Efficiency of pre-inoculation of soybeans with Bradyrhizobium up to 60 days before sowing

The inoculation of nitrogen-fixing bacteria in soybean crops allows the achievement of high crop yield by reducing or eliminating the application of nitrogen fertilizers. Pre-inoculation of seeds can reduce costs and increase the efficacy of this agricultural practice. The objective of this study was to assess the efficacy of pre-inoculation of soybean seeds with a commercial inoculant of Bradyrhizobium (RIZOLIQ LLI ® manufactured by Rizobacter do Brasil) at 60, 45, and 30 days before sowing using a cell protector and different chemical treatments. The study was conducted in four municipalities in the state of Paraná (Londrina, Pato Branco, Ponta Grossa, and Santa Tereza do Oeste) and included 14 treatments, a negative control, standard inoculation of the recommended bacterial strains on the day of sowing, and pre-inoculation at 30, 45, and 60 days before sowing using the commercial inoculant RIZOLIQ LLI ® together with the cell protector and chemical treatment of seeds with Imidacloprid (Rocks ® ), Fipronil+Thiophanate-methyl+Pyraclostrobin+ (Standack ® Top), Metalaxyl-M+Fludioxonil (Maxim ® XL), and Metalaxyl-M+Fludioxonil+Thiabendazole with Thiamethoxam (Maxim ® Advanced+Cruiser ® ). The nodulation, plant biomass, nitrogen concentration in shoot and grain, and grain yield were evaluated. The cell protector was efficient in maintaining the bacterial inoculant viable in the seed for up to 60 days. All treatments of pre-inoculation of soybean chemically-treated seeds up to 60 days before sowing could be performed without impairment of nodulation, plant biomass, nitrogen concentration in shoot and grain, and grain yield. Therefore, pre-inoculation of soybean seeds up to 60 days before sowing is an efficient and practical inoculation strategy for sowing soybean crops.


INTRODUCTION
Glycine max (L.) Merr.(soybean) is actually the most cultivated oilseed plant worldwide.Soybean production in Brazil was the second largest worldwide and reached 113.9 million tons of grains in the 2016/2017 harvest (CONAB, 2017).
Nitrogen (N 2 ) is the most required nutrient in soybean crops and biological nitrogen fixation (BNF) under the growth conditions of Brazil can provide all the nitrogen needed by the crop (Mourtzinis et al., 2018), making this process indispensable for the economic and environmental viability of soybean cultivation in Brazil.This biological process is performed by bacteria of the genus Bradyrhizobium, which can fix the atmospheric N 2 naturally and in synchrony with crop requirements and cycle, supplying more nitrogen to grains than chemical fertilizers (Kaschuk et al., 2010).
The efficiency of BNF in leguminous plants, especially in tropical regions, has been affected by several edaphoclimatic factors, management practices and seed treatment with fungicides, and these factors could compromise the viability of bacterial cells and symbiosis (Hungria et al., 2007).For this reason, there is a growing interest in research on innovative strategies that can help maximize BNF and increase crop productivity.
The practice of the inoculation of Bradyrhizobium species has been recognized for its advantages and economic and environmental gains.However, some factors limit the application of this technology, including inoculation on the day of sowing, which generates extra work for the producer and increase labor and time to prepare the seeds (Aguiar et al., 2014).These difficulties have stimulated the non-utilization of this inoculation practice in the soybean crop by producers (Zilli et al., 2010).
Pre-inoculation or early inoculation of Bradyrhizobium strains recommended for soybeans might be the solution as long as the bacterial cells are viable on the day of sowing and the chemical treatment of the seeds does not reduce the viability of the inoculant.Therefore, preinoculation allows amplifying the effects of BNF in soybeans by the increased inoculation of soybeans with Bradyrhizobium by farmers because the seeds could be commercialized in the pre-inoculated homogenized form and seed quality is guaranteed (Anghinoni et al., 2017;Araújo et al., 2017).
The objective of this study was to assess the agronomic efficiency of soybean pre-inoculation with the inoculant Rizoliq LLI + Premax cell protector at 30, 45, and 60 days before sowing seeds chemically treated with different products.
The study included four replicates arranged in a completely randomized block design and was conducted in plots of 4.05 × 6.00 m (24.3 m 2 ) with nine crop rows each with 6 m in length, spaced at 0.45 m, with approximately 12 plants per linear meter.The soybean varieties cultivated in the state of Paraná were BMX Potência RR in the municipalities of Londrina and Pato Branco and BMX Apolo RR in Santa Tereza do Oeste and Ponta Grossa.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License treatment without addition of inoculant or mineral nitrogen were used (Table 2).The inoculated and treated seeds were stored in paper bags and kept at ambient temperature (25 ± 3°C) protected from the sun and humidity until the time of sowing (MAPA, 2010).

