Background

Symbiotic N2 fixation (SNF), a biological process of transforming the atmospheric N2 by mutual interaction of the host plant with soil bacteria is an essential environmentally and economically sustainable sources of N to the soil (Silva and Uchida 2000), thereby reducing the use of chemical N fertilizer. Different rhizobial species belonging to the genera Rhizobium, Agrobacterium, Ensifer, Bradyrhizobium, and Ochrobactrum are able to produce nodules with common bean plants (Wang et al. 2016). Inoculation is a key biological input to improve crop productivity and soil fertility through increasing the rhizobia in the plant rhizosphere (Keyser and Li 1992; Remans et al. 2008), thereby improving nodulation and N2-fixation (Peoples et al. 1995) and it can also fix to exceed 200 kg N ha year−1 (Giller 2001). The symbiotic N2relationship between common bean and Rhizobium contributed up to 90 kg N ha−1 which was 40–50% of the total N near physiological maturity (Westermann et al. 1981). Several studies indicated the promising potential of common bean to fix N2 derived from the atmosphere (Asadi Rahmani et al. 2005; García et al. 2004; Remans et al. 2008).

The efficacy of rhizobial strains in nodulating and fixing atmospheric N with common bean varies with both the host genotypes and the Bacterium strains (Aguilar et al. 1998; Caballero-Mellado and Martinez-Romero 1999; Farid and Navabi 2015; Michiels et al. 1998; Moawad et al. 1998). The prevailing environmental conditions significantly shape the diversity and distribution of indigenous rhizobia nodulating common bean (Wang et al. 2016). Deficiency of different essential nutrients have also been reported as legume-Rhizobium symbiosis limiting environmental factors, which may limit the nodulation and N2 derived from the atmosphere (Divito and Sadras 2014). Soil water availability, which is one of the major environmental factors, also influences the N2 fixation derived from the atmosphere by common bean (Devi et al. 2013) and soybean (Collino et al. 2015). This variability often limits the nitrogen-fixing performance of soil native rhizobia or use of commercially available inocula. Strains of rhizobia widely differed in their abilities to survive, nodulate and fix Nin soil environments (Slattery et al. 2001). Considering the high level of adaptation by native rhizobia to local soil conditions, it is important to characterize the indigenous rhizobial collection for use in inoculant production.

Many research reports indicated that host genotypic factors affect nodulation and nodule activity in Phaseolus vulgaris (Graham and Temple 1984; Rennie and Kemp 1983). Nleya et al. (2001) also illustrated the different response of common bean genotypes to the application of Rhizobium inoculant. Hardarson et al. (1993) also found that N derived from the atmosphere (% Ndfa) varied from 35 to 70% among different common bean genotypes. Usually, bushy growth habit of common bean has the lowest N fixation efficiency among all legume crops (Bliss 1993; Hardarson et al. 1993; Isoi and Yoshida 1991; Martinez-Romero 2003). Indeterminate genotypes generally can fix more nitrogen than determinate genotypes due to the greater “sink” in the indeterminate variety (Ofori and Stern 1987). Bliss (1993) identified common bean genotypes capable of fixing enough atmospheric N2 to support the grain yield of 1000–2000 kg ha−1. Therefore, improvement of bean BNF requires a multidisciplinary approach that will increase the host capacity to fix N (Giller 2001) and selection of effective Rhizobium strains that can compete for nodulation with native populations of bacteria present in most soils. So far, the effect of environmental condition on Rhizobium-common bean genotypes is not well known. Almost no attempt has also been made on effective bushy type common bean genotypes (with variable maturity time)-Rhizobium symbiosis, which can give higher responses in different environment conditions. Hence, the objective of this work was to evaluate the effect of bushy type common bean genotypes, Rhizobium strains and environment interaction on the nodulation, yield and yield traits of common bean in soils of eastern Ethiopia.

Methods

Description of experimental sites

Field experiments were conducted at four locations, including Hirna (09°13.157′N and 041°06.488′E at an altitude of 1779.6 m above sea level [m.a.s.l.]), Fedis (09°06.941′N and 042°04.835′E at an altitude of 1642.8 m.a.s.l.), Babile (09°13.234′N and 042°19.407′E 1643.4 m.a.s.l.) and Haramaya (09°24.954′N and 042°02.037′E at an altitude of 1999.4 m.a.s.l.) agricultural research centers representing the major common bean cultivating areas of Ethiopia in 2013. The fields were located in the eastern parts of Ethiopia where common bean had long been grown intercropped with sorghum and maize without inoculation. The location map of the study site was previously indicated in Argaw (2016).

