Pre-inoculation of soybean seeds : Effects on survival of Bradyrhizobium elkanii , nodulation and crop yield

1 Department of Agronomy, Universidade Estadual do Centro-Oeste (UNICENTRO), Rua Simeão Camargo Varela de Sá, 03, Vila Carli, 85040-080, Guarapuava, Paraná, Brazil. 2 BASF S/A, Alameda Adélia Salvador Bernardo, 243, 13917-196, Jaguariúna, São Paulo, Brazil. 3 Department of Veterinary Medicine, Universidade Estadual do Centro-Oeste (UNICENTRO), Rua Simeão Camargo Varela de Sá, 03, Vila Carli, 85040-080, Guarapuava, Paraná, Brazil. 4 College of Agronomy and Veterinary Medicine – FAMV, Universidade de Passo Fundo (UPF), Avenida Brasil Leste, 285, São José, 99052-900, Passo Fundo, Rio Grande do Sul, Brazil.


Inoculation
of nitrogen-fixing bacteria of the Bradyrhizobium genus in soybean is a common practice in Brazil (Alves et al., 2003;Zilli et al., 2010).These bacteria are capable of fixing atmospheric nitrogen and provide it to the plant in exchange for photoassimilates (Bergersen, 1997;Fagan et al., 2007).As a result, the *Corresponding author.E-mail: isandini@hotmail.com.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License inoculation has allowed achievement of high soybean yields while reducing or even eliminating the need for application of nitrogen fertilizers (Mendes et al., 2003;Hungria et al., 2006;Machineski et al., 2018).
In order to ensure the viability of bacterial cells, farmers usually apply the inoculant on the soybean seeds at the day or up to 24 h before sowing (Silva Júnior et al., 2016).Application in-furrow during the sowing process can also be performed (Campo et al., 2010), but it is a less common practice due to the need for adaptation of planting machines and costs associated with it.The time required to inoculate the seeds on-farm has been considered a challenge from an operational point of view.In addition, seed treatment with insecticides and fungicides has often been performed to protect the seeds from pests, seed-borne and soil-borne pathogens (Goulart 1998), but the toxicity of these products has been considered a risk to people who manipulate the treated seeds as well as for the bacteria inoculated on the seeds with the purpose of fixing nitrogen.Therefore, new technologies need to be explored with the aim of overcoming these challenges.
As an alternative, seed companies have been adopting the industrial seed treatment, in which the company offers soybean seed already treated with insecticides, fungicides and pre-inoculated with Bradyrhizobium spp.(Brzezinski et al., 2015).This practice has allowed to speed up the operational process at sowing as well as to reduce the risk of exposure to toxic products.However, when fungicides and insecticides are added to the seed, the viability of the nitrogen-fixing bacteria could be jeopardized to a point where the number of viable bacteria is not enough to infect the plant effectively and form nodules (Hartley et al., 2012;Araujo et al., 2017).Thus, the compatibility of fungicides and insecticides represents a great challenge to the seed treatment (Campo et al., 2009;Pereira et al., 2010a;Silva Neto et al., 2013;Gomes et al., 2017).Very few studies were performed to evaluate the association of pre-inoculation and industrial seed treatment (Anghinoni et al., 2017;Machineski et al., 2018), and little is known about the maximum period of time before sowing that the seed could be treated without compromising the viability of the bacteria, nodulation of the plants and soybean yield (Pereira et al., 2010b;Zilli et al., 2010).Some authors have mentioned up to five days (Zilli et al., 2010), while other demonstrated that the period could be expanded to 30 days (Araujo et al., 2017) or even up to 60 days (Machineski et al., 2018) when a protectant is added to the inoculant.
Technologies that include polymers associated with the inoculation of Bradyrhizobium spp.could be an alternative to the conventional inoculants by avoiding the dehydration of the bacterial cells (Deaker et al., 2007) and protect them from toxic effects of fungicides and insecticides (Zilli et al. 2009;Campo et al. 2010).As a result, survival of the nitrogen-fixing bacteria could occur at levels that do not compromise the establishment of the symbiotic relationship and supply of nitrogen to the plant (Denardin and Freire, 2000;Fernandes Júnior et al., 2009;Pereira et al., 2010c;Fernandes Júnior, 2012;Marks et al., 2013;Araújo et al., 2017).Since little is known about the use of a new technology (HiCoat ® ) developed with the aim of enabling pre-inoculation of soybean seeds treated with fungicides and insecticides, the objectives of this study were i) to determine whether the HiCoat ® technology allows the survival of B. elkanii after storage of the inoculated seeds; and ii) to evaluate whether the pre-inoculation and seed treatment with fungicides and insecticides reduces the physiological quality of seed and soybean yield.

