MYCORRHIZATION STIMULANT IN SOYBEAN ASSOCIATED WITH PHOSPHATE FERTILIZATION IN OXISOLS

The use of stimulants in the establishment of arbuscular mycorrhizal fungi has great potential in contributing to P uptake by plants. This study aimed to evaluate the effect of isoflavonoid formononetin as mycorrhization stimulant on soybean associated with phosphate fertilizer in Oxisols with intermediate (OPi) and low (OPl) phosphorus availability, in the Cerrado region of Piauí. The experiment in each soil consisted of randomized blocks design in a 4×4 factorial scheme, with four replications. The treatments comprised of four phosphorus doses (0, 26.66, 40 and 80 kg ha P2O5), and four isoflavonoid formononetin doses (0, 0.5, 0.9 and 1.8 g kg soybean seeds). The variables evaluated were soybean growth, yield, nodulation and mycorrhizal colonization rate. With the exception of mycorrhizal colonization rate in OPi, plant height and number of pods per plant in OPl, the other variables were not affected by the application of isoflavonoid formononetin, in both soils. Yield linearly increased with increase in the phosphorus doses in OPl, and presented agronomic efficiency of ~15 kg ha grain for each kg ha P2O5 applied to the soil.


INTRODUCTION
Soybean [Glycine max (L.) Merrill] is one of the most important crops in the Cerrado region of Piauí and has been indicated in the national scenario as a result of its tremendous grain yield potential (PONTES et al., 2017). Agricultural productivity in tropical regions such as the Brazilian Cerrado, has been limited by the low natural fertility of the soil (LOPES; GUILHERME, 2016). In these soils, phosphorus (P) deficiency is due to the fixation by 1:1 clay and iron and aluminum oxides (NOVAIS; SMYTH, 1999;ABDALA et al., 2015;LOPES;GUILHERME, 2016). Therefore, the implementation of technologies aiming to increase efficiency in the use of P by plants is important for the cultivation of this agro-ecosystem (TEIXEIRA et al., 2016).
In plants, P plays an important role in photosynthesis, respiration, energy storage and transfer, and consequently acts in cell division and growth, among other processes of plant development (HAWKESFORD et al., 2012). Thus, P deficiency causes reduction in shoots by limitation of the number and expansion of the leaves, shoots branching, reduction in carbon assimilation rate, and acceleration of leaf senescence (HAWKESFORD et al., 2012;SINGH;REDDY, 2015), which explains the low production of biomass of plants cultivated in the absence of P 2 O 5 .
The management of mycorrhizal symbiosis in the improvement of P nutrition is an alternative to be taken into account, due to the stimuli relevant to plant growth (NOURI et al., 2014), which are attributed to arbuscular mycorrhizal fungi (AMF). Thus, due to the exploitation of a greater soil amount and capacity of hyphae to absorb P at low concentrations (SMITH et al., 2011), AMF favors plant growth in low availability conditions of this element, contributing to 80% of P absorption (MARSCHNER;DELL, 1994).
The technological development for commercial use of AMF has been quite limited, since they are obligate biotrophic (BERRUTI et al., 2016). However, the use of stimulants to establish native AMF in roots has great potential in extensive agriculture. Ribeiro et al. (2016) evaluated field application of isoflavonoid in soils with different P doses, and concluded that formononetin isoflavonoid provided greater symbiosis between plants and mycorrhizal fungi, increasing colonization and maximizing the benefits they provide to plants. It is noteworthy that the use of formononetin has promoted grain yield in several crops, such as beans (LAMBAIS; RÍOS-RUIZ;ANDRADE, 2003), potato (DAVIES;CALDERON;HUAMAN, 2005) and soybean (CORDEIRO et al., 2015) in greenhouse condition. Thus, similar responses are expected in field conditions after the application of AMF stimulants.
It should be noted that mycorrhization management cannot be a substitute for phosphate fertilizer, since crops require high P doses. However, it may contribute to improvement of the efficiency of the fertilizer.
The objective of this study was to evaluate the effect of isoflavonoid formononetin as mycorrhization stimulant on soybean associated with different P 2 O 5 doses in two Oxisols at the Cerrado region of Piauí.
According to Köppen (1931), the local climate is classified as Aw tropical, with two well defined seasons: dry season, from May to September, and rainy season, from October to April. Figure 1 shows the rainfall and temperature during the experiments.
