The plant material used in this study consists of seeds of Glycine max L. cv. Docko, sensitive to NaCl [18] , from the IRAD of Wakwa in the Adamawa region. The choice of products (organic and inorganic) is based on their low cost and accessibility, as well as their ability to provide nitrogen and carbon for rhizobial growth [19] . Thus, the materials used for the formulation of biofertilizers are composed of by-products of the agro-food industry (digestate), shell powder, and yeast mannitol broth (YMB) media; the different associations of these formulations made it possible to classify them into three groups ( Table 1 ). The coating material for soybeans is palm kernel oil and maize powder. The BOSD rhizobial consortium, provided by the Soil Microbiology Laboratory of the Biotechnology Centre of Yaoundé I, consists of 4 strains isolated in the soybean rhizosphere, namely Bosd1, Bosd2, Bosd4 and Bosd6 [20] .

The physicochemical parameters are determined on the one hand on the solid supports used: biochar, charcoal, coating substrate, white powder, soil sample from the Garoua field (production site of the Docko variety), Cajun weight and cow dung, and on the other hand on liquid supports: digestate and sugarcane molasses at the Laboratory of the Soil Analysis and Environmental Chemistry Research Unit (URASCE) of the University of Dschang using the methods of [21] .

The biofertilizer formulations based on the combinations of these different media mentioned above are listed in Table 1 . The Yeast Mannitol Broth (YMB) medium noted B0 is prepared according to the standard method of [22] . As for the formulations noted B1 and B2, the solid materials are ground, passed through an 80 µm diameter sieve, and mixed in the proportions indicated in Table 1 for a final mass of 100 g autoclaved at 121˚C for 30 min; each of these two formulations is added to 40 ml of consortium inoculum [23] were the standard YMB medium (yeast mannitol brother) was prepared from the protocol developed by [22] ; Subsequently, four colonies of each isolate (bosd1, bosd2, bosd4 and bosd6) are added to this medium. Formulation B3 is made by the modified method of [24] ; thus, approximately 2 ml of sugarcane molasses is added to 98 ml of pigeon pea extract. Formula B4 is obtained by diluting 80 ml of digested with 20 ml of distilled water. Formulations B5 and B6 are B3 and B4 respectively with the addition of 1 g of white powder in each of them. The pH of the formulations, namely B3, B4, B5, and B6, are adjusted to 6.8; they are also autoclaved at 121˚C for 30 minutes and finally, four colonies are inoculated. All these formulations are divided into three groups ( Table 1 ) and stored at room temperature during the evaluation of the growth and viability of the rhizobia.

The consortium’s growth is evaluated four days after inoculation for 56 days, followed by its monthly viability using the counting method after decimal dilution described by [25] . The individualized colonies are evaluated by manual counting after 24 h of incubation, and the bacterial load (CN) is obtained by calculating the number of rhizobia per ml of solution and expressed in log CFU/ml.

CN (log CFU/ml) = Number of colonies counted × Dilution factor

The volume of inoculum deposited

The viability of the consortium in the different formulations is evaluated after the growth of the bacteria every 7 days (7th, 14th, 21st, 28th, 35th, 42nd, 49th, and 56th day) over 2 months [26] .

The consortium development assessment presented in Table 2 is carried out according to the modified method of [24] described by [27] .

Yeast extract is prepared using sterile distilled water and the solution obtained is brought to the boil. As for the soybean, Cajun pea and bean extracts, approximately 100 g of power of each material are mixed with 1000 ml of water and mixed. The mixture is then boiled for two hours, cooled and autoclaved at 121˚C for 30 minutes.

From the previous Petri dishes, four colonies of each strain are taken and inoculated into the different growth media ( Table 2 ).

Optical densities are read using a UV-visible spectrophotometer (Jenway 6305) at 600 nm every 24 hours for 4 days, at the end of which the media are centrifuged [26] . The pellet obtained after centrifugation is seeded in Petri dishes containing Yeast Extract Mannitol Agar (YEMA) medium to determine the fresh biomass (mg/ml) of the consortium used.

Each previously prepared nitrogen-fixing rhizobia biofertilizer formulation is used for coating apparently healthy and morphologically similar soybeans with palm kernel oil as an adhesive for one block and corn powder for the other by the effluent method. This method consists of mixing 0.05 g (solid) or 0.05 ml (liquid) of each biofertilizer formulation with 0.15 ml of palm kernel oil or 0.15 g of maize powder, then adding 5 g of soya beans. The mixture is transferred to a container, stirred carefully and spread on a clean surface in the shade between 26 ± 2˚C, 55% and 64% humidity. These coated seeds are finally sown in the field 24 hours later.