Seed treatments with fungicides and insecticides
All the fungicides and insecticides used in this study were commercial products.For application of the chemical treatments, 1 kg of seed of the soybean cultivar ND5909RR was placed into a plastic bag.For each treatment, the products were mixed according to their respective doses and water was added to bring the volume to 5 ml (Table 1).The mixture was applied on the seeds followed by agitation to uniformly distribute the products.The bags were maintained on a laboratory bench for the seeds to dry for 1 h.In all experiments, an untreated check without application of fungicide and insecticide was included.

Inoculant and inoculation procedure
The inoculants evaluated in this study were commercial products manufactured and commercialized by BASF S/A.Inoculant was applied using the HiCoat ® technology, which is a mixture of 300 g of peat inoculant (Adhere HC ® , B. elkanii strain SEMIA 587 and 5019 at 5 x 10 9 CFU ml -1 ), 300 ml of liquid inoculant (Gelfix ® , B. elkanii strain SEMIA 587 and 5019 at 5 x 10 9 CFU ml -1 ), 150 ml of polymer S30 and 200 ml of distilled water.A volume of 9.5 ml kg -1 was applied on the seeds.For all experiments, an untreated check without inoculation was added along with a standard inoculation performed only with liquid inoculant (Gelfix ® , B. elkanii strain SEMIA 587 and 5019 at 5 x 10 9 CFU ml -1 ).

Survival of bacterial cells
Experiments were carried out under controlled conditions in the Laboratory of Field Crops of the Universidade Estadual do Centro-Oeste (UNICENTRO) in Guarapuava, Paraná, Brazil.The survival of B. elkanii was determined as the number of colony-forming units (CFU) recovered from soybean seeds according to the methodology established by Brazilian regulations (Mapa, 2010).A hundred seeds were placed into a 250-ml sterilized Erlenmeyer that contained 100 ml of 0.85% (w/v) NaCl solution with three drops of Tween 80. Flasks were agitated in an orbital shaker for 15 min.Serial dilutions in 0.85% (w/v) NaCl were prepared.A 100-L aliquot from each dilution was placed onto sterilized Petri dishes containing CRYMA culture media (Fred andWaksman, 1928 cited by Mapa, 2010).After inoculation, the plates were maintained at 28°C for 10 days.The number of CFU was counted under an optical microscope, and only the plates between 30 to 300 colonies were considered.The results were expressed as number of CFU Type of product: I = insecticide; F = fungicide; N = nematicide.b All products were mixed and distilled water was used to bring the volume to 5 ml that was applied to one kilogram of seed.c HiCoat ® Technology: a mixture of 300 g of peat inoculant (Adhere HC ® , Bradyrhizobium elkanii strain SEMIA 587 and 5019 at 5 x 10 9 CFU ml -1 ), 300 ml of liquid inoculant (Gelfix ® , Bradyrhizobium elkanii strain SEMIA 587 and 5019 at 5 x 10 9 CFU ml -1 ), 150 ml of polymer S30 and 200 ml of distilled water.d Treatment 10: standard inoculation was performed at the day of planting with liquid inoculant (Gelfix ® .Bradyrhizobium elkanii strain SEMIA 587 and 5019 at 5 x 10 9 CFU ml -1 ).e Treatment 11: no inoculation with Bradyrhizobium elkanii was performed.
per seed.