Bacterial survival in inoculated seeds
After storage of pre-inoculated seeds, viable cells of Bradyrhizobium spp.were recovered and quantified on the surface of the seeds according to Brazilian regulations (MAPA, 2010).Standard inoculation with RIZOLIQ ® was also subjected to analysis.For this procedure, samples of 100 seeds of each treatment were aseptically transferred to an Erlenmeyer containing 100 mL of a solution of 0.85% NaCl (w/v) and 0.01% Tween 80 (w/v) and shaken in an orbital shaker for 15 min at 150 rpm.The obtained suspension was serially diluted ten-fold in a 0.85% NaCl solution (w/v).After that, 0.1-mL aliquots of the dilutions 10 -1 to 10 -9 were transferred to Petri dishes containing Ikuta semi-selective culture medium (IKUTA, 1995), Congo red (0.25 g/100 mL), and the antimicrobials nalidixic acid (20 mg L -1 ), neomycin (20 mg L -1 ), chloramphenicol (20 mg L -1 ), actidione (10 mg L -1 ), and triazole (2.5%).The aliquots were spread over the culture medium using a drigalski loop with three replicates on distinct plates.The plates were kept in an oven at 28 ± 2°C for 10 days and then the colonies were counted.Only the dilutions whose average of the three plates were 30 to 300 CFU were considered in the count.The number of B. japonicum recovered from the seeds was transformed into log10 (MAPA, 2010).

Experimental procedures and analyses
The execution of the experiments followed the agroclimatic zoning of each region (MAPA, 2015) and sowing were performed in October.The cultivation practices adopted for managing weeds, pests, and diseases complied with the recommendations for soybean cultivation.During flowering (phenological stage R1), five plants of each treatment were collected and the following parameters were evaluated: number of nodules per plant, dry mass of nodules per plant, dry mass of root and shoot, and concentration of nitrogen (N) of the shoot following the methodology described by Miyazawa et al. (1992).Grain yield and nitrogen concentration in grains (Miyazawa et al., 1992) were evaluated after 50% of the crop reached the R8 phenological stage.

Data analysis
The results were subjected to analysis of variance (ANOVA) with a level of significance of p<0.1.The mean number of viable bacterial cells on the seeds were compared using Duncan's test (p<0.1).The mean values of the pre-inoculation treatments for the variables nodulation (number and dry mass of nodules), biomass (root and shoot dry mass), nitrogen concentration in shoot and grain, and grain yield were compared by pairs with the standard treatment (inoculation at sowing using RIZOLIQ) using the bilateral Dunnett test (p<0.1)(Dunnet, 1964).

Bacterial survival on inoculated seeds
The recovery and quantification of viable Bradyrhizobium cells on the seed surface are shown in Table 3.The mean number of viable bacterial cells on seeds inoculated with RIZOLIQ ® on the day of sowing but not treated chemically (standard inoculation) was 3.3 log 10 CFU of seed -1 .

Pre-inoculation with RIZOLIQ LLI
® + Premax ® cell protector at 30 days before sowing seeds chemically treated with different products ensured the survival of The means followed by the same letter were not significantly different using the Duncan test (p<0.05).Data transformed to log10 (CFU + 1).CV=Coefficient of variation.
In addition, the mean bacterial population on seeds pre-inoculated at 45 days before sowing and treated with Imidacloprid (Rocks ® ) (LLI45R) and seeds pre-inoculated at 60 days before sowing and treated with Imidacloprid and Fipronil+Thiophanate-methyl+Pyraclostrobin (Rocks ® and Standak ® Top) (LLI60R and LLI60S) was not significantly different from the population on seeds subjected to standard inoculation.The bacterial population in seeds of other treatments pre-inoculated at 45 and 60 days before sowing (LLI45S, LLI45MC, LLI45M, LLI60MC, and LLI60M) was smaller than that on seeds inoculated with RIZOLIQ ® on the day of sowing.

