Combined inoculation of rhizobia on the cowpea development in the soil of Cerrado 1

Biological nitrogen fixation (BNF) plays an important role in the cowpea cultivation. This study aimed to assess the foliar levels of chlorophyll and the yield components of cowpea subjected to combined inoculation of rhizobia in Cerrado soil. The experiment was carried out in field conditions, from December of 2014 to the march of 2015, in Rondonópolis, Mato Grosso, using randomized block design with ten treatments and three replicates. Were tested the single strains MT8 and MT15 (both of R. tropici), MT16 (R. leguminosarum), BR3267 (Bradyrhizobium japonicum), the strains combinations MT8+MT15, MT8+MT16, MT15+MT16, and MT8+MT15+MT16, nitrogen fertilization (70 kg ha-1 of N-urea), and absolute control (without inoculation of rhizobia and without nitrogen fertilization). Data were subjected to analysis of variance and the averages were compared by orthogonal contrast and F test (p≤0.05). Were assessed the Falker chlorophyll index (at the 40 and 60 days after sowing), number and dry matter of nodules (at the 40 days after sowing), number of pods per plant, grains yield, the concentration and accumulation of nitrogen in grains, and crude protein. The number of pods per plant was increased 33.6% in the combination MT8+MT15. The same effect was observed for the yield grain, which presented an increase of 8.7%, 13.8%, and 16.7% in the combinations MT8+MT16, MT15+MT16, and MT8+MT15, respectively. The nitrogen accumulation in the grains increased 42.7% with the inoculation of MT15 strain. The cowpea responds positively to the usage of combinations of rhizobia strains.


INTRODUCTION
In Brazil, cowpea is cultivated in an area of 1.3 million ha in North, Northeast, and Central-West regions, with a total production of 482,000 tons yr -1 (RUFINI et al., 2014).This crop plays an important role in the feed of the most of the population at those regions, supplying the daily demand of vitamins, minerals, and proteins (FRIGO et al., 2014).
Particularly, the biological nitrogen fixation (BNF) can contribute to the nutrition of cowpea, once this process replaces up to 80 kg ha -1 of N, increasing remarkably the cowpea yield in the Brazilian agroecosystems (MELO;ZILLI, 2009;RUFINI et al., 2014).
Alternatively, a strategy for increasing the efficiency of inoculants recommended to cowpea is the use of combinations of rhizobia strains.Owing to little information existent about this technique for cowpea crop, studies have been necessary for understanding the contribution of combined rhizobia strains in the same formulation.Traditionally, are well-known studies involving the combination of rhizobia strains with mycorrhizal fungi (LIMA et al., 2011;OMIROUA;FASOULAB;IOANNIDES, 2016), or rhizobia strains with plant growth promoting bacteria (PGPB) (ARAÚJO et al., 2010;RODRIGUES et al., 2013a).
According to Trabelsi and Mhamdi (2013), the use of combined rhizobia inoculation (two or more strains) can be beneficial to the establishment of bacteria present in the inoculant at the soil in relation to native rhizobia population, and it also increases the efficiency of BNF, avoiding the technical problems of the single strains inoculation.In addition, nitrogen participates of chlorophyll structure and correlates positively with nitrogen content in the leaves of cowpea (MORAIS;FONTES;GONÇALVES, 2013).
In this context, this study aimed to assess the effect of combined inoculation of rhizobia strains on the foliar chlorophyll levels and the yield components of cowpea grown in Cerrado soil.

Experimental location and soil characteristics
The experiment was carried out from December 2014 to March 2015, at the experimental site of Federal University of Mato Grosso, Rondonopolis Campus, located at latitude 16º27'44" S and longitude 54º34'48" W, 290 m.According to Köppen classification (ALVARES et al., 2013), the local climate is type Aw (humid tropical, with a temperature average of 25.6 °C, rainy summer with average rainfall amongst 1,400 to 1,500 mm yr -1 , and dry winter).The climate data during the experimental period are shown in Figure 1.

Experimental design and treatments
The experimental design used was in randomized block, with ten treatments and three replicates.The treatments consisted of inoculations with three strains isolated from cowpea (MT8 and MT15 of Rhizobium tropici, and MT16 of R. leguminosarum), a recommended strain for cowpea (BR3267, of Bradyrhizobium japonicum), four combinations (MT8+MT15, MT8+MT16, MT15+MT16, and MT8+MT15+MT16), nitrogen fertilization (70 kg N ha -1 , using urea as source), and absolute control (without rhizobia inoculation and without N-urea application).