The seeds coated with these different formulations were sown in triplets on 7 kg of substrate consisting of a mixture of black earth and sand with a diameter of 0.7⁻¹ mm in the ratios (3 v/v) contained in 5 L culture pots. The pots were arranged in two completely randomized factorial blocks corresponding to the two coating adhesives (palm kernel oil and corn powder), and each block consists of 10 treatments corresponding to the different formulations with six repetitions each. The treatments carried out were: T0 (negative control), T1 (treatment with bacterial inoculum), T1 + h (bacterial inoculum + oil), T1 + pb (bacterial inoculum + white powder), T2 + h (B1 + oil), T3 + h (B2 + oil), T4 + h (B3 + oil), T5 + h (B4 + oil), T6 + h (B5 + oil) and T7 + h (B6 + oil). The same treatments were repeated with corn powder (m), i.e. 20 treatments in total. The plants were irrigated every day with tap water and once a week with this water enriched with 3 g/l of NaCl until harvest.

Growth parameters are measured during flowering [22] . Thus, the number of leaves is determined by counting, the height of the stems by direct measurement using a tape measure, the average leaf area [28] method, S = 2/3 * L *l * n * 0.80 * 0.662 (cm²) where n is the number of leaves, L is the length of the leaf, and l is the sum of the maximum widths of its leaflets. The nodules are collected and counted after harvest, and the masses of the different organs were determined using a precision balance (KERN 572) to 0.001 g.

Chlorophyll is measured in non-senescent adult leaves harvested 4 weeks after sowing by the modified [29] descried by [30] . The total nitrogen in these leaves is measured according to the method described by [31] .

The obtained effect of each result was the difference in quotient between treated and control blocks by control block multiplied by 100.

The data for each parameter has been collected and treated into SPSS version 23.0 software for analysis of variances (ANOVA) using Duncan’s test at the 1% threshold. Microsoft Excel 2016 software was used to present the results in the form of histograms and curves with means ± standard deviation.

The data recorded during the characterization of the biofertilizer formulation materials are recorded in Table 3 . The analysis of this table shows that: some materials used are acidic because of their pH between 4.82 and 6.7 while others are basic with a pH between 7.8 and 9.68. The carbon and total nitrogen contents are higher in sugarcane molasses compared to the coating substrate. The phosphorus content is more abundant in Ox dung ( Table 3 ). The major components of the biofertilizer are material supports of organic, inorganic, or synthetic nature which have appropriate physicochemical properties to maintain microbial strains in good physiological storage conditions and the field [32] ; Bacterial growth considers several factors including an appropriate pH buffer, high water holding capacity, and chemical and physical uniformity [33] . Thus, the pH of different materials used for biofertilizer formulation ranges from 6 (Garoua soil) to 9.68 (digestat), this growth pH is generally close to neutral, is known to support a large population of rhizobia and maintain their viability [34] .

The growth of the consortium in the different formulations is shown in Figure 1 . It emerges from its analysis that the loads in the different formulas increase exponentially and there is no significant difference between treatments B1, B4 (8.73 and 8.74 log CFU/ml) and the control treatment B0 (8.72 log CFU/ml). Formula B3 (8.89 log CFU/ml) has the highest load compared to the other formulas. The Bradyrhizobium strain would grow better in alternative media compared to the standard YMB medium [35] . This observation is made by an increase of + 30 to + 36% in these media over 4 days of observation. The formula enriched with Cajun pea extract supplemented with sugarcane molasses has the highest loading (8.89 log CFU/ml) compared to the standard YMB medium (8.71 log CFU/ml). Media enriched with sugarcane molasses would significantly increase the growth and viability of rhizobial strains [36] .

The viability of the rhizobial consortium in the different formulations is shown in Figure 2 . Its analysis shows that: the curves have a generally decreasing shape; this is reflected by a progressive decrease in the viability of the consortium in the different biofertilizer formulations depending on the shelf life. All formulas (B1 to B6) have loads that vary between 6.5 and 6.74 log CFU/ml compared to the control B0 (6.49 log CFU/ml) after 56 days of counting. Formula B5 (6.74 log CFU/ml) has the highest bacterial load compared to the others and formula B3 (6.5 log CFU/ml) is the lowest. Determining the lifespan of a microbial strain in a formulation is an important step in the development of biofertilizers [37] and prolonged survival of inocula during storage is a desirable effect capable of increasing the application of biofertilizer formulations in industry [38] . In addition, variations in the charges in these formulations differ depending on the nature of the latter. They would probably be due to differences in the physicochemical properties of the transport materials used [39] .

The growth of the BOSD rhizobial consortium in alternative media is shown in Figure 3 . Its analysis shows that: the supply of nitrogen sources or extracts enriched with nitrogen sources allows an increase in the bacterial load in all the consortium media during the evaluation period except soybean extract, the loads of which remain constant. Alternative media thus appear to be beneficial due to their local availability and affordable cost [27] . The growth of the consortium in the developed media is reflected by its development with a preference for the medium enriched with soybean extract as a nitrogen source and powdered milk as a carbon source compared to the yeast extract/mannitol medium, which is the standard YMB medium. Indeed, the VUID1 strain would use all the alternative sources of carbon available (sucrose, glycerol, and cow’s milk) to develop [27] . Fast-growing and intermediate-growing species of rhizobia have been reported to be able to grow in media containing sucrose as a carbon source [40] .

