Inoculation with plant-growth promoting bacteria improves seed germination and initial development of Brachiaria decumbens

: The objective of this research was to evaluate the inoculation and co-inoculation of bacteria with biotechnological potential, isolated from Brachiaria decumbens Stapf. and Brachiaria humidicola (Rendle) Schweickerdt, under germination and seedling growth of B. decumbens cv. Basilisk, as well as, to verify the infl uence of the co-inoculum in the soil indigenous bacterial community. For this, two assays in a completely randomized design were set up. The experimental period was 21 days. In a germination chamber, 25 treatments were evaluated (24 bacterial inoculants and a control – without inoculation). In greenhouse assay, were used fi ve co-inoculations (bacterial consortium). The bacterial consortium was obtained based on the bacterial strain performance in the germination test. In addition, the control and one treatment with mineral fertilizer (NPK) were tested. In germination test, the seed inoculation promoted increases of 61, 40, 144, 82, 6, 96, 91 and 52% in germination vigor, speed germination index, number of absorbent hairs, number of plumules, primary root length, hypocotyl length, total length, and dry matter of seedlings, respectively, when compared to control. The co-inoculation also increased the growth parameters of B. decumbens plants when compared to the control treatment. In addition, promoted changes in the soil bacterial community structure. Becoming an important strategy to increase the germination rate and germination speed of B. decumbens plants.


INTRODUCTION
Forage grasses belonging to Urochloa genera, commonly known as Brachiaria, are cultivated worldwide (Cheruiyot et al. 2018).In Brazil alone, there are approximately 200 million hectares of pasture, which has a high degradation degree caused mainly by non-replenishment of soil nutrient, inadequate management and low seed germination rates under fi eld conditions (Hungria et al. 2016, Bono et al. 2019).
The forage grass establishment in the fi eld depends of the seed germination potential, being a crucial phase in determining pasture uniformity and yield potential (Silva et al. 2019a).Thus, the technologies that aim to increase, besides contributing to the plant-growth are essential for more effi cient pasture management (Freitas et al. 2019).In this scenario, plant growth promoting bacteria (PGPB) can be an excellent strategy for improving seed germination rate and establishment of forage grasses.The PGPBs act as biological control agent against diseases, producing phytohormones and hydrolytic enzymes necessary to seed germination and plant-growth, favoring the fast plant establishment (Andrade et al. 2019, Araújo et al. 2012, Kim et al. 2012, Lima et al. 2018, Oliveira et al. 2018, Sammauria et al. 2020, Souza et al. 2015, Terra et al. 2019).
Although the single inoculation (or the inoculum using only one microorganism specie), can promote the plant-growth, the coinoculation with more than one microorganism species per inoculum, can increase the survival of the inoculated microbial population, increasing the probability of success due to the synergistic effect of different growth-promoting mechanisms (Sánchez et al. 2014, Sá et al. 2019).
The effect of PGPB inoculation depends on the plant genotype (Santos et al. 2019).This is because the plant species to be inoculated has a strong influence on the microbial community due to a wide variety of organic compounds that recruit the most distinct microbial groups to the rhizosphere soil.These microbial groups are often beneficial to the plant, contributing to nutrient acquisition, pathogen protection and resistance to environmental stress (Araújo et al. 2012, Wemheuer et al. 2016, Murphy et al. 2016, Santos et al. 2019).However, the inoculation of bacteria isolated from plant species and/or specific cultivars can facilitate colonization due to adaptation to biotic conditions, favoring the competition with native microbiota the plant and soil (Bashan et al. 2014, Sammauria et al. 2020).
Thus, our aim was to evaluate the inoculation and co-inoculation of B. decumbens cv.Basilisk seeds with growth-promoting bacteria isolated from B. decumbens Stapf.and B. humidicola (Rendle.)Schweickerdt on germination speed and seedling growth, as well as verifying the influence of co-inoculation on the native soil bacterial community.