Effect of seed packing and storage period on survival of bacterial cells
The treatments consisted of two types of seed packing (plastic and paper), three seed treatments (Stadank ® Top + standard inoculation; HiCoat ® technology only; Standak ® Top + HiCoat ® technology), and seven storage periods (0, 3, 9, 15, 30, 45 and 60 days after inoculation).The bags used for seed packing were transparent plastic bags and brown paper bags.The dose of Standak ® Top (fipronil + thiophanate-methyl + pyraclostrobin) was of 2 ml kg -1 of seed.The survival of B. elkanii was determined as described above.The study was carried out in a completely randomized design with three replications.

Effect of pre-inoculation on physiological quality of soybean seed
The experiment design was a complete randomized with eight replications.Eleven seed treatments were evaluated (Table 1).Fifty soybean seeds were placed on a "germitest" paper sheet previously soaked in distilled water, and another sheet was used to cover the seeds.The paper sheets were rolled and placed in an incubator at 25C in the dark.Each sheet was considered as a replication.On the fifth day, seed vigor was evaluated according to Brasil (2009).The rolled papers were returned to the incubator for three more days and then germination, normal seedlings and non-germinated seeds were determined according to Brasil (2009).All data were expressed as percentage of the total seeds placed on each paper sheet.

Effect of pre-inoculation on nodulation of soybean plants
The experiment was carried out under field conditions in 1 m wide and 0.20 m height flowerbeds.The experimental design was a randomized complete block with four replications and eleven treatments (Table 1).Each experimental plot was composed of one row, and the rows were spaced 0.50 m apart.Twenty-five seeds were manually placed in each row at 2 cm deep and covered with soil.After emergence, only 10 seedlings were maintained in each row.At full flowering (R2) stage, three soybean plants were selected from each row.Plant root system was collected by placing a cylindrical metal device with 10 cm diameter and 10 cm height.
The soil was washed off the roots and the nodules were collected with a metal screen.The number of nodules was determined.The nodules were placed in paper bags that were maintained in a drying oven at 65ºC for 72 h, and after weighting the dry mass was determined.

Effect of pre-inoculation on soybean yield
The experiments were conducted under field conditions in Guarapuava, Paraná, Brazil, during the 2010/2011 growing season.The experimental site is located in a region with approximately 1,100 m altitude and humid subtropical climate (Cfb).The soil was classified as a Brown Latosol (Embrapa, 2013).The experiment was carried out in a randomized complete block design, six replications and eleven treatments (Table 1).The experimental plot was composed of four rows spaced 0.40 m apart and 5.5 m length.
The experiment was carried out twice.The first experiment was sowed on 28 Nov 2010 corresponding to the seed stored for 59 days, and the second experiment was sowed on 10 Dec 2010 with the seeds stored for 71 days.Maize was grown during the previous summer season, while the experimental area remained under fallowing during the previous winter season.Desiccation was performed with glyphosate herbicide (720 g ha -1 a.i.) 30 days before sowing of soybean.Fertilization was performed at the sowing with 46 kg ha -1 of P2O5 and 60 kg ha -1 of K2O.Management of weeds, pests and diseases was performed when needed.
Ten plants were collected from the two central rows of each plot.The number of pods and number of grains per plant were determined.The remaining plants in the two central rows of the experimental plot were manually harvested and threshed.The soybean seeds were weighted and seed moisture was determined.Soybean yield was determined in kg ha -1 at 13% moisture.A sample was taken and used to determine the thousand-grain mass by counting and weighing 300 grains.

Data analysis
Homogeneity of variances was analyzed by Bartlett's test.As a result, percentage values and counted values were transformed to√x+1.Analysis of variance was then performed, and means were Sandini et al. 2683 compared by Tukey's Test at 5% of probability using the statistical program Sisvar (Ferreira, 2000).