Soil sampling

The initial soil samples were collected from the top 0–20 cm for analysis of the soil physico-chemical properties. A composite soil comprising 20 auguring sampling points from each experimental site was taken and transported back to the laboratory within a day. Representative subsamples of 1 kg each were prepared for most probable number (MPN) assay and stored in a refrigerator at 4 °C until used for enumerating indigenous rhizobial population. The soil physico-chemical properties were analyzed using standard procedures employed by Sahlemedhin and Taye (2000).

Soil properties

The soils of the study sites had clay, sandy loam, sandy clay loam and silty clay loam in Hirna, Babile, Haramaya and Fedis sites, respectively. The pH(H2O) of the study sites ranged from 6.66 to 7.84 which is within the suitable pH ranges for Rhizobium species. All experiment sites had the electric conductivity less than 0.14 ms cm−1. The soil organic carbon and total N content were 1.65 and 0.06%, 0.56 and 0.06%, 1.96 and 0.12%; and 1.32 and 0.12% in Hirna, Babile, Haramaya and Fedis sites, respectively. The soil had the CEC ranging from 6.59 cmol(+) kg−1 in Babile to 39.88 cmol(+) kg−1 in Hirna site. The soil of the study sites had exchangeable Ca+2, Mg+2, Na+1 and K+1 with ranges of 39.88–4.18, 12.87–3.5, 0.33–0.12 and 1.09–0.14 cmol(+) kg−1, respectively.

Source of the isolates and common bean seed genotypes

Eight isolates of Rhizobium spp. were obtained from Biofertilizer Research and Production Project (BRPP), Haramaya University (Haramaya, Ethiopia). The isolates were designated as HUCBR-1, HUCBR-2, HUCBR-3, HUCBR-4, HUCBR-5, HUCBR-6, HUCBR-7, and HUCBR-8. All isolates used in this study were obtained from Ethiopian soils. All isolates were previously characterized as superior isolates in nodule formation and shoot biomass production of common bean under greenhouse conditions (Argaw 2007).

Seeds of Phaseolus vulgaris genotypes used in this study were obtained from Lowland Pulse Research Program, Haramaya University, Haramaya, Ethiopia. The selected genotypes were characterized as highly productive genotypes in the study sites. Beside this, maturity time was also considered for selection of genotypes for this experiment. Accordingly, Gofta, Kufanzik and Dursitu genotypes belong to early, medium and late maturing categories, respectively.

Preparation of inocula

The pure cultures of Rhizobium isolates were obtained from the laboratory in slant culture. The bacteria were purified by culturing in YEM (Yeast extract mannitol) agar medium and then single pure colony was transferred into YEM broth medium and kept at 30 °C for 7 days on a rotary shaker at 120 rpm. About 400 ml of culture liquid medium containing appropriate Rhizobium sp. were added to 1 kg of the carrier (sterile fine filter mud) and mixed thoroughly and then packed in plastic bags. Filtermud-base inoculum was incubated at 26–28 °C for 15 days. At the time of inoculation, the number of rhizobia in the inoculum was estimated using plate count method. One ml samples of serially diluted inocula from 10−6 dilution were plated in YEMA medium. Colonies that developed after incubation at 28 °C for 5–7 days were recorded. This test indicated that the number of rhizobia was more than 1 × 109 g−1 inocula.

Experimental layout and treatments

The experimental fields were plowed thoroughly twice with a tractor and divided into sub-plots in accordance with the treatments. The net size of each experimental sub-plot was 3 × 2 m2. There were five rows per plot and the spacing was 1 m between plots, 40 cm between rows and 10 cm between plants. Ten levels of inoculation containing eight Rhizobium isolates (NSCBR-14, NSCBR-(25)2, NSCBR-59, NSCBR-31, NSCBR-16, NSCBR-18, NSCBR-57 and NSCBR-25) with uninoculated and N-fertilized (20 kg N ha−1) control and three common bean genotypes were factorially combined. Before sowing, 20 kg P ha−1 as tri superphosphate for all experimental plots were applied in furrows. Identical field experiments were carried out in four locations.

