Viability of liquid medium-inoculation of Rhizobium etli in planting furrows with common bean

The objective of this work was to evaluate the viability of liquid medium inoculation of Rhizobium etli in the planting furrow and to certify the efficiency of its strain UFLA 02-100 as a potential inoculant for common bean (Phaseolus vulgaris). The treatments consisted of three application rates of liquid inoculant applied in the planting furrows and one rate applied in common bean seed, besides two controls without inoculation. The inoculant contained 109 cells of Rhizobium etli mL-1. Regardless of the application method, the yield obtained with the inoculation was equivalent to that of N from urea; however, the application in the furrows, at 0.6 L ha-1, is more advantageous due to the operational practicality and reduced costs.

In recent years, numerous studies indicate that common bean (Phaseolus vulgaris L.) can truly benefit from biological nitrogen fixation (BNF) under field conditions, as long as good inoculation practices are performed.Nonetheless, questions regarding the conventional use of peat inoculant in seed still persist, and the effects of mineral N fertilization on BNF are still being assessed.Current results with peat inoculant on seed have indicated that the application of 20 kg ha -1 N, at planting, favors the symbiosis, which increases yield and reduces costs (Soares et al., 2016).In this sense, other studies have shown that common bean inoculation, complemented by leaf fertilization with Mo, increases the cost-effectiveness of the crop (Valadão et al., 2009;Figueiredo et al., 2016).
In tropical soils, where acidity conditions are more accentuated, the use of adapted strains is necessary to maximize the contribution of BNF to crops.Researches on peat inoculation have shown a high performance of the R. tropici strains CIAT 899 and H 12, and also of the R. freirei strain PRF 81.These strains are already recommended for use; nonetheless other strains, such as UFLA 02-127 (R. leguminosarum bv.phaseoli), UFLA 02-68 (R. etli bv.mimosa), UFLA 04-173 (R. miluonense), and UFLA 02-100 (R. etli), have shown a high competitive ability, and presented an equivalent (Ferreira et al., 2009) and even greater effectiveness than that of the native rhizobia population (Soares et al., 2006;Nogueira et al., 2017).Actually, field experiments with this last strain (UFLA 02-100) show similar results to that of CIAT 899 (frequently used in commercial inoculants), and to those of control treatments fertilized with mineral N (Soares et al., 2006;Nogueira et al., 2017).All these studies show the possibility of economic savings in relation to N fertilizers.
Despite the efficiency of peat inoculation in seed, there are many producers that question the practicality of this operation in large scale, especially with the current trend of increasing the amounts and types of products applied to seed, such as fungicides, micronutrients, and microorganisms.For this reason, soybean producers and large producers of common bean have shown interest in the use of liquid inoculant in the planting furrow.
The possible incompatibility between products used in seed treatment (mainly of fungicides) and rhizobium survival from the inoculant is a matter of great concern for these producers.The pertinent literature, however, shows some disagreement on this issue (Araújo & Araújo, 2006;Oliveira et al., 2016).In cases of incompatibility between products, the producer should treat seed by himself, since it would be risky to acquire it already treated.Seed inoculation with peat or liquid inoculant subjected to drying should be undertaken in the shade, with an interval for proper adherence, since rhizobium is greatly exposed to the incompatible product, strengthening the possibility of deleterious effect of the practice (Araújo & Araújo, 2006;Araújo et al., 2007).In fact, it has been shown that under liquid inoculant and Mo applications to seed, the number of Bradyrhizobium cells can be significantly reduced (Albino & Campo, 2001), with negative effects on BNF.Besides, planting with moist or wrinkled seed can cause difficulties to the sowing operation, which requires an adequate time for drying the inoculated seed.
One of the limitations of liquid application to the planting furrows is related to the small number of research studies involving this practice, and to the difficulties in determining the best application rates for good performance of BNF (Vieira Neto et al., 2008a, 2008b;Zilli et al., 2010).Publications involving this alternative method of inoculation for common bean could not be found; therefore, the present work can help the producer in deciding on which method to adopt for this crop.
The objective of this work was to evaluate the viability of liquid medium inoculation in planting furrows, and to certify the efficiency of the UFLA 02-100 strain of Rhizobium etli as a potential inoculant for common bean.
The field experiment was carried out in the dry season of 2014, in the municipality of Lavras, in the state of Minas Gerais, Brazil, in a Latossolo Vermelho eutroférrico (Santos et al., 2013), i.e., Eutric Acrudox, according to the soil taxonomy.Results of the chemical and physical analyses of the soil, before planting, are shown in Table 1.A randomized complete block design with six treatments was used, with four replicates.The treatments involved three application rates of liquid inoculant in planting furrows, and one rate applied in seed, and two controls without inoculation.Crop management was performed under no-till, over corn residues, with periodic sprinkler irrigation.The population of native rhizobium capable of nodulating the crop was of approximately 10 3 colony forming units (CFU) per gram of soil, estimated according to Rufini et al. (2011).