Survival of bacterial cells
The number of CFU recovered from soybean seeds was significantly affected by seed packing, seed treatment, storage period and all the interactions (Table 2).The survival of bacterial cells was reduced over time when soybean seeds received the standard seed treatment independently of seed packing, and at 60 days after inoculation no bacterial cells were recovered (Figure 1).However, when seeds were inoculated using the HiCoat ® technology with and without the fungicide and insecticide, the number of CFU was maintained above 2 x 10 6 CFU per seed even at 60 days after inoculation (Figure 1).
Certain combinations of fungicides (Campo et al., 2009;Gomes et al., 2017;Costa et al., 2013) as well as insecticides (Pereira et al., 2010c) can be detrimental to Bradyrhizobium spp.due to the toxicity of the active ingredient, pH and solvents used in the formulations (Hungria and Campo, 2000), but the magnitude of the effect depends on the type and the period of contact between the chemicals and the inoculant (Costa et al., 2013).A previous study using a standard inoculation protocol with liquid products without polymers or any other type of protectant, demonstrated that 62% of bacteria inoculated on soybean seeds treated with fungicides died two hours after inoculation, and 95% of the bacteria died after 24 hours (Campo et al., 2009).In another study, the number of bacterial cells recovered from soybean seeds decreased in the first 2 h after inoculation, but it stabilized at 24, 48, 72 and 96 h after inoculation (Costa et al., 2013).
Nevertheless, in this study when the HiCoat ® technology was used for inoculation, no significant adverse effect was observed on the survival of the bacterial cells.This technology includes a polymer, which may have acted as a protectant for the bacteria mainly against dehydration (Deaker et al., 2007) or toxicity of the chemical products (Pereira et al., 2010a).Previous observations confirmed the ability of polymers to protect cells from changes in the environmental conditions after inoculation of the seeds (Tittabutr et al. 2007).Use of polymers as carriers for inoculants allowed survival of B. japonicum for up to 180 days after inoculation of cowpea (Fernandes Júnior et al., 2009).Moreover, Araujo et al. (2017) pre-inoculated soybean seeds with commercial products that contained protectants, and verified that satisfactory populations of the Bradyrhizobium spp.were recovered from the seeds for up to 30 days after inoculation.All these results indicate that pre-inoculation can be used without compromising the survival of Bradyrhizobium spp.cells.a Variables in percentage were transformed to √x+1 for statistical analysis.

Physiological quality of soybean seed
The seed treatments had significant effect on seed vigor and non-germinated seeds, but they did not affect the seed germination and abnormal seedlings (Table 3).Seed vigor reduced while the percentage of nongerminated seeds increased when imidacloprid + thiodicarbe + fludioxonil + metalaxyl-M + abamectin + HiCoat ® were applied to the soybean seeds compared to fipronil + fludioxonil + metalaxyl-M + HiCoat ® .However, when the treatment with imidacloprid + thiodicarbe + fludioxonil + metalaxyl-M + HiCoat ® was applied, there was no significant difference from the other seed treatments (Table 4).Hence, the negative effect is due to the association of abamectin with imidacloprid + thiodicarbe + fludioxonil + metalaxyl-M rather than the active ingredients by themselves because no significant difference from other treatments was observed when abamectin was added with other combinations of fungicides and insecticides.
In this study, seed germination was not affected by the pre-inoculation of B. elkanii.In addition, none of the treatments with fungicides and insecticides associated with HiCoat ® had lower seed germination when compared to the standard inoculation after 60 days of storage of the seeds.Similarly, neither peat-based inoculant nor liquid inoculant containing polymeric additives had an effect on the germination of soybean seeds (Tittabutr et al., 2007).It was previously showed that polymers can interfere with insecticide phytotoxicity, but did not affect the physiological quality of soybean seeds (Kumar et al., 2007;Pereira et al., 2010c;Camargo et al., 2017;Fagundes et al., 2017).The absence of a negative effect on the seed germination demonstrates that preinoculation can be performed on soybean seeds.However, caution must be taken to avoid the use of the imidacloprid + thiodicarbe + fludioxonil + metalaxyl-M + abamectin + HiCoat ® due to the reduction of seed vigor.Low seed vigor can negatively affect the initial establishment of the plants and this can reduce soybean yield (Scheeren et al., 2010).