Common bean seeds were sterilized using 70% ethanol for 1 min and NaClO solution (0.25% as available Cl) for 3 min. The seeds were then washed carefully in sterilized deionized water five times before sowing. Then, 20 g of the different rhizobia inoculants was added to different polyethylene bags containing 200 g of common bean seeds. A 10% (w/v) sucrose solution to increase adherence was added to each bag to enhance proper mixing and adhesion of the rhizobia carrier material to the common bean seeds. After mixing, seeds were allowed to air-dry in the shade for 15 min and sown field layout. Two seeds were planted by hand per hole and later thinned down to one per hole 1 week after germination. A total of 30 treatment combinations were used in the experiment. The experiments were designed as two-factor experiments in a randomized complete block design (RCBD). There were three replications of each treatment. All standard local cultural practices were accomplished throughout the growth period. Manual weeding was done whenever required.

Nodulation, yield and yield attributes

At late flowering and early pod setting stage, five plants were randomly chosen from central three rows for the evaluation of nodulation and plant growth. Adhered soil on the sampled plants were loosen by placing into plastic buckets filled with water. Thereafter, nodules from roots were picked and following data were recorded: (1) Nodule number plant−1, and (2) nodule dry weight plant−1. Shoot dry weight was also measured after drying the samples at 70 °C in the electrical oven until the weight of the samples became constant. Shoots of the plants were later ground to pass through a 0.5 cm sieve. Total N determinations were done by the Kjeldahl method of Bremner (1965). At full maturity stage, numbers of pods plant−1, the number of seed pod−1, plant height at harvest and total biomass were recorded. Grain yield was corrected for 13% moisture content after determining humidity level with a grain moisture tester.

Data analysis

Data were subjected to analysis of variance (ANOVA) using Statistical Analysis System (SAS Institute Inc 1999). Statistically significant differences between treatment means were also determined using the least significant difference (LSD) test at 5% probability level of significance (SAS Institute Inc 1999). Figures were prepared using excel Microsoft of version 10.

Results

Nodule number

Analysis of variance (ANOVA) showed that Rhizobium inoculation, experimental location, the genotypes and their interaction significantly affected the nodule number (NN) at P ≤ 0.05 (Table 1). The effect of Rhizobium inoculation treatments on NN varied due to different varieties and experimental locations (Table 2). At Haramaya site, most isolates, except NSCBR-59 and NSCBR-31 inoculations, resulted in significant increase in NN with Dursitu. With Gofta variety, all tested isolates with the exception of isolate NSCBR-59, significantly increased the NN. With exception of NSCBR-31, all isolates resulted in significant increase in NN with Kufanzik genotype.

Table 1 Summary of ANOVA results for all investigated traits of common bean affected by Rhizobium inoculation, locations and common bean genotypes and their interaction, in Haramaya, eastern Ethiopia, during 2012/2013 main cropping season
Full size table
Table 2 Nodule number of common bean investigated from three genotypes (Dursitu, Kufanzik and Gofta) over four experimental locations (Babile, Fedis, Haramaya and Hirna), in eastern Ethiopia, in 2012/2013 main cropping season
Full size table

At Hirna site, NSCBR-59, NSCBR-31, and NSCBR-57 inoculations had significantly higher NN than the control check with Dursitu genotype. However, most isolates, except NSCBR-59, NSCBR-18, and NSCBR-57, increased the NN significantly with Gofta. With Kufanzik genotype, a significant increase in NN was recorded in NSCBR-14, NSCBR-16, and NSCBR-18 treatments.

At Babile site, significantly higher NN was recorded with inoculation of NSCBR-(25) and NSCBR-18 with Dursitu than the uninoculated control. Significant increase in NN of Gofta was recorded with NSCBR-14, NSCBR-(25)2, NSCBR-59 and NSCBR-31 treatments. However, NSCBR-(25)2 isolate significantly increased the NN of Kufanzik. At Fedis site, a significant increase in NN of Dursitu inoculated with all isolates with the exception of NSCBR-31 and NSCBR-18 was recorded while NSCBR-14, NSCBR-(25)2, NSCBR-59 and NSCBR-31 isolates inoculated Gofta resulted in an increase in NN.

In general, Dursitu inoculated with all isolates except NSCBR-31, produced the highest number of nodules increase over the uninoculated control while this highest increase with Kufanzik was recorded at NSCBR-31 inoculation (Fig. 1a). However, the highest increase in NN of Gofta over the control check was obtained from NSCBR-59. The highest means of NN (216.17, 221.93, 106.27 and 152.37) were induced with Dursitu at Haramaya, Hirna, Babile and Fedis sites, respectively, over other treatments while the lowest were recorded from uninoculated control at all sites.

