertilization occurred at 20 DAE (V4 stage), with a recommended application of 500 ml per hectare, obtaining 7.2 ml of the following products: Maxi Zinc^® (zinc 50% w/w), Booster^® (molybdenum + Zinc 5% w/w) and Broadracre Mn^+® (manganese 27% w/w). All foliar sprays performed, both for phytosanitary controls and foliar fertilization, were carried out with the aid of a 20 L manual backpack sprayer with a conical nozzle. The local climate and soil moisture conditions were ideal when applying the micronutrient foliar fertilizer (Table 2).

Cultural treatments were carried out whenever necessary, with weed control occurring through manual mowing without herbicide application  for  control.  An   application   of   52.8 g ha^-1 a.i.  of  the insecticide based on Dinotefuran and Lambda-Cyhalothrin was carried out to control the attack of froghoppers (Deois flavopicta) at 47 DAE. In addition, 84 g p.a.i. ha^-1 of fungicide based on Picoxystrobin and Cyproconazole for the preventive control of brown spot (Bipolaris oryzae) was applied at 58 DAE. In its from V5 to V6 stage, a sulfur deficiency was diagnosed, requiring the application of ammonium sulfate in the coverage at a dose of 80 kg ha^-1.

Manual harvesting was carried out using a cleaver at a cutting height of 15 to 20 cm above the ground when 90% of the panicles had the typical coloration of mature grains. The drying was then done in the shade, followed by the mechanized threshing. The harvest of the cultivar BRS A502 was carried out at 100 DAE; cultivars ANa 8001 and AN Cambará at 110 DAE and ANa 9005 CL occurred at 117 DAE.

The following characteristics were evaluated: number of panicles per m², determined by counting panicles in two linear meters in each plot; number of spikelets per panicle, obtained utilizing an average count of 15 panicles per plot; 1000 grains weight, performed by weighing three samples from each plot, each with 100 grains, and then multiplying by 10 (weight of 100 grains × 10); and productivity in t ha^-1, obtained by weighing the grains with husk from the useful area of the plots, correcting the humidity to 13% and converting it into t ha^-1.

The evaluation of the whole grain yield was performed in the RICAL classification laboratory. For such, a sample of 100 g of paddy rice grains was collected and a testing mill (Zaccaria), model PAZ-1-DTA, was used for 1 min and 15 s; then, the burnished (polished) grains were weighed and the values found were considered as yield of benefit, with data in percentage. Afterwards, the burnished (polished) grains were placed in the "trieur" at 5 and the separation of the grains was processed, for 60 s; the grains that remained in the "trieur" were weighed and the value obtained was considered as yield of whole grains and the rest, broken grains, both expressed in percentage. The experimental data were submitted to individual and joint analysis of variance, applying the F-test. The joint analysis was performed under conditions of homogeneity of residual variances. The Tukey test was applied at a 5% probability for comparisons between treatment means. All analyses were performed using the AgroEstat statistical software (Barbosa and Maldonado Junior, 2015).

The results show that there was no statistical difference between the levels (with and without) of the foliar fertilization factor in the number of panicles per m² (NP), thousand-grain mass (TGM), percentage of whole grains (WG) and productivity (PROD). However, it was significant for the number of spikelets per panicle (NSP) (Table 3).

Concerning the variable NSP, it was observed that the application of foliar fertilizer promoted an increase of 8% for cultivar ANa 8001. These results are corroborated by Lahijani et al. (2020), who evaluated the effect of applying foliar fertilizers based on Fe, B, Zn, Cu, and Mn and also observed increases in NSP ranging from 92 (17%) to 159 (72%) in a study developed in Iran in an area with irrigated rice cultivation. According to Buzetti et al. (2006), the total number of spikelets is influenced by genetic factors and external conditions during the vegetative phase until five days before flowering, which is essential for achieving good crop yields. 

When analyzing Table 3, it was discovered that the NP parameter was not influenced by the application of foliar fertilizers, resulting in no significant yields. However, only BRS A502 showed the highest NP when comparing this parameter between cultivars, significantly differing from the other cultivars. This can be explained by the genetic characteristics of this material since it has a high value for this parameter, as well as in relation to PROD for edaphoclimatic conditions in the Cerrado. According to Zayed et al. (2011), NP can be influenced by the form of application of micronutrients. The authors conducted a study in Egypt and discovered a 15% increase in NP compared to the control treatment (no application) when foliar fertilizer (a combination of Zn, Fe, and Mn) was applied via soil. In comparison, the increment of this parameter was only 8% via foliage. Additionally, Lahijani et al. (2020) highlight that foliar spraying of Fe, Zn, B, Cu, and Mn elements can delay plant metabolism procedures compared to soil application, which can compromise rice yields.