Nodulation of plants
Biological nitrogen fixation by Bradyrhizobium spp. is complex and the efficiency of the nodulation is mediated by internal factors, for example plant hormones and availability of photoassimilates, along with external factors such as temperature, oxygen, water, nutrients and products applied during seed treatment (Fagan et al., 2007).In this study, there was no significant effect of preinoculation associated with fungicides and insecticides on the number of nodules per plant, mass of nodules per plant and mass per nodule (Table 5).Similarly, in a study performed by Pereira et al. (2010a), the treatment of soybean seeds with the fungicides carbendazin + thiram or thiabendazole + thiram associated or not with polymer, and independently of the application time, did not affect the establishment and development of nodules when the seeds were inoculated with Bradyrhizobium.Moreover, there was no significant difference among the preinoculation of cowpea seeds with either polymer or peat inoculant until 14 days after inoculation, but at 35 days the nodulation and dry mass with polymer inoculant was significant higher than the peat inoculant (Silva Júnior et al., 2016).The non-inoculated control (Table 6) had the lowest number and mass of nodules per plant, but the difference from the other treatments was not significant due to the high variation of the data.The presence of nodules on the plants for the non-inoculated control (Table 6) was due to natural contamination by nitrogen-fixing bacteria as observed in the study by Pereira et al. (2010a).Naturalized population of Bradyrhizobium spp. in the soil of a field previously grown with soybean can provide satisfactory levels of nodulation and no response by the application of different doses of inoculant may be observed (Campos 1999).However, Nishi and Hungria (1996) verified that even in soils with elevated population of Bradyrhizobium spp.there was increase in the number and mass of nodules by the inoculation.

Soybean yield
Soybean yield and thousand grain mass were significantly affected by experiment (Table 7).The means of these two variables were significantly lower in the experiment 2 compared to Experiment 1 (Table 8).The difference in yield between the two experiments was likely due to the planting time since soybean is sensitive to photoperiod (Farias et al., 2007) rather than the difference in the storage period of the seeds because there was significant lower yield in the standard inoculation treatment (Table 8).At the latitude in the Southern Hemisphere that this study was performed, the photoperiod in December is longer than November due to the approach of summer solstice.Thus, sowing carried out in November leads to development of taller plants that are also more productive when compared to plantings in December, in which earlier flowering occurs even when foliar index is still low (Fietz and Rangel, 2008).
There was no significant effect of pre-inoculation and seed treatment with fungicides and insecticides as well as the interaction between seed treatment and experiment on soybean yield, thousand grain mass, number of pods per plant and number of grains per pod (Table 7).Similarly, Machineski et al. (2018) verified that pre-inoculation of soybean seeds using a cell protector was efficient in maintaining the bacterial inoculant viable on the seed for up to 60 days and did not negatively affect soybean yield.Anghinoni et al. (2017) also verified that soybean seeds industrially treated with fludioxonil fungicide and thiametoxan insecticide can be inoculated and stored up to 10 days before sowing, with no adverse effects on grain yield.In addition, Silva Júnior et al. (2016) found out that cowpea yield was not significantly different from the standard inoculation and positive control (70 kg ha -1 of N) when peat and polymer inoculants were used (Silva Júnior et al., 2016).All these results are contrary to what was found by Brzezinski et al. (2015) and Zilli et al. (2009), where the authors mentioned that treatment of soybean seeds with insecticide and fungicides before sowing hinders the establishment of soybean in the field, and reduces nodulation of the plants, respectively.However, the difference may be due to the absence of polymers in both studies and the fact that seeds were stored for 240 days before sowing in the work by Brzezinski et al. (2015).
This study was performed in a Brown Latosol, with  to be overcome to allow the use of new seed treatments at a global scale (O'Callaghan, 2016).Some authors have suggested that if pre-inoculation of soybean seeds is desired, some protective inoculant technology must be used (Araujo et al. 2017).Our findings indicated that the fungicide and insecticide associated with the HiCoat ® technology can be used in the industrial seed treatment and seeds can stored for up to 71 days without compromising the nodulation of plants and soybean yield.Further studies are required to investigate the effect of pre-inoculation using the HiCoat ® technology with other species and strains of Bradryrhizobium, different soybean cultivars, and conditions of sandy soil with low organic matter content and no history of cultivation of soybean.