MATERIALS AND METHODS
Two experiments were carried out.The first was assembled in the germination chamber (experiment 1) and, the second, was carried out in a greenhouse (experiment 2) at the Universidade Federal do Agreste de Pernambuco,Brazil (8°54'23.7"S 36°29'39.7"W).Were used commercial seeds of B. decumbens cv.Basilisk and 24 bacteria strains with different PGPB properties previously evaluated by Oliveira et al. (2018) and summarized in Supplementary Material -Table SI.The bacterial strains belonged to Microbial Genetics and Biotechnology Laboratory collection, located at Universidade Federal do Agreste de Pernambuco, Brazil.Twelve bacterial strains of each plant species were selected (B.decumbens Stapf and B. humidicola (Rendle.)Schweickerdt), being six from root (endophytic) and six from the rhizosphere (Table SI).
The experiment 1 was performed in a completely randomized design with 25 treatments, 24 inoculation and a control treatment (without inoculum), with four replicates, contained 50 seeds each.To inoculum were prepared from pure colonies, which were incubated in the TSB 10% (Trypticase Soy Broth) (1.7 g L -1 of tryptonepancreatic digest of casein; 0.3 g L -1 of soytone -peptic digest of soybean; 0.25 g L -1 of glucose; 0.5 g L -1 of NaCl; 0.25 g L -1 of Na 2 HPO 4 ; and pH = 7.3), this culture media was supplemented with 0.05% of tryptophan.
The cultures were diluted in a phosphate buffered saline solution (PBS) (8.0 g L -1 of NaCl.200 mg L -1 of KCl; 1.44 g L -1 of Na 2 HPO 4 ; 240 mg L -1 of KH 2 PO 4 ; and pH = 7.4) and the optical density (OD) was adjusted in a spectrophotometer at 630 nm, corresponding to 10 6 CFU mL -1 .The seeds were disinfected using NaCl solution (1%) for 5 min, washed in distilled water and immersed in the inoculum for 30 min under gentle shaking.
The seeds were placed on a Germitest paper substrate moistened with distilled water 2.5 times its weight and maintained in a germination chamber at 25±5° C under a 12 h photoperiod, for 21 days.The germinated seed count was performed daily, from the seventh day after sowing, until the end of the experimental period.The germination rate was carried out following Brasil (2009) and Maguire (1962).
The plant-growth promotion evaluation was done at the 21° days after sowing.The parameters evaluated were: number of absorbent hairs (root hairs 3 mm in length); number of plumules per seedlings; length of primary root; distance between cotyledon and end of extended primary root; hypocotyl length; distance between the cotyledon and the base of the first extended seedling plumule, and the total length of the extended seedling.The dry mass of the seedlings was determined in the forced circulation greenhouse at 55° C for 72 h.
The experiment 2 was assembled following completely randomized design with seven treatments, being five bacterial strain (called MIX), a treatment using only mineral fertilization, and a control treatment without inoculum and mineral fertilization (Table I).The microcosm was assembled in pots containing 7.5 L of soil and sown with 15 seeds per pot, resulting in a total of 30 replicates.The bacterial MIX was composed by bacterial strains isolated from roots and rhizosphere, except for MIX 1, which was formulated according to the bacterial strain performance in the experiment 1. Fifteen coinoculated seeds were sown in each microcosm.The seed emergency was evaluated at 7 days after planting (DAP).Plant-growth characteristics were analyzed in the 21 DAP.
The bacterial cultivation, individual inoculum preparation and seed superficial disinfection were carried out in a similar way to experiment 1.The preparation of the coinoculant formulations was performed by mixing and homogenizing individual cultures according to Table I.The seeds were immersed in the coinoculant for 30 min under light agitation (50 rpm).The sowing depth was 2 cm.After seed coverage, 50 mL of co-inoculant was applied to the soil.
For emergency and initial growth of seedlings, the first count of emerged seedlings occurred on the 7 DAP, considering as emergence the seedlings that exposed the first plumule.The speed index was determined as proposed by Maguire (1962), and on the 21 DAP, the percentage of emerged seedlings was calculated, accounting the number of seedlings per pot, based on the number of seeds deposited per pot.Finally, we counted the number of plumules per seedlings; seedling height using graduated ruler, considering the distance between the soil to the tip of the highest plumule not extended; and the pseudoculm diameter at 5 cm from the ground.
The leaf area of the primary plumules was estimated according to Bianco et al. (2000), considering the dimensional parameters, width and plumule length.The green plumules intensity was evaluated using the portable meter SPAD-502 (Soil Plant Analysis Development), chlorophyll meter, measuring the primary plumules in the intermediate portion.The dry mass of the seedlings was made by cutting at the base of the plants, weighing and drying the plant material in the greenhouse at 55° C for 72 h.
The rhizospheric bacterial community was evaluated at 21 days after planting.For this, soil rhizospheric samples were collected from each plot.Then, every 10 plots, the samples were homogenized to form a composite sample, which totaled 3 composite samples per treatment.Three soil samples were collected before the experiment was implemented (initial control).The soil DNA total was extracted with Power Soil DNA kit (MoBio; USA), followed by the integrity evaluation by 1.2% agarose gel electrophoresis in TAE 1x buffer (40 mM of Tris-acetate; 1 MM of EDTA).
The 16S rDNA gene was amplified using the primers set 027F (5'-GAGAGTTTGATCCTGGCTCAG-3') and 1387R (5'-CGGTGTGTACAAGGCCCGGAACG-3'), and the amplification reaction follow the Heuer et al. (1997) recommendations.The PCR product was evaluated in 1.2% agarose gel electrophoresis in TAE 1x buffer and stained with blue green loading.For the restriction fragment length polymorphism analysis (RFLP), the PCR products of the 16s rDNA gene were digested with restriction enzymes HindIII, HaeIII and MboI.For each enzyme, individual blends were prepared containing 7 μL of the PCR product, 10 μL of the specific buffer for each enzyme; 2 U of restriction enzyme and 2.7 μL ultrapure water.The digestions were performed at 37° C for 10 h.The digestion products were separated by 2.5% agarose gel electrophoresis in TAE 1x buffer and stained with blue green loading.
Statistical analysis, in experiment 1, the differences between groups were evaluated by orthogonal contrast using the t-test at 5% of probability.Subsequently, the averages of all treatments were compared in relation to the control by the Dunnett test at 5%, the treatments that stood out were compared for each variable by the Tukey test at 5%.In the experiment 2, the differences between groups were analyzed by orthogonal contrast using the 5% t test, followed by the Tukey test at 5% between treatments, using the statistical software SISVAR ® version 5.7 (Ferreira 2007).The groups of the rhizospheric bacterial community was evaluated by Principal Coordinates Analysis (PCoA) and the significance between groups was tested by the ANOSIM test, using the using the Bray-Curtis similarity matrix (Ramette 2007).Both analyses were performed in the statistical software PAST ® version 4.0 (Hammer et al. 2001).