Inoculant preparation
The strains were cultivated using the YMA solid medium (in 1000 mL of distilled water: 10 g of mannitol, 0.5 g of K 2 HPO 4 , 0.2 g of MgSO 4 .7H 2 O, 0.1 g of NaCl, 0.5 g of yeast extract, 5 mL of 0.5% blue bromothymol in 0.2N KOH, 16 g of agar) with pH 6.8 (FRED; WAKSMAN, 1928).Initially, the strains were multiplied in Petri plates, in a bacteriological incubator (72 h, 28 °C).After this, strains were transferred to Erlenmeyer flasks containing 150 mL of YMA liquid media for obtaining of bacterial broth, which was maintained in benchtop shaker (28 °C and 80 rpm) for 24 h.The peat-based inoculant was prepared in the proportion of 35 g of peat to 10 mL of bacterial broth when strains were in log phase of growth, aiming a minimal concentration of 10 9 viable cells g -1 inoculant, which was applied to seed little before sowing (FRIGO et al., 2014).
The urea was applied in plots of the treatment with nitrogen fertilization in a single dose at 15 days after sowing.The BRS Tumucumaque cultivar was sowed in the plots after inoculation with the density of ten seeds per meter.The experimental plots they consisted of 5 m x 3 m (15 m 2 ), being spaced 0.5 m inter-row with 1.0 m far from each other, wherein were used 9 m 2 as the useful area to assess the studied parameters.

Studied parameters and determinations
Were evaluated the Falker chlorophyll index at the 40 and 60 DAS (days after sowing), number and dry matter of nodules per plant, the number of pods per plant, grain yield, nitrogen concentration and accumulation in grains, and crude protein.
For measure the Falker chlorophyll index, a portable chlorophyll meter was used, model ClorofiLOG® CFL 1030, in which the readings were performed in five leaves per plant.Two readings were performed, the first one being in flowering period (40 DAS), and the second one being in grain filling period (60 DAS), wherein five plants were sampled for each reading.
The nodule number was estimated through removal and counting of nodules present in roots of ten plants at the 40 DAS.The dry matter of nodules was estimated through the weighing of nodular material in semi-analytical weighing scale, which was previously oven-dried (72 h, 65 °C).
The number of pods per plant was estimated by counting the pods from ten chosen randomly plants.All cowpea plants were harvested to estimate the grain yield, being the grain moisture initially standardized at 13% by oven method at 105 °C, according to Brazilian rules for seed analysis (BRASIL, 2009).
The concentration of nitrogen and crude protein in grains was estimated by Kjeldahl method, according to the methodology described by Frigo et al. (2014).The nitrogen accumulation in the grains was calculated by multiplying weight of the sample by N percentage obtained in titration, divided by 1000.

Statistical analysis
In order to investigate the effects of combined inoculation of rhizobia strains on the foliar chlorophyll levels and the yield components of cowpea, the data were subjected to analysis of variance (ANOVA) using the statistical software SISVAR Version 5.3 (FERREIRA, 2011).The average of the treatment was compared by orthogonal contrasts and F test (p<0.05).The use of orthogonal contrasts makes possible a detailed analysis of experimental data, being useful to estimate the main and nested effects, as well as interaction effects in comparisons between average groups (NOGUEIRA, 2004).

RESULTS AND DISCUSSION
The results of the Falker chlorophyll index at flowering period (40 DAS) of the cowpea plants did not present significant differences in contrasts accomplished among the treatments (Table 2).
Based on the equation (Y = 17.695 + 0.3362x -0.0015x 2 , with R 2 = 0.89), proposed by Morais, Fontes and Gonçalves (2013), the observed values of nitrogen levels in the leaves of cowpea plants varied from 32.84 g kg -1 in the MT16 treatment to 33.50 g kg -1 in the BR3267 treatment.According to Crusciol et al. (2007), values between 30-50 g kg -1 indicates a satisfactory nutritional status of the plants, and shows the cowpea plants presented an adequate amount of nitrogen at the leaves.This indicates the effectiveness of the rhizobia strains, and reinforces the positive correlation between chlorophyll index (Falker) and the nitrogen content in the plant tissues as an indicator of the nutritional state of plants (MORAIS; FONTES; GONÇALVES, 2013).
In a similar study, Frigo et al. (2014) evaluated the productive characteristics of cowpea inoculated with rhizobia lineages in Red Oxisol during the flowering period, and they observed significant difference between nitrogen fertilization treatment (63.21) and BR3267 strain (57.85).According to these authors, the plant response was associated with the availability of nitrogen in the urea form, which allowed the major uptake of this nutrient.
In the present study, the high soil moisture probably contributed to more nitrogen leaching in depth, reducing the availability of this nutrient and its uptake by plants, according to climate data observed in February 2015, once was the period with most rainfall (Figure 1), corroborating the conclusions of Rufini et al. 2014, who they observed that nitrogen fertilization can be substituted by BNF, decreasing the nitrogen losses in agroecosystems.
In the second reading (60 DAS), the absolute control presented the highest value of Falker chlorophyll index,