BRS A502 was one of the cultivars with the lowest NSP; however, it was the one with the highest NP 335.7 (32%), which was compensated in the TGM and PROD of this material. In addition, this result can be explained by the plants showing less sterility of spikelets and a greater amount of whole grains, allowing excellent productivity, as observed by Fageria et al. (1995).

It was noted that three of the four cultivars evaluated had a TGM of approximately 25.02 g, 9% higher than the cultivar ANa 8001, which average reached was 23.1 g. According to Boldieri et al. (2010), this slight variation of TGM is directly linked to the genetic characteristics of the evaluated cultivars. This observation is very important because, according to Fornasieri Filho and Fornasieri (2006), TGM is a stable varietal trait and is basically dependent on the husk size.

Concerning WG, there is a statistical difference between cultivars BRS A502, ANa 8001, and AN Cambará in relation to ANa 9005CL. It is approximately 10% lower for the last cultivar, implying a percentage of integers below that established by the company Agro Norte Pesquisa e Sementes Ltda. It is responsible for producing this cultivar, which establishes an acceptable range for this material between 61 and 70%. This may be associated with the time the grains remained in the field after complete physiological maturation, exposed to excess rain, solar radiation, and high temperatures, which caused changes in the moisture content of the grains and a low rate of whole grains (54.83%), as observed by Juliano and Duff (1991) and Smiderle and Pereira (2008).

A highlight regarding the yields of the cultivars studied is that all of them showed higher yields than those obtained by Agro Norte and EMBRAPA in their field experiments, which were carried out in the region covered by the FARON area. All productions are also higher  than  the  state  of  Rondônia's  average  (4 t ha^-1) (CONAB, 2020). This consolidates the good adaptability of these cultivars to the soil and climate conditions of Vilhena. Even the soil presenting a pH value of 4.9, classified as high acidity (Table 1), did not interfere in PROD, as reported by Moro et al. (2013). The PROD seen in the ANa 9005CL cultivar should be observed with reservations, as there were attacks in three plots by wild animals natives from the forest located in the area adjacent to the experiment site. In addition, it should be noted that the plots affected by the attacks were located on the edge of the experiment, which facilitated the degradation of these experimental plots.

Even performing foliar fertilization at the stage V4, the cultivars presented results above the expected compared to the results of Agro Norte and EMBRAPA in their field experiments mentioned above. According to Garcia and Hanway (1976) and Rosolém and Boaretto (1989), the time when annual crops need to increase or maintain nutrient concentrations in the leaves is close to flowering and at the start of flowering for grain filling to prevail because root absorption of nutrients is practically zero at this stage. Therefore, if the application of foliar fertilizers were carried out at their phenological stage close to flowering and at the beginning of flowering, the cultivars used could have higher yields than those observed.

It is noteworthy that, for the experimental field conditions considered in the present work, the foliar fertilization with micronutrients performed in the vegetative stage (V4) in the production of upland rice did not influence the evaluated production components, except for NSP, which is in agreement with the results found by Andrade et al. (1997, 1998), Dynia and Moraes (1998), Marchezan et al. (2001), Camargo et al. (2008), Fang et al. (2008), Wei et al. (2012), Dario et al. (2012), and Maldonado Junior et al. (2013). Some differences in the results found in the literature can be attributed to the different locations where the field tests were conducted, particularly when considering the physical-chemical and water characteristics of the soil, as well as the local climate conditions, which vary from year to year in the various regions where rice sowing occurs. Furthermore, the form of application of the foliar fertilizer directly influences the physiological and phenological responses of rice subjected to either soil or foliar application (Marchezan et al., 2001). Nutritional deficiency is usually found in soils with a more sandy texture, which was not the case in the present work, once the soil was the Latossolo Vermelho-Amarelo distrófico type of clayey textural class.

The application of foliar fertilizers Maxi Zinc^®, Booster^®, and Broadracre^® did not significantly affect the yield components, except for NSP and productivity of the evaluated rice cultivars. Under the region's edaphoclimatic conditions, the cultivar ANa 8001 produced the best results in terms of productivity and grain physical quality.

The authors have not declared any conflict of interests.

The authors thank Faculdade Marechal Rondon - FARON, the research unit of the Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA) in Vilhena and Rical Rack Indústria e Comércio de Arroz Ltda. for providing all the support necessary to carry out this research work.