RESULTS
The bacteria inoculation promoted a significant increase in germination and seedling growth of B. decumbens, showing the largest increases in the vigor and germination speed index, as well as number of absorbent hairs, primary root length, hypocotyl and total seedling, confirming our initial hypotheses (Table II).
In general, the inoculation of Brachiaria seeds increased the germination vigor (GV) and germination speed index (GSI) by 61 and 40% in relation to the control, respectively (Table III).For these two variables, only 5 bacterial strains performed better than the control treatment, named UAGC10, UAGC71, UAGC150, UAGC 154 and UAGC167 (Table IV) by the Dunnett test at 5% probability.
Also, the inoculations with Klebsiella (UAGB 154) and Rhizobium (UAGB 167), giving the largest increases as compared to the control, with 108% and 80%, respectively.For number of absorbent hairs (NAH) and plumules (NPL), only 29% (7 strains) of the bacterial inoculum differed from the control (Table IV).The inoculants increasing the NAH and NPL at 144 and 82% (Table III).The Klebsiella (UAGB 156) strain promoted an increase of 300% in the NAH when compared to control.
For the primary root length (PRL), only three bacterial inoculants differed statistically from the control (Table IV), with the average of all inoculants providing an increase of PRL at 6% (Table III).The largest increases compared to the control were observed with the Klebsiella (UAGB 154) strain at primary root length and total seedling length, with increases of 39% and 129%, respectively, and with Sinomonas (UAGB 71) in the hypocotyl length, with an increase of 134% (Table V).