The experiment in each soil is comprised of randomized blocks design with four replications and 16 treatments, in a factorial 4 × 4. The factors consisted of four P doses (0, 26.66, 40 and 80 kg ha -1 P 2 O 5 ), and four doses of the isoflavonoid formononetin (0, 0.45, 0.9 and 1.8 g kg -1 seeds), formulated like the commercial product Myconate ® (7-hydroxy, 4'-methoxy-isoflavone). For P doses, the values correspond to the application of 0, 25, 50 and 100% of the recommended dose for intermediate initial P in soil, whereas for doses of isoflavonoid formononetin, the values correspond to the application of 0, 25, 50 and 100% of the recommended doses of 77.28 g ha -1 according to the manufacturers' recommendations (Plant Health Care). Super simple phosphate fertilizer (18% P 2 O 5 ) was used. Each plot was 6 m long and 3 m wide, with planting rows spaced 0.50 m apart, and the area of each plot was 18 m 2 (6 x 3 m). The three central rows were considered as useful plot, excluding 1 m of each extremity. The soybean cultivar used was M-soy 8766 RR, of late cycle and determinate growth, recommended for a Cerrado opening area with intermediate and low fertility soil. Seeds were treated with insecticide fipronil + fungicides pyraclostrobin and tiofanato metílico (Standak Top ® ), and at the time of planting, they were inoculated with Bradyrhizobium japonicum, SEMIA 5079 and 5080 strains (5.0 x 10 9 viable cells mL -1 ) at the dose of 2 mL kg -1 seed, and subsequently, isoflavonoid doses were applied. Before the soybean seeding in both years, glyphosate herbicide was applied at a rate of 1,080 g ha -1 a.e. to eliminate weeds. Weeds control at post-emergence was carried out with soybean at the trifoliate stage (V3), using glyphosate at a rate of 1,080 g ha -1 a.e.
Potassium fertilization in OPi consisted of 80 kg ha -1 K 2 O applied in the planting row. In OPl, 130 kg ha -1 K 2 O was used, where 80 kg was applied in the planting row and the remaining was applied in topdressing when the crop was between phenological stages V4 and V5. Pest control was carried out at phenological stages V3, V6 and R2, using flubenzurom insecticide (Nomolt ® ) at a rate of 22.5 g ha -1 a.e., for stinkbugs control, and permethrin was added to the last application (Pounce 384 CE ® ) at a rate of 49.9 gha -1 a.e. Disease control was performed with the combined application of carbendazim and azoxystrobin + cyproconazole (Carbendazim ® and Priori Extra ® ), in a mixture, at the phenological stage R1, and at 21 and 35 days after the first application, at a rate of 400 and 84 gha -1 a.e., respectively.
The soybean plant stand in the useful area of the plot was evaluated 14 and 28 days after emergence. When the crop was at the phenological stage R1, plant height, shoot dry biomass, relative chlorophyll content, leaf P content, number of nodules per plant, nodules fresh weight, nodules dry weight and mycorrhizal colonization rate were measured. The relative chlorophyll content was determined using a portable Clorofilog CFL 1030 Falker device. For sampling, the third fully developed trifoliate was used, from the apex to the base (diagnostic leaf), by sampling three trifoliates per plot.
Leaf P content was determined by collecting the diagnostic leaf in 10 plants of each plot. This material was dried in a forced air circulation oven at 60°C for 72 h and ground in a Willey mill equipped with a 40 mesh sieve. To obtain the leaf P content, the tissues were digested using the nitropercloric digestion method and the determination was done by the molybdate blue method (MURPHY; RILEY, 1962). Shoot dry biomass, plant height, number of nodules per plant, fresh and dry weight of nodules, as well as mycorrhizal colonization rate were determined using five plants per plot. Plant height was determined by direct measurement with the aid of a graduated metric tape in centimeters and nodules fresh weight after count was weighed in a precision scale balance. The dry biomass of plants shoots and nodules dry weight were obtained by drying the plants and nodules in a forced air circulation oven at 60°C for 72 h and followed by weighing in a precision scale balance.
To determine the mycorrhizal colonization rate, roots of plants were collected and washed with water, and stored in plastic flasks containing a solution of 5% formaldehyde, 90% ethanol and 5% acetic acid (PHILLIPS; HAYMAN, 1970). Thereafter, one gram of fine roots was clarified using 5% KOH solution for 30 min. Then, the material was washed in tap water and shaken for four minutes in 1% HCl. Roots with fungal structures were subsequently stained with 0.05% (w/v) trypan blue in lactoglycerol (1:1:1 lactic acid, glycerol and water) following procedures described by Phillips and Hayman (1970). To estimate the percentage of colonized roots, the grid-line intersect method described by Giovanetti and Mosse (1980) was used, followed by observation under a stereoscopic microscope.