Symbiotic efficiency and grain yield
The number of nodules, dry mass of nodules, dry mass of roots and shoot, nitrogen concentration in shoot and grain, and grain yield in Londrina, Santa Tereza do Oeste, Pato Branco and Ponta Grossa are shown in Tables 4, 5, 6 and 7, respectively.Nodulation in the flowering period was high in all tested locations and all treatments.In most cases, preinoculation with RIZOLIQ LLI ® + Premax ® cell protector at 30, 45 and 60 days before sowing did not reduce nodule number and dry mass compared with inoculation with RIZOLIQ® on the day of sowing.In Londrina, preinoculation at 45 days before sowing and chemical treatment with Imidacloprid (Rocks ® ) (LLI45R) and preinoculation at 30 days before sowing and treatment with Fipronil + Thiophanate-methyl + Pyraclostrobin (Standak ® Top) (LLI30S) caused a signifcant increase in the number of nodules compared with standard inoculation.In addition, pre-inoculation at 45 days before sowing and treatment with Fipronil+Thiophanate-methyl+Pyraclostrobin (Standak ® Top) (LLI45S) caused a significant increase in nodule dry matter compared with standard inoculation (Table 4).In Santa Tereza do Oeste, pre-inoculation at 30 days before sowing and chemical treatment with Imidacloprid (Rocks ® ) (LLI30R) caused a significant increase in nodule dry mass (Table 5).In Pato Branco, pre-inoculation at 30 and 60 days before sowing and treatment with Imidacloprid (Rocks ® ) (LLI30R and LLI60R) increased nodule number and dry mass, respectively, compared to standard inoculation (  and LLI60MC), preinoculation at 45 days before sowing and treatment with Metalaxyl-M+Fludioxonil (Maxim ® XL) (LLI45M), and pre-inoculation at 30 days before sowing and treatment with Imidacloprid (Rocks ® ) (LLI30R) yielded a lower nodule dry mass when compared with the standard inoculation (Table 7).The results of the other treatments did not differ from the standard inoculation.
Plant biomass also showed a good response to preinoculation with RIZOLIQ LLI ® + Premax ® cell protector.Pre-inoculation at 45 days before sowing and treatment with Imidacloprid (Rocks ® ) (LLI45R) in Londrina, and preinoculation at 45 days before sowing and treatment with Fipronil+Thiophanate-methyl+Pyraclostrobin (Standak ® Top) The means followed by the same letter were not significantly different using the Duncan test (p<0.05).CV: Coefficient of variation; HSD: honestly significant difference.The means followed by the same letter were not significantly different using the Duncan test (p<0.05).CV=Coefficient of variation.HSD: Honestly significant difference.The means followed by the same letter were not significantly different using the Duncan test (p<0.05).CV: Coefficient of variation; HSD: honestly significant difference.
The amount of nitrogen provided to the shoot and grain by pre-inoculation at 30 days before sowing with RIZOLIQ LLI ® + Premax ® cell protector was similar that of the standard inoculation in all analyzed areas.In Londrina, pre-inoculation at 45 and 60 days before sowing and treatment with Metalaxyl-M+Fludioxonil (Maxim ® XL) (LLI60M and LLI45M) and pre-inoculation at 45 days before sowing and treatment with Imidacloprid (Rocks ® ) (LLI45R) caused a decrease in nitrogen concentration in the shoot.Pre-inoculation at 60 days before sowing and treatment with Imidacloprid (Rocks ® ) (LLI60R) and pre-inoculation at 45 days before sowing and treatment with Metalaxyl-M+Fludioxonil+Thiabendazole with Thiamethoxam (Maxim ® Advanced+Cruiser ® ) (LLI45MC) promoted a decrease in the concentration of nitrogen in the grains (Table 4).In Santa Tereza do Oeste, pre-inoculation at 60 days before sowing and treatment with ® , respectively) (LLI45S and LLI45R) caused an increase in nitrogen concentration in the grains, and the concentration was similar to that of the standard inoculation (Table 6).However, pre-inoculation at 45 days before sowing and treatment with Imidacloprid (Rocks ® ) (LLI45R) decreased the nitrogen concentration in the grains (Table 7).
Although pre-inoculation treatments showed differences in nodulation, plant biomass, and nitrogen concentration in Londrina, Santa Tereza do Oeste, and Ponta Grossa, grain yield by pre-inoculation with different chemical products at 30, 45, and 60 days before sowing was statistically similar to that of the standard inoculation.In Pato Branco, because of the occurrence of pests and diseases in the study period, the mean grain yield in the region was lower than that in the other regions and presented variability between the evaluated treatments.In Pato Branco, only pre-inoculation at 30 and 45 days before sowing and treatment with Imidacloprid (Rocks ® ) (LLI45R and LLI30R) and pre-inoculation at 30 days before sowing and treatment with Metalaxyl-M+Fludioxonil (Maxim ® XL) (LLI30M) caused lower yields compared with the standard inoculation, whereas grain yield in nine of the twelve pre-inoculation treatments tested was not significantly different from that in the standard inoculation, suggesting the potential use of this strategy even under unfavorable conditions.