Treatments
Average Contrast and coefficients Combined inoculation of rhizobia on the cowpea development in the soil of Cerrado when compared to others treatments (C1, p≤0.05).In this same reading, the MT15 strain also presented significant difference in relation to MT8 strain (C9, p≤0.05).In other contrasts, have not were observed significant difference amidst treatments (Table 3).
According to the equation proposed by Morais, Fontes and Gonçalves (2013), the plants of absolute control presented 33.70 g kg -1 of nitrogen in their leaves and an increase of 7.1% in the chlorophyll index, whereas the plants of MT15 treatment presented 33.49g kg -1 of nitrogen in their leaves and an increase of 4.65%, both in relation to nitrogen fertilization.This result shows that the measurement with chlorophyll meter is capable of informing the nitrogen levels in the leaves of cowpea plants.
The high chlorophyll index of absolute control plants can be related to the small of this plants in relation to plants of other treatments, once these plants were not inoculated, neither received nitrogen fertilizer.At the 60 DAS (grain filling period), decreases in the chlorophyll index are expected during grain filling period, once nitrogen is remobilized from leaves to the grains (FRIGO et al., 2014), which, in this study, led to a low nitrogen accumulation in the grains (11.86 kg ha -1 ) in these plants.
Cowpea is able to form, promiscuously, a symbiosis with soil native rhizobia, besides the  population of soil native rhizobia may vary in composition and effectiveness (ZILLI et al., 2006).In a study with cowpea inoculation, Rufini et al. (2014) did not observe significant differences in the nitrogen accumulation in the grains, probably due to inefficient BNF by rhizobia strain of soil.
The highest nitrogen accumulation in the grains (65.30kg ha -1 ) in MT15 treatment suggest the effectiveness of this strain in the fixation of nitrogen, and and corroborates the observation of Melo and Zilli (2009), where the BNF contributes for an adequate cowpea nutrition in Brazilian agroecosystems, and may replaces the nitrogen fertilization in the cowpea cultivation.
The number of nodules did not present significant differences in contrasts accomplished amongst treatments (Table 4).
Regarding the number of nodules, the results testified that nodulation of cowpea can be affected by the association of this crop with strains of soil native rhizobia.The presence of nodules in the absolute control plants (uninoculated plants) proves this hypothesis, suggesting native rhizobia strains established in soil where this experiment was conducted, being supported by the results found by Zilli et al. (2006), who proved the existence of a large genetic diversity of rhizobia capable of nodulate the cowpea in the soil of Brazilian Cerrado.Similar result was observed for the dry matter of nodules, in which do not presented significant differences in contrasts accomplished amongst treatments (Table 5).

Treatments Average Contrast and coefficients
The absence of significant difference may be related to the diverse size of nodules within the same treatment.The values observed in MT8+MT16 (1.16 g plant -1 ), MT8 (1.03 g plant -1 ) and absolute control (1.00 g plant -1 ) treatments, show that these treatments showed higher nodules in relation to those found in other treatments, as in the absolute control, confirming the existence of a miscellany of rhizobia strain able to form symbiosis with this legume in the soils of Cerrado (ZILLI et al., 2006).
These results also corroborate those found by Rebeschini et al. (2014), who assess the effect of inoculation of rhizobia and nitrogen application on cowpea in two traditionally growing area of bean (Cafeara and Florestopolis, both of Parana state, Brazil), in a ultisol,  and they did not observe significant difference between inoculated and uninoculated treatments, neither between the areas wherein the experiment was carried out.
The values observed in the present study for the MT8+MT15 combination (33.7% greater than treatment with 70 kg ha -1 N-urea) and in the MT15 strain (15% greater than treatment with 70 kg ha -1 N-urea) was greatest those reported by Farias et al. (2016), who evaluate the agronomic efficiency of two strains approved as inoculants for cowpea, in Maranhão State, and observed significant differences for the number of pods per plant in the control (with 80 kg ha -1 mineral N-urea) the value of 4.83 pods per plant, following for strain UFLA03-153 with 4.09 pods per plant.
These observations can be explained based on the interaction legume-rhizobia-environment, which makes possible to obtain further representative responses in the field, and due to the nitrogen supplying during the preflowering and grain filling periods, once that are the periods when to occur major assimilation rates of CO 2 by plants, increasing the pod yield (MOREIRA; SIQUEIRA, 2006;OLIVEIRA et al., 2012).
Combined inoculation of rhizobia on the cowpea development in the soil of Cerrado