At the end of the crop cycle, five plants were collected for evaluation of the number of pods and grains per plant and weight of a thousand grains. Afterwards, plants of the useful area of the plot were manually collected, mechanically threshed, standardized to 13% humidity and the grains yield (kg ha -1 ) was obtained. The grain harvest index was determined as follows: GHI = grain yield/grain yield + straw The measured data were subjected to a variance analysis and in the case of significance (p =0.05), factors were subjected to polynomial regression analysis for the effect of individual factors (P or isoflavonoid formononetin doses) and multiple linear regression when there was interaction of these factors, by using the statistical software R 3.2.3, and ExpDes package (FERREIRA; CAVALCANTI; NOGUEIRA, 2013). In order to obtain homoscedasticity, data of shoot dry biomass weight, number of nodules, and nodules fresh and dry weight were transformed by the equation (x + 1) 0.5 . The values presented in tables and figures are the original means.

RESULTS AND DISCUSSION
Mycorrhizal colonization rate (Figure 2A) was influenced by the interaction between P 2 O 5 doses and formononetin in the OPi, demonstrating the positive influence of formononetin application, and the negative influence of high P 2 O 5 doses, thereby corroborating the results of Cordeiro et al. (2015) in soybean cultivated in Oxisol with intermediate P content. Therefore, it should be noted that the use of high P doses reduces the stimulation of mycorrhizal colonization (BALZERGUE et al., 2013), since during the symbiotic association, the plant provides assimilates to the fungi (BAGO; PFEFFER; SHACHAR-HILL, 2000). However, in a condition of full availability of P in the soil, this relation is no longer advantageous for plant, since P uptake occurs without extra energy expenditure. Although, a mycorrhizal colonization was expected in soil with low P content, this did not occur. This may be related to the environmental factors and not only the P content since the colonization by AMF depends on both biotic and abiotic factors to stimulate the germination of spores and association with the roots of plants (SOUZA, 2015).
For plant height ( Figure 2B), there was interaction between P 2 O 5 doses and formononetin in OPl, such that the significant and positive influence of P 2 O 5 doses on the promotion of plant growth is evident, proving that the P 2 O 5 factor is preponderant in the growth of soybean plants, when compared with the formononetin factor. Thus, in general, higher values for plant height (~70 cm) were obtained using 80 kg ha -1 P 2 O 5 combined with the doses of 0.45 or 1.8 g kg seed -1 formononetin. It is noteworthy that the absence of both factors significantly reduced plant height by ~20%, when compared with the maximum doses of P 2 O 5 and the doses of 0.45 or 1.8 g kg seed -1 formononetin. Cordeiro et al. (2015) reported that in Oxisol with intermediate P doses, there was no interaction between the application of P 2 O 5 and formononetin for plant height, differing from the results obtained in this study. However, the same author observed individual effects of formononetin, and greater height was obtained with the application of 0.5 mg seed -1 formononetin in a greenhouse experiment. This occurred because P is an important constituent of energy compounds (ATP/NADPH), phospholipids, and other esters which play important functions in plant (HAWKESFORD et al., 2012), by influencing proper plant development. The number of nodules ( Figure 3A) and shoot dry biomass ( Figure 3C) had a quadratic curve adjustment in relation to the P 2 O 5 doses, reaching a greater number of nodules and shoot dry biomass with estimated doses of 44.2 and 70 kg ha -1 P 2 O 5 , respectively. However, there was a reduction in the number of nodules per plant at a dose of 80 kg ha -1 P 2 O 5 . This was similar to those obtained by Abbasi et al. (2008), who observed in a greenhouse that, the use of a mixture of two strains of Bradyrhizobium japonicum (S377 + S379) and 100 kg ha -1 P 2 O 5 reduced the number of nodules, and a dose of 50 kg ha -1 P 2 O 5 increased the number of nodules, which is similar to the values obtained in the present study. Thus, it should be noted that P is essential for the proper development of nodules and for nitrogen fixation, especially in its formation, resulting in a greater number of nodules (KARIKARI; ARKORFUL; ADDY, 2015). The determination of P 2 O 5 doses, with the aim of reaching the maximum biomass production has proven to be uncertain, ranging from 30 to 100 kg ha -1 P 2 O 5 (MABAPA et al., 2010;RIBEIRO et al., 2016).