DISCUSSION
Soybean crops require high amounts of nitrogen, and it is estimated that 80 kg of nitrogen is needed for producing 1000 kg of grains (Hungria et al., 2007).The supply of nitrogen to soybean culture in Brazil is provided primarily by inoculation of Bradyrhizobium spp., and the major factor involved in the successful supply of nitrogen by these bacteria to crops is the guarantee of bacterial survival on the seeds (Dall'Agnol, 2016).
In this study, the pre-inoculation of soybean seeds was tested with RIZOLIQ-LLI ® + Premax ® cell protector before sowing.Pre-inoculation at 30 days before sowing and treatment with Metalaxyl-M+Fludioxonil and Imidacloprid (Maxim ® XL and Rocks ® , respectively) promoted an increase in the population of Bradyrhizobium spp. on the seeds.The increase in the bacterial population may have been stimulated by the integrated action of Premax ® cell protector combined with a positive effect of the products used in seed treatment (Maxim ® XL and Rocks ® ).However, this hypothesis was not investigated.Although in the pre-inoculation treatments LLI45S, LLI45MC, LLI45M, LLI60MC, and LLI60M, bacterial survival on the seeds was lower than that in the standard inoculation in all the studied areas (Table 3), in all treatments were recovered at least 10 5 viable cells seed -1 as recommended for soybean crop (Hungria et al., 2007) and this difference did not significantly decrease nodulation or grain yield, evidencing that the survival of viable cells is enough to ensure the nitrogen supply required by the crop (Tables 4 to 7) .
Although some authors report the toxicity of plant protection products used in seed treatment to the bacteria present in the inoculants (Araujo and Araújo, 2006;Hartley et al., 2012) Nevertheless, in the present study, the use of the cell protector ensured the survival of bacteria even with chemical treatments and these results are in accordance with Araújo et al. (2017), who hypothesize that for establishing pre-inoculation technologies, especially in seeds treated chemically with crop protection products, the use of chemicals that improve bacterial survival is necessary.
Pre-inoculation of soybean seeds before sowing may stimulate the use of inoculants in these crops and optimize sowing.Some studies have shown the efficacy of seed pre-inoculation on bacterial survival (Gemell et al., 2005) and crop yield (Zilli et al., 2010;Anghinoni et al., 2017) even with a cell protector (Marks et al., 2013).However, to date, few inoculation technologies used before sowing ensured bacterial survival on seeds treated chemically with crop protection products under long storage periods and ensured good agronomic performance in the field.Therefore, in this study, the use of the Premax ® cell protector promoted the survival of the bacteria present in the RIZOLIQ LLI ® inoculant, even on seeds treated chemically with crop protection agents during storage.
The crop yield was the parameter that best reflected the efficacy of pre-inoculation.The grain yield achieved by pre-inoculation at 30, 45, and 60 days before sowing in the three study areas was similar to that of the standard inoculation, demonstrating the effectiveness of pre-inoculation with these two producs.Positive results in soybean grain yield with pre-inoculation technology have already been found in other studies.Silva et al. (2018) also observed that even with a drastic reduction in numbers of cells with chemical treatmens of the seeds, pre-inoculation 10 days before sowing did not reduce nodulation nor grain yield.Zilli et al. (2016) verified that the pre-inoculation five days before sowing had a similar performance to the standard inoculation.
Although grain yield in three treatments (LLI45M, LLI30M, and LLI30R) in Pato Branco was lower than that in the standard inoculation, grain yield in nine of the twelve pre-inoculation treatments tested was not significantly different from that of the standard inoculation.Araujo et al. (2017) observed that preinoculation of soybeans with cell protector was effective for up to 30 days before sowing, and grain yields were similar to that of the standard inoculation even under adverse environmental conditions, indicating the potential of using this technology even under unfavorable conditions.
Several studies had reported the effectiveness of the pre-inoculation technology in soybean crop for five days before sowing (Zilli et al., 2016), up to 10 days (Anghinoni et al., 2017;Silva et al., 2018) and up to 30 days before sowing (Araujo et al., 2017).However, this study is the first to report the efficacy of pre-inoculation of soybean seeds for up to 60 days before sowing.Therefore, we showed that the pre-inoculation of soybeans with RIZOLIQ LLI ® +Premax ® cell protector can be used in seeds chemically treated with Fipronil+Thiophanate-methyl+Pyraclostrobin (Standack ® Top), Metalaxyl-M+Fludioxonil+Thiabendazole with Thiamethoxam (Maxim ® Advanced+Cruiser ® ), Metalaxyl-M+Fludioxonil (Maxim ® XL) and Imidacloprid (Rocks ® ) and stored for up to 60 days at room temperature without loss in nodulation, plant biomass, nitrogen concentration in the plant shoot and grains, and grain yield (Tables 4 to 7).