Groups
Under greenhouse conditions, the coinoculation group promoted increase as compared to the control and were inferior to chemical fertilization treatment.Among the bacterial MIXs, the inoculated Brachiaria species promoted greater emergence and initial plant development, leaf area of primary plumule, the green intensity of primary plumule and dry mass of seedling aerial part (Table VI).MIX 1, although the strains stand out under germination chamber conditions, did not promote increments in greenhouse conditions compared to other MIX treatments (Table VI).-% -------------------------------cm --------g  Similar to observed under germination chamber conditions, the MIX 3, bacteria isolated from B. decumbens, presented the best results, resembling the mineral fertilization and significantly superior to the control, except for the percentage of emerged seedlings (Table VII).
Regarding the rhizosphere bacteria community, co-inoculations and chemical fertilization promoted changes in the bacterial community structure (Figure 1).The MIX 2, 3 and 4 showed overlapping, demonstrated by the ANOSIM analysis a greater similarity of the communities among the inoculated soils (R<0.111)(Table VIII).It should be noted that only the germination of the seeds provided alteration in the bacterial community.the root system, with improvements in the development and architecture of the root system, besides the increase of germination and emergence (Araújo et al. 2012, Mia et al. 2012).
Plants infected by growth promoting bacteria that produce auxin-like compounds increase the water and nutrient uptake capacity and, consequently, can potentiate its development and the chances of establishing the crop (Jochum et al. 2019).However, plant growth may not interfere with dry mass, due to the action of this phytohormone on cell stretching and vacuolar turgor (Conceição et al. 2008), as observed in this study under conditions of germination chamber and greenhouse.
The highest increases in the germination were observed by strains Klebsiella (UAGB 60, UAGB 156 and UAGB 154), Rhizobium (UAGB 167) and Sinomonas (UAGB 71).The latter three showed in vitro indol auxin-like compounds production higher than 100.08 μg mL -1 (Table SI).Possibly, favoring the variables increased vigor and germination speed index, besides the characteristics of primary root length, hypocotyl and total length of seedlings.In the other treatments, the production of indol acetic acid varied from 4.98 to 67.17 μg mL -1 (Table SI), probably the interaction of plant microorganism assumed a fundamental role in the positive biological results of this interaction.
Among the co-inoculations, MIX 1, formed by the five best strains under germination chamber, did not favor the emergence of the initial seedling's development in a greenhouse environment (Table VI and VII).The low performance of bacterial inoculants has been observed in several studies in greenhouse and field conditions.This is due to the lower competitiveness of the bacterial strains that make up the inoculant compared to the soil native microbial community (Ramakrishna et al. 2019, Rilling et al. 2019), the physical and chemical conditions of the soil or even the efficiency of application of the inoculant (Santos et al. 2019, Souza et al. 2015).
Although almost all plant tissues are capable of producing low levels of indole acetic acid, apical meristems of stems and young leaves are the main synthesis sites of these phytohormones.In leaf primordia, the auxin accumulates at the apex, as the leaves develop, accumulations of these phytohormones can be detected on the leaf margins (Taiz et al. 2017).Thus, bacteria that produce auxin-like compounds may exert a positive effect on the dimensional parameters of the leaves (Mutai et al. 2017).
Because nitrogen is one of the constituents of chlorophyll, the content of this pigment can be used as an indicator of the nitrogen level in the leaves (Ghimire et al. 2017, Pedreira et al. 2017).Plant-associated diazotrophic bacteria can supply nitrogen more effectively, raising their levels and consequently raising chlorophyll levels (Kelemu et al. 2011, Sammauria et al. 2020).Similar results were observed by Chauhan et al. (2013) to co-inoculate different bacterial strains in sugarcane (Saccharum officinarum L.).
The productivity increase in plants inoculated with diazotrophic bacteria possibly occurred due to the increase of the available nitrogen (Sammauria et al. 2020, Shahverdi et al. 2014).This nutrient supplied through biological fixation, besides stimulating plant production, facilitates the production of phytohormones by bacteria, which in turn favor the development of roots and consequently the absorption of nitrogen, phosphorus and other nutrient indispensable to plant development (Lima et al. 2018, Tripathi et al. 2013).
According to Salamone et al. ( 2012), the inoculation of plant-growth-promoting bacteria alters the indigenous community of the soil, favoring the heterogeneity of the bacterial community (Trabelsi et al. 2011), as observed in this study.The MIXs 2, 3 and 4 showed overlapping, demonstrated by the ANOSIM analysis a greater similarity of the communities among the inoculated soils (R<0.111)(Table VIII).The ANOSIM analysis observes significant differences based on algorithms of mean distances between groups generating R correlation.Values R>0.75 indicates that the groups are well separated, for groups 0.25>R<0.75,but overlapping groups exist, and R<0.25 values there is no separation between the groups (Clarke & Gorley 2001, Ramette 2007).
Rare are the studies evaluating the inoculation of different strains of bacteria with the potential to promote plant growth in B. decumbens seeds, evidencing the importance of the present study, with unpublished and quite promising results.In both inoculation and co-inoculation of plant growth promoting, bacteria in seeds of B. decumbens cv.Basilisk promoted higher germination, emergence and initial development of seedlings, having of B. decumbens homologous bacteria and with a high yield of indole acetic acid providing higher growths in the vegetable.
The co-inoculation and chemical fertilization processes alter the bacterial community in the soil.However, it is important to emphasize the importance of studies that evaluate the entire life cycle of the plant, as well as the stress caused by cutting the plant, and/or grazing.Assuming that the use of inoculums and coinoculants of plant growth promoting bacteria in forage programs makes it possible to glimpse the reduction of production costs, as well as the increase of pasture productivity.