Fig. 1
figure 1

The effect of Rhizobium isolates inoculation on a Nodule number, b Nodule dry weight, c Total biomass yield, d Grain yield and e Total N accumulation of three varieties of common bean, in eastern Ethiopia, 2012/2013 cropping season

Full size image

Nodule dry weight

The effect of Rhizobium inoculation, the genotypes, experimental locations and their interaction was significant on nodule dry weight (NDW) (Table 1). The effect of inoculated isolates on NDW varied with different genotypes and in the different experimental sites similar to the result obtained in NN (Table 3). At Haramaya site, Dursitu inoculated with all Rhizobium inoculation treatments, except NSCBR-(25)2, NSCBR-59 and NSCBR-18, produced significantly higher NDW than the control check. Inoculating NSCBR-14, NSCBR-(25)2 and NSCBR-16 on the genotype Gofta increased significantly the NDW. However, only NSCBR-59 inoculated with Kufanzik significantly increased NDW when compared to the control check.

Table 3 Nodule dry weight of common bean investigated from three genotypes (Dursitu, Kufanzik and Gofta) over four experimental locations (Babile, Fedis, Haramaya and Hirna), in eastern Ethiopian, in 2012/2013 main cropping season
Full size table

At Hirna site, NSCBR-(25)2, NSCBR-59, NSCBR-31 and NSCBR-57 isolates significantly increased the NDW of Dursitu genotype. All, except NSCBR-(25)2 and NSCBR-18 isolates, significantly increased more NDW with Gofta than with the uninoculated control. With Kufanzik, inoculating NSCBR-16, NSCBR-18 and NSCBR-57 more significantly increased NDW than uninoculated control. At Babile site, none of the isolates with Dursitu and Kufanzik significantly affected the NDW when compared to the control. However, only NSCBR-(25)2 inoculated to Gofta significantly increased the NDW.

At Fedis site, most of the isolates excluding NSCBR-31 and NSCBR-18, significantly improved NDW with Gofta. A significant increase in NDW of Gofta was observed with NSCBR-14, NSCBR-59, NSCBR-31 and NSCBR-16. InoculatingNSCBR-14, NSCBR-59, and NSCBR-16 with Kufanzik were significantly (P ≤ 0.05) enhanced the NDW. With Dursitu, all isolates with the exception of NSCBR-59, NSCBR-31, and NSCBR-18, resulted in the highest increase in NDW over the control check, while better NDW of Kufanzik was obtained with NSCBR-59 and NSCBR-18 (Fig. 1b). Only NSCBR-31 inoculated with Gofta recorded the highest NDW over the control check. The highest NDW across the inoculation treatments was produced with Dursitu.

Total biomass yield

ANOVA revealed that the main effect of Rhizobium inoculation, the genotypes, experimental locations and their interaction were significant (P ≤ 0.05) on total biomass yield (TBY) (Table 1). At Haramaya site, NSCBR-16, NSCBR-57, and NSCBR-25 inoculated to Dursitu significantly increased in the TBY(Table 4). Isolate NSCBR-14 inoculated to Gofta and none of the isolates with Kufanzik resulted in a significant increase in the TBY. At Hirna site, significantly higher TBY of Dursitu was recorded in response to NSCBR-14, NSCBR-59, and NSCBR-31 inoculation than that of the control, while NSCBR-59 and NSCBR-18 inoculations significantly increased TBY of Kufanzik. However, the data exhibited the non-significant effect of inoculation on TBY of Gofta.

Table 4 Total biomass yield of common bean investigated from three genotypes (Dursitu, Kufanzik and Gofta) over four experimental locations (Babile, Fedis, Haramaya and Hirna), in eastern Ethiopian, in 2012/2013 main cropping season
Full size table

At Babile site, a significant increase in TBY of Dursitu inoculated with NSCBR-14 was obtained. With Gofta, better TBY than from the uninoculated control was recorded with NSCBR-(25)2 and NSCBR-(16). Kufanzik inoculated with NSCBR-(25)2 gave significantly better TBY than that of the control. Inorganic N application with all genotypes at Babile site produced the highest TBY over the other treatments.