Each experimental unit consisted of six rows of 4.0 m length each, spaced at 0.6 m, with an useful area corresponding to the four center rows.All plots received 20 kg ha -1 N (urea), 110 kg ha -1 P 2 O 5 , and 40 kg ha -1 K 2 O.In addition, N-fertilized control plots received another 20 kg ha -1 N, at planting, and 40 kg ha -1 as topdressing, between the V3 and V4 stages.Planting was performed with a "handjab planter" at density of 15 seed per linear meter, using the common bean BRSMG Madrepérola cultivar.
The strain UFLA 02-100 was isolated from acid soils from the Amazon region, and selected for its high efficiency in common bean BNF, with a performance comparable to that of strains frequently used in commercial inoculants for this crop (Soares et al., 2006;Ferreira et al., 2009;Nogueira et al., 2017).The inoculant was prepared in liquid medium "79", and its quality was monitored by counting the number of CFUs, that should be according to the legal minimum number of viable Rhizobium cells, which is approximately 10 9 mL -1 inoculant (Brasil, 2011).The inoculant rate applied in seed was equivalent to 7.5 mL kg -1 , and inoculation occurred a short time Pesq.agropec.bras., Brasília, v.53, n.3, p.394-398, Mar.2018 DOI: 10.1590/S0100-204X2018000300015 before planting.The application rates in the furrows were determined to obtain populations equivalent to ½, 1, and 2 times the reference application rate of 1.2 L ha -1 commonly recommended for commercial inoculants.The inoculant distribution in the furrows was performed with a manual backpack sprayer, with a flat fan spray nozzle XR 110.02, and spray volume equivalent to 20 L ha -1 .
At full blossom (R6 stage), a sample of 10 plants was removed at random from rows 2 and 3 of each plot, for the determination of the number of nodules (NN), dry matter of nodules (DMN), shoot dry matter (SDM), shoot-N concentration (SNC), and shoot-N accumulation (SNA).At maturation (R9 stage), final stand (FS), grain yield and its primary components (number of pods per plant, grains per pod, and mean 100 grain weight, W100), and grain-N concentration (GNC) and accumulation (GNA) were determined in 10 plant samples from the rows 4 and 5.The final stand was obtained by counting the number of plants in the useful area of the plot.Grain yield was obtained from total weight of grains produced in the useful area of the plot (lines 4 and 5), with moisture corrected to 130 g kg -1 .The SNC, GNC, SNA, and GNA were calculated according to Soares et al. (2006).
Data were subjected to analysis of variance, using the software Sisvar 4.0, after meeting the presupositions of normality (Shapiro-Wilks test) and homoscedasticity (Bartlett test).NN and DMN were transformed into (x+1) 0.5 prior to the analyses.The means were grouped by the Scott-Knott test, at 5% probability.
Neither nodulation nor BNF were affected by the inoculation method, which can be inferred by the SNC and SNA variables (Table 2).In fact, SNC, which is one of the traits closely related to BNF efficiency, was greater than 2.8%, within the sufficiency range indicated by Oliveira et al. (2016).Moreover, the effectiveness of the UFLA 02-100 strain on GNA, W100, and grain yield, whose values were similar to those of mineral fertilization with 80 kg ha -1 N, is noteworthy.The higher values of these variables compared to those of the control without additional N indicate a greater use of N for nutrition and grain filling, which resulted in yields of 1,600 kg ha -1 , which represents twice the mean value of the absolute control.
These results clearly show the important contribution of the UFLA 02-100 strain to common bean and support the considerations of high efficiency of bacteria in relation to the nitrogen fertilizer control reported by other authors (Soares et al., 2006;Nogueira et al., 2017).They also justify the continuity of studies with this strain, which has a considerable potential for future approval as an inoculant strain.
Research studies with liquid medium inoculation in the planting furrows for common bean were not found, which made it impossible to compare our results.This fact, however, reinforces the relevance of the present study.In soybean, liquid medium inoculation in the furrow has been reported not only as practical, but also as compatible with fungicide treatment in seed (Zilli et al., 2010).Potential advantages of application of liquid inoculant in the planting furrows had also been reported for soybean grown in native field (Voss, 2002).
Based on the effect of inoculation in the furrows on the studied variables, mainly on grain yield (Table 2), our results for common bean are even more encouraging due to the fact that low-application rates of the inoculant were already sufficient to bring about yields much higher than the mean productivity of the crop (932 kg ha -1 ) in the state of Minas Gerais, Brazil.Application rates greater than 0.6 L ha -1 in the furrows provided a similar yield to that obtained with the inoculation in the seed, but with higher costs.Using a commercial inoculant for soybean, at 150, 300, and 450 mL ha -1 , Vieira Neto et al. (2008a, 2008b) obtained similar yields to that with inoculation in the seeds.The authors recognized the viability of furrow applications, even at the lowest rate.
Further studies should evaluate commercial inoculant strains already available in the market, and application rates appropriate for the best productive performance.Applications of these strains should also be compared with that of UFLA 02-100 in the planting furrows.This information will surely assist the producer in choosing the most profitable method and rhizobium strain for growing common bean.

Table 1 .
Results of the chemical and physical analyses of soil samples from the experimental area (0-0.20 m soil depths), collected before planting.