Treatments
Average (g plant -1 ) Contrast and coefficients  Analysis of the grain yield data showed that all nitrogen sources influenced this parameter when compared to absolute control (C1, p≤0.001).Of all combinations, the MT8+MT15 presented highest value of grain yield, followed    The MT8+MT15 combination presented an increase of 16.7% in relation to nitrogen fertilization treatment, whereas the combinations MT15+MT16 and MT8+MT16 stand out regarding combination MT8+MT15+MT16, with increases of 13.8% and 8.7% in relation to nitrogen fertilization treatment, respectively.These results corroborate the positive effect of the use of combined rhizobia strains allows the effective establishment in soil compared to native rhizobia strains (TRABELSI; MHAMDI, 2013).
This results are similar to those found by Farias et al. (2016), who evaluated the agronomic efficiency of strains approved as inoculants for cowpea and observed highest grain yields in the inoculated treatment 41 kg ha -1 ), differing to the control (uninoculated and with 80 kg ha -1 mineral N-urea), which presented the highest grain yield of all treatments.
Regarding nitrogen concentration and crude protein in grain, significant difference was observed in C9 contrast (p≤0.05), in which the MT15 strain was found 50.40 g kg -1 of nitrogen concentration in grain and 315 g kg -1 of crude protein in grain.For others, contrasts did not observe significant differences between treatments (Table 8).
The increase observed in MT15 strain for both parameters was 30.9% in relation to treatment with nitrogen fertilization, and confirmed that the BNF contributes to The values found in crude protein in grain were higher than expected for BRS Tumucumaque (235.3 g kg -1 ) (EMBRAPA, 2014).The results of this study differ from those found by Ddamulira et al. (2015) in other Brazilian cultivars of cowpea, who observed values of 252 g kg -1 (BRS Carijó), 258 g kg -1 (BRS Tapaihum), and 280 g kg -1 (BRS Pujante).
The data of nitrogen accumulation in the grains showed that all nitrogen sources influenced this parameter (C1, p≤0.001).The MT8+MT15 combination presented significant differences (C4, p≤0.05) when compared to all others combinations.The MT8+MT16 and MT15+MT16 combinations were significantly higher regarding to triple combination MT8+MT15+MT16 (C5, p≤0.01).The inoculated plants with MT15 strain presented significant difference in relation to MT8 strain (C9, p≤0.05).In other contrasts did not was observe significant differences (Table 9).
The MT8+MT15 combination presented an increase of 27.8% in relation to nitrogen fertilization, whereas the MT8+MT16 and MT15+MT16 combinations presented an increase of 23.3% and 14.5% in relation to nitrogen fertilization treatment, respectively.The MT15 strain presented an increase of 42.7% in relation Combined inoculation of rhizobia on the cowpea development in the soil of Cerrado Nitrogen concentration in grains

Treatments
Average (g kg -1 ) Contrast and coefficients   to treatment with nitrogen fertilization.These results show that the BNF process is responsible for the largest accumulation of this nutrient in cowpea plants (RODRIGUES et al., 2013b).
Different results were found by Rufini et al. (2014), who evaluate the agronomic efficiency of rhizobia strains approved as inoculant for cowpea in a Rhodic Eutrudox in Lavras (Minas Gerais) and did not observe significant differences among the inoculated and noninoculated treatments.This response shows that inoculation of single rhizobia strains, in some cases, it not influences positively the nitrogen accumulation in the grains of cowpea (FRIGO et al., 2014).

CONCLUSION
1.The combined inoculation of rhizobia strains can be used to improve the cowpea cultivation, and present potential for increase the grain yield of this crop; 2. The combinations MT8+MT15, MT8+MT16, and MT15+MT16 were capable to provide for cowpea plants a nitrogen amount equivalent to nitrogen fertilization with 70 kg N ha -1 urea.

Table 1 -
Physicochemical properties from the soil at the depth of 0.0-0.20 m

Table 2 -
Average values of chlorophyll index assessed at flowering period (40 days after sowing)

Table 3 -
Average values of chlorophyll index assessed at grain filling period (60 days after sowing)

Table 4 -
Average number of nodules in cowpea plants subjected to combined inoculation of rhizobia strains in Cerrado soil

Table 5 -
Average values of dry matter of nodules in cowpea plants subjected to rhizobia inoculation in Cerrado soil

Table 6 -
Average values of number of pods per plant of cowpea plants subjected to rhizobia inoculation in Cerrado soil

Table 7 -
Average values of grain yield in cowpea plants subjected to rhizobia inoculation in Cerrado soil

Table 8 -
Average values of nitrogen concentration and crude protein in grains of cowpea plants subjected to rhizobia inoculation in ns e *-Non-significant contrast, and significant at 5% of probability by F-test, respectively