At Fedis site, NSCBR-59, NSCBR-16, and NSCBR-57 inoculated to Dursitu produced significantly higher TBY than the control. Gofta inoculated with NSCBR-(25)2, NSCBR-31 and NSCBR-16 gave significantly higher TBY than the uninoculated control. Inoculating NSCBR-(25)2 resulted in a significant increase in TBY with Kufanzik. In contrast to nodulation, the highest TBY (2589.44 and 5036.48 kg ha−1) across the treatments were produced with Gofta at Babile and Fedis sites. At Haramaya and Hirna sites, all genotypes produced almost similar amount of TBY. With control check, Gofta recorded the highest TBY in all sites. Across locations, Gofta when compared to the other varieties recorded the highest TBY with all treatments (Fig. 1c).

Grain yields

The grain yield (GY) of common bean was significantly (P ≤ 0.05) affected by Rhizobium inoculation, the genotypes, experimental sites and their interaction (Table 1). The effects of isolates on GY were significantly variable among the different genotypes and experimental locations (Table 5). At Haramaya site, Dursitu inoculated with NSCBR-14, NSCBR-16 and NSCBR-57 produced significantly higher GY than the uninoculated control. With Gofta, applying NSCBR-14 resulted in a significant increase in GY compared with the uninoculated control. The response of Kufanzik to inoculation with NSCBR-14, NSCBR-59, NSCBR-16 and NSCBR-18 significantly affected GY.

Table 5 Grain yield of common bean investigated from three genotypes (Dursitu, Kufanzik and Gofta) over four experimental locations (Babile, Fedis, Haramaya and Hirna), in eastern Ethiopian, in 2012/2013 cropping season
Full size table

At Hirna site, all isolates, except NSCBR-18, NSCBR-57, and NSCBR-25 with Dursitu, resulted in a significant increase in GY while none of the isolates significantly affected the GY of Gofta. Kufanzik inoculated with NSCBR-14, NSCBR-16 and NSCBR-18 significantly increased the GY. At Babile site, NSCBR-14 with Dursitu gave significantly higher GY than the uninoculated control. With Gofta, NSCBR-(25)2 and NSCBR-16 inoculation increased GY significantly. However, the data revealed the non-significant effect of inoculation on the GY of Kufanzik.

At Fedis site, a significant improvement of GY for Dursitu was obtained from inoculation with NSCBR-59, while Rhizobium inoculations did not affect the GY of Gofta and Kufanzik. The highest mean GY of 2932.3, 2739.4, 1490.0 and 2065.6 kg ha−1 were recorded with Gofta in Haramaya, Hirna, Babile and Fedis sites, respectively. In all experimental sites with all treatments including uninoculated control, Gofta produced the highest GY of 3498.4, 3257.82, 1499.25 and 2204.82 kg ha−1 at Haramaya, Hirna, Babile and Fedis over Kufanzik and Dursitu (Fig. 1d).

Total plant N accumulation

ANOVA showed significant (P ≤ 0.05) effect due to Rhizobium inoculation, the genotype, experimental locations and their interaction on total plant N accumulation (TPNA) (Table 1). The effect of Rhizobium inoculation was non-significant on plant N accumulation in Dursitu at Haramaya site (Table 6). At this experimental site, inoculation with NSCBR-16, NSCBR-57 and NSCBR-25 to Gofta significantly improved the plant N accumulation, while this trait was higher in Kufanzik inoculated with NSCBR-59, NSCBR-31, and NSCBR-16 than uninoculated control.

Table 6 Total N accumulation of common bean investigated from three genotypes (Dursitu, Kufanzik and Gofta) over four experimental locations (Babile, Fedis, Haramaya and Hirna), in eastern Ethiopian, in 2012/2013 main cropping season
Full size table

At Hirna site, a significant increase in plant N accumulation by NSCBR-(25)2, NSCBR-59, NSCBR-57 and NSCBR-25 inoculated with Dursitu was recorded. None of the Rhizobium inoculations significantly affected the plant N accumulation with Gofta and Kufanzik. At Babile site, all Rhizobium inoculations did not improve the TPNA of all the tested genotypes. At Fedis site, all isolates, excluding NSCBR-14 and NSCBR-59 with Dursitu were significantly higher in plant N accumulation than the uninoculated control. However, this trait did not significantly affect when isolates were inoculated to Gofta and Kufanzik. The highest mean total plant N accumulation values of 3.6257, 3.9950, 2.8543 and 3.5637% were recorded with Dursitu in Haramaya, Hirna, Babile and Fedis sites, respectively. Like nodulation, the highest plant tissue N accumulation in all treatments including uninoculated control was recorded with Dursitu genotype (Fig. 1e).