Conclusions
The inoculation of soybean seeds with RIZOLIQ LLI ® +Premax ® cell protector allowed the survival of bacterial cells on seeds chemically treated with crop protection agents and stored for up to 60 days before sowing.
The pre-inoculation of soybean seeds with RIZOLIQ LLI ® +Premax ® cell protector up to 60 days before sowing can be performed on seeds treated chemically without impairment of nodulation, plant biomass, nitrogen concentration in shoot and grain, and crop yield and is an efficient and practical inoculation strategy for sowing soybean crops.
) (LLI60S and LLI60MC) and pre-inoculation at 45 days before sowing and treatment with Metalaxyl-M+Fludioxonil (Maxim ® XL) (LLI45 M) decreased the nitrogen concentration in the shoot.However, these pre-inoculation treatments did not decrease nitrogen concentration in the grains.In Pato Branco, pre-inoculation at 45 days before sowing and treatment

Table 1 .
Chemical characteristics of the soil of the experimental areas located in Londrina, Santa Tereza, Pato Branco, and Ponta Grossa, state of Paraná, Brazil.

Table 2 .
Pre-inoculation treatments of soybean seeds.

Table 3 .
Viability (colony forming units-CFU) of Bradyrhizobium japonicum in soybean seeds inoculated with RIZOLIQ® on the day of sowing or pre-inoculated with RIZOLIQ-LLI ® + Premax ® cell protector.

Table 5 .
Number of nodules.dry mass (DM) of the nodules.dry mass of roots and shoot.nitrogen concentration in shoot and grains.and yield of soybean cultivar BMX Apolo RR treated with different pesticides and pre-inoculated in Santa Tereza do Oeste, Paraná, Brazil. in the 2015/2016 harvest.

Table 6 .
Number of nodules.dry mass (DM) of the nodules.dry mass of roots and shoot.nitrogen concentration in shoot and grains.and yield of soybean cultivar BMX Potência RR treated with different pesticides and pre-inoculated in Pato Branco, Paraná, Brazil. in the 2015/2016 harvest.

Table 7 .
Number of nodules.dry mass (DM) of the nodules.dry mass of roots and shoot.nitrogen concentration in shoot and grains.and yield of soybean cultivar BMX Apolo RR treated with different pesticides and pre-inoculated in Ponta Grossa, Paraná, Brazil. in the 2015/2016 harvest.