Figure 1 .
Figure 1.Mean colony forming-units (log10 of CFU per seed) of Bradyrhizobum elkanii for each seed packing (paper bag or plastic bag) and pre-inoculation with the HiCoat ® technology or standard inoculation over different storage periods.Chemical seed treatment was performed with Standak ® Top (fipronil + thiophanate-methyl + pyraclostrobin).

Table 1 .
Seed treatments used in the experiments to evaluate the pre-inoculation of Bradyrhizobium elkanii on soybean seeds using the HiCoat ® technology.

Table 2 .
Probability values from the analysis of variance for the effect of seed packing, seed treatment and storage period for the number of colony forming-units (CFU) of Bradyrhizobium elkanii.

Table 3 .
Mean square and probability values from analysis of variance for the effects of seed treatment on the physiological quality of soybean seeds.

Table 4 .
Physiological quality of soybean seeds treated with different insecticides and fungicides and pre-inoculation at 71 days before sowing.
a Means with the same letter do not statistically differ at P ≤ 0.05 by Tukey's test.All statistical analysis and mean comparisons were performed on square-root-transformed data.b Treatment 10: standard inoculation was performed at the day of planting.c Treatment 11: no inoculation with Bradyrhizobium elkanii was performed.

Table 5 .
Mean square and probability values from analysis of variance for the effects of seed treatment on the nodulation of soybean plants.

Table 6 .
Nodulation of soybean plants with pre-inoculation of Bradyrhizobium using the HiCoat ® technology and different fungicide and insecticide seed treatments.
a Means followed by the same letter do not statistically differ by Tukey's Test at P ≤ 0.05.b Treatment 10: standard inoculation was performed at the day of planting.c Treatment 11: no inoculation with Bradyrhizobium elkanii was performed.

Table 7 .
Mean square and probability values from analysis of variance of the effects of seed treatment on the soybean yield and yield components.

Degress of fredom Mean square (P values) Yield (kg ha -1 ) Thousand grain mass (g) Number of grains per plant Number of pods per plant
Zilli et al. (2010) clay and high organic matter.In a study performed under field conditions with sandy soil with low organic matter content and with soybean grown for the first time,Zilli et al. (2010)did not observe significant difference between the pre-inoculation at 5 days before planting compared to the standard inoculation on the number and mass of nodule formed on soybean plants, and both were superior to the check and nitrogen fertilization.Therefore, studies need to be carried out under different field conditions to effectively determine the effect of pre-inoculation of soybean seeds.Scientific, technological and commercial challenges need

Table 8 .
Yield and yield components of soybean with pre-inoculation of Bradyrhizobium elkanii using the HiCoat ® technology and different fungicide and insecticide seed treatments in two experiments under field conditions.
a Means followed by the same capital letter for experiments and lower case for seed treatment do not statistically differ by Tukey's test at P ≤ 0.05.b Treatment 10: standard inoculation was performed at the day of planting.c Treatment 11: no inoculation with Bradyrhizobium elkanii was performed.