CONCLUSIONS
The inoculation and co-inoculation of plant growth promoting bacteria in seeds of B. decumbens cv.Basilisk provided an increase in the vigor of germinated seeds, germination speed index and the promotion of seedling growth at 21 days.The bacteria isolated from B. decumbens with high production of indole acetic acid providing greater increments to the development and growth of seedlings.The processes of co-inoculation and mineral fertilization change the bacterial community in the soil.Thus, we provided evidence that co-inoculation promotes an increase in seed performance in grass pasture in controlled conditions.
The highest challenge for the use of bacterial inoculants in pastures is: 1) lack of consistent results on the use of inoculants formulated with PGBPs in field conditions; 2) absence of adequate inoculation technologies for pasture areas; 3) absence of monitoring of strains bacterial which compose the inoculant to verify their survival and permanence over time and space variations.In front of the above-mentioned challenges, our research can be a promising strategy for the management and recovery of degraded pastures through the selection of PGPBs efficient in field trials and subsequent formulation of a bioproduct (inoculant).

Figure 1 .
Figure 1.Principal coordinates analysis (PCoA) of the rhizospheric bacterial community seedlings of Brachiaria decumbens cv.Basilisk at 21 days after co-inoculation PGPB.MIX 1: Co-inoculant formulated with the five best bacteria strains in the germination and growth promotion of B. decumbens cv.Basilisk test under germinating chamber conditions; MIX 2 and 4: Co-inoculant formulated with bacterial strains isolated from B. humidicola (Rendle.)Schweickerdt; MIX 3 and 5: Coinoculant formulated with bacterial strains isolated in B. decumbens Stapf.MF: Mineral fertilization; CON: Control, and SZ: Soil zero, before the experimental period.

Table I .
Bacterial isolates for formulation of co-inoculant (MIX) used on seeds Brachiaria decumbens cv.Basilisk under greenhouse conditions.

Table III .
Performance of bacterial strains in the germination test in relation to control (without inoculant).

seedlings -1
MBI: mean of bacterial isolates; MC: mean of control; BSP: bacterial strains performance -percentage increase of MBI in relation to MC; GV: Germination vigor; SGI: Speed germination index; NAH: Number of absorbent hairs; NPL: Number of plumules; PRL: Primary root length; HYL: Hypocotyl length; TLS: Total length of seedling; DMS: Dry matter of seedlings.

Table IV .
Germination characteristics and initial growth of Brachiaria decumbens cv.Basilisk seedlings after seed inoculation of potentially plant growth promoting bacteria, with 21 days in germination chamber at 25±5° C under 12 h photoperiod.

Table V .
Germination characteristics and initial growth of Brachiaria decumbens cv.Basilisk seedlings after seed inoculation of potentially plant-growth promoting bacteria, with 21 days in germination chamber at 25± 5° under 12 h photoperiod of the treatment's superior to the control, pre-selected by the Dunnett test.

Table VI .
Comparison between groups of means by orthogonal contrasts for emergence characteristics and promotion of seedling growth at 21 days in the greenhouse after co-inoculation (MIX) of potentially plant-growth promoting bacteria in seeds Brachiaria decumbens cv.Basilisk.

Table VII .
Emergence characteristics and promotion of seedling growth at 21 days of greenhouse cultivation after co-inoculation (MIX) of plant-growth promoting bacteria in Brachiaria decumbens cv.Basilisk seeds.ESI: Emergency speed index; PSE: Percentage of seedlings emerged; NPL: Number of plumules; SEH: Seedling height; PSD: Pseudoculm diameter; LAP: Leaf area of the primary plumule; SPAD: Measurement of the green intensity of primary plumule; SSDM: Seedling shoot dry mass; MIX 1: Co-inoculant formulated with the five best bacteria strains used in the germination and growth promotion of B. decumbens cv.Basilisk test under germinating chamber conditions; MIX 2 and 4: Co-inoculant formulated with bacterial strains isolated from B. humidicola (Rendle.)Schweickerdt; MIX 3 and 5: Co-inoculant formulated with bacterial strains isolated in B. decumbens Stapf and MF: Mineral fertilization.Means followed by the same letter in the column do not differ from each other by the Tukey test at 5% probability.

Table VIII .
Analysis of similarity (ANOSIM) of the rhizosphere bacterial groups presents on the seedlings Brachiaria decumbens cv.Basilisk at 21 days after co-inoculation with plant-growth promoting bacteria.Co-inoculum formulated with the five best bacteria strains in the germination and growth promotion of B. decumbens cv.Basilisk test under germinating chamber conditions; MIX 2 and 4: Co-inoculant formulated with bacterial strains isolated from B. humidicola (Rendle.)Schweickerdt; MIX 3 and 5: Co-inoculum formulated with bacterial strains isolated in B. decumbens Stapf; MF: Mineral fertilization; CON: Control and SZ: Soil zero, before the experimental period.