Values in the same column followed by the same letter are not significantly different at the 1% level.

The responses of soybean plants in the different formulations and coating methods are presented in Table 4 . Analysis of the table shows that all morphological parameters of soybean plant growth significantly increased (P ≤ 0.01) in the treatment (T4 + h) with the highest and lowest values in the treatment (T6 + m) compared to the control T0. The contribution of rhizobial strains in the soybean rhizosphere would promote hydromineral and nitrogen nutrition. This supply of water and nutrients would instantly stimulate the phenomenon of photosynthesis, thus inducing good growth and development of these crops, namely the size, number of leaves, and their leaf surface [40] . This is also the case for Cowpea (Vigna unguiculata L.) after fertilization with consortium biofertilizer (CBF) [41] .

Values in the same column followed by the same letter are not significantly different at the 1% level.

The different values of the assessment of the effect of formulations on the number, and weight of nodules as well as the effect on soybean plants are presented in Table 5 . The analysis of this table shows that the contribution of formulated biofertilizers leads to a significant increase in the number and weight of nodules regardless of the coating method ( Table 5 ). As for the effect of nodulation on soybean plants, treatments (T2 + m), (T3 + m) and (T4 + h) have more significant effects namely: +163%, +191%, and +178% respectively) then the other formulations and compared to (T1). Legumes in general and beans in particular produce a large number of nodules at the root or stem level when the appropriate symbiotic partners are provided [42] [43] . This symbiotic association is used in several legume species, thus increasing their productivity [19] [20] . The growth parameters of soybean plants all show a significant increase in formulations with palm kernel oil as a coating adhesive. However, the nodulation effects are greater on the one hand with charcoal (+191%) and on the other hand with pigeon pea powder sugarcane molasses (+178%); this result would reflect the beneficial effect of biofertilizers on plant nodulation [27] .

Values in the same column followed by the same letter are not significantly different at the 1% level.

The effect of formulations and coating methods on the total biomass of soybean plants harvested in an open environment is presented in Table 6 . Its analysis shows that: plants of different formulations have a similar biomass effect (+200%) to that of the positive control medium (T1) and +350% compared to that of plants of the negative control medium (T0). The biomass effect is more significant with the Cajun pea powder + sugarcane molasses associated with palm kernel oil (+350%). [27] also showed the beneficial effect of biofertilizers on the biomass and nodulation of peanut plants (Arachis hypogaea L.). The rhizobia-legume symbiosis at the root level is materialized by the formation of nodules, the seat of the biological fixation of nitrogen essential for the growth of the plant. Nodulation and biomass production depend on the compatibility between the plant genotype, as in the case of faba bean where the rhizobium strain and its interaction with the biophysical conditions of the soil had led to an increase of between 69 and 114% [42] ; similar responses were observed in cowpea with rhizobia strains [20] .

The total chlorophyll contents determined 4 weeks after sowing are presented in Figure 4 . The analysis of this shows that: the T4 + h treatment has a high total chlorophyll content (125) with an effect of (+11%) compared to the control (T0), while the T4 + m treatment represents the lowest content (113) compared to T0. However, in each treatment, the leaves of soybean plants whose seeds had been coated with palm kernel oil have the highest total chlorophyll content compared to those of plants whose seeds had been coated with corn powder, highlighting the importance of stickiness on the sowing quality of soybeans. The palm kernel oil used causes sufficient humidification of the seeds, reducing physiological accidents and improving the germination quality and good growth of the plant.

The total nitrogen (N) contents determined after harvest in soybean leaves are reported in Table 7 . It emerges from his analysis that the contribution of biofertilizers leads to a significant increase (P ≤ 0.01) in the nitrogen content with an effect of +861% for the treatment (T4 + h) compared to the positive control (T1) and +1102% compared to the negative control (T0). The total nitrogen abundance is observed at this level of the treatment added to palm kernel oil. The biological fixation of nitrogen by bean, soybean, and cowpea plants has a positive effect on the nitrogen content compared to the control according to the work of [44] . These beneficial microorganisms are capable of fertilizing the soil enriched with all micro and macronutrient elements via nitrogen fixation, solubilization or mineralization of phosphorus and potassium, plant growth regulation factors, and degradation of organic matter in the soil [45] . The same is true for the protein content which presents a significant (P ≤ 0.01) high value in the treatment (T4 + h), namely 195.31 which presents effects of +1165% and +935% respectively compared to T1 and T0.

Values in the same column followed by the same letter are not significantly different at the 1% level.