Rev
In general, both AMF and the nitrogen fixing bacteria have a synergistic relationship, although the responsible mechanisms are not yet elucidated. Silva et al. (2017) demonstrated more active soybean nodules when colonized by AMF under field conditions. Studies have shown the effect of formononetin on soybean mycorrhization (CORDEIRO et al., 2015;RIBEIRO et al., 2016). However, in relation to nodulation, the results of these studies are contradictory regarding isoflavonoid application. In the study of Ribeiro et al. (2016), which was carried out in the field, there was no influence of formononetin on the number and weight of nodules of soybean plants; on the other hand, in the study of Cordeiro et al. (2015), which was also carried out in the field, it was found that formononetin application increased the number of nodules.
The increase in P 2 O 5 doses promoted linear increase in relative chlorophyll content and the number of pods in plants ( Figure 3B and D). Tairo and Ndakidemi (2013) obtained results contrary to those of this study, since no effect on the chlorophyll content with P doses ranging from 0 to 80 kg ha -1 , were found. According to Hawkesford et al. (2012), P may indirectly contribute to the relative chlorophyll contents of the leaves since it acts in the photosynthesis process, and the nutrient plays important roles in the formation of NADPH and ATP. Fageria, Moreira and Castro (2011) did not find linear growth for the number of pods when increasing doses of P 2 O 5 (0 to 120 kg ha -1 ) were used in Oxisol with low P content, which is different from the results of this study.
The number of pods increased linearly in response to the application of formononetin ( Figure 3E). These results corroborate those obtained by Cordeiro et al. (2015), in which the positive effect of formononetin application on the number of pods in soybean plants was found. Although the effect of formononetin application is associated with root AMF colonization, its exogenous application can reduce the activities of peroxidase (FRIES; PACOVSKY; SAFIR, 1998) and catalase (LAMBAIS; RÍOS-RUIZ;ANDRADE, 2003), which are enzymes associated with the elimination of reactive oxygen species and contribute to increased number of pods in the soybean plant.
The number of grains per plant ( Figure 4A) was adjusted to the quadratic model of the curve, reaching a maximum limit of 222 grains per plant, with the application of an estimated dose of ~66 kg ha -1 P 2 O 5 . The weight of one thousand grains, grain yield and grain harvest index ( Figure  The stabilization in number of grains per plant ( Figure 4A), occurred with 66 kg ha -1 P 2 O 5 . However, there was still an increase in the weight ( Figure 4B), suggesting that soybean plant prioritizes the transport of assimilates to the existing grains in detriment to new grains. Similarly, the increase in grain yield and grain harvest index in function of the P 2 O 5 doses is consistent with the linear increases observed in the number of pods per plant ( Figure 3D) and weight of one thousand grains ( Figure 4B). According to Grant et al. (2001), under P stress, the total number of seeds produced by the plant is reduced as compared to the seed size. Thus, both the abortion of flowers and legumes occur to maintain plant production equilibrium within the available assimilates (BOARD; HARVILLE, 1994).
Soybean grain yield was ~1.200 kg ha -1 without phosphorus fertilization. However, with the addition of P to the soil, the agronomic efficiency was ~15 kg ha -1 grains for each kg ha -1 P 2 O 5 applied, which was similar to the value obtained by Valadão Júnior et al. (2008), who worked on increasing P 2 O 5 doses in the state of Rondônia. In the Cerrado region of Piauí, with low initial P contents in the soil, Alcântara Neto et al. (2010) observed quadratic response to phosphate doses application, and the maximum efficiency was obtained at a dose of 94.8 kg ha -1 P 2 O 5 , for yield of 2.614 kg ha -1 grains, which was is similar to the value obtained in this study (2,395 kg ha -1 grain).  The adequate P supply from the beginning of plant growth stimulated root development, formation of the reproductive structures, good formation of legumes and seeds, as well as yield (SHARMA et al., 2013). Therefore, it is possible that during the growth and development of plants, the P 2 O 5 dose responsible for the maximum yield, together with other nutrients added to the soil, provided the nutritional needs of the crop in a balanced way.

CONCLUSIONS
The initial phosphorus content influenced soybean grain yield in Oxisol, and the effect of phosphate fertilizer occurred only in soil with low phosphorus content.
In Oxisol with low P 2 O 5 doses, positive response of phosphate fertilizer was observed for the number of nodules, shoot dry biomass, plant height, relative chlorophyll content, number of pods, number of grains, weight of one thousand grains, grain harvest index and grain yield with the application of P 2 O 5 doses ranging between 40 and 80 kg ha -1 .
Formononetin application was efficient in the development of soybean in Oxisol with intermediate initial P 2 O 5 doses, with regards to mycorrhizal colonization rate, and in Oxisol, with low initial P 2 O 5 doses, there was development of soybean with regards to plant height and number of pods per plant, respectively, at formonetin doses ranging between 0.90 and 1.80 g kg -1 seeds.