Discussion

Utilizing Rhizobium inoculation for pulses production is a common practice in different part of the world including some countries in sub-Saharan Africa (SSA). However, the success of this inoculant technology in common bean is variable from location to location. Besides, it depends on common bean genotypes. Due to different rhizobia population size and its competitiveness in different locations and presence of specificity between Rhizobium strain-common bean genotypes (Aouani et al. 1997), we need to develop genotype and location specific Rhizobium inoculant. Hence, this study was initiated to evaluate the effect of genotypes, Rhizobium inoculation and environmental locations on nodulation and productivity of common bean in major common bean growing areas of eastern Ethiopia.

In general, the Rhizobium inoculation, the locations, the common bean genotypes and their interaction effect was significant (P ≤ 0.05) on nodulation, yield and yield traits of common bean (Table 1). This indicates the need for specific Rhizobium isolate development for each of common bean genotype when cultivating in different locations. Similar findings were previously reported on common bean (Handley et al. 1998; Mostasso et al. 2002; Popescu 1998; Remans et al. 2008). This specificity could be due to the fact that the exchanges of chemical signals between the two partners are present. The legume roots exude organic compounds (flavonoids) (Hungria et al. 1997; Long 2001), which differ between plant species and genotypes. Then after, rhizobial bacteria respond with lipo-chitin oligosaccharides, known as Nod factors, which act as specific morphogenetic signal molecules to induce the roots nodule formation (Oldroyd and Downie 2008). In addition, the result of the current study revealed the need for location specific Rhizobium development.

The present study revealed that isolates performed better in improving NN, NDW, TBY, GY and TPNA with one of the tested genotypes did not consistently exhibit with other genotypes, indicating the presence of specificity of Rhizobium isolates and common bean genotypes. Similarly, Bouhmouch et al. (2005) reported the common bean genotypes-Rhizobium specificity. This indicates the presence different infectivity potential of Rhizobium isolates with different common bean genotypes (Neila et al. 2014).

We found that relatively more number of inoculated Rhizobium performed better in NN than the background rhizobia in the Haramaya site than in the other study sites. This indicates the presence of less competitive background rhizobia in infectiveness at Haramaya site than the other study sites. The current study showed that those isolates that performed better in improving NN did not perform similarly in NDW enhancement in all study sites, suggesting that better in infectiveness is not always good in effectiveness. The present work indicated that all isolates including the uninoculated control produced the lowest mean NN and NDW in all genotypes at Babile. This was probably due to low rhizobial population in this site (Ojo et al. 2015) and this consequently reduced the nodule formation. Low nodulation formation might be also attributed to the prevailed adverse environmental condition at Babile site (Hungria et al. 2003). Elias and Herridge (2015) found that rhizobial population was positively correlated with soil moisture. Besides, the soil textural class of Babile soil was sand and had low SOM (Table 1), which could reduce the survival of inoculated Rhizobium in the soil (Hagedorn 1978; Mahler and Wollum 1981). However, Bliss (1993) suggested that the limitation of N2 fixation imposed by environmental factors could be resolved through the selection and breeding of improved common bean cultivars.

The highest NN and NDW in the control without inoculation were produced with Dursitu at Haramaya and Hirna sites and Kufanzik at Babile and Gofta at Fedis site. This suggests the presence of appropriate indigenous rhizobia, which could be different in infectiveness and effectiveness in different soils. Rodiño et al. (2011) determined common bean variety and variety × environment interaction effect on nodulation. A similar finding was reported in common bean in Canadian Prairie by Nleya et al. (2009) who found that common bean genotypes differed in nodulation formation. In addition, Ikeda (1999) found that the number of nodules was directly controlled by host genotype. This preference could have a major significance in resolving strain competition problem in Phaseolus vulgaris (Raposeiras et al. 2006).

The result of the present work indicated that those isolates induced the highest nodulation with one genotype was not consistently performed with the other genotypes. Similarly, Bonish and MacFarlane (1987) demonstrated that isolates mean effectiveness of 12% with ‘Tamar’ variety was recorded and 87% mean effectiveness with Huia variety. Differences in host variety among clover lines influence the effectiveness of the symbiosis (Hagedorn and Caldwell 1981; Sherwood and Masterson 1974).

The highest mean NN and NDW across locations and with all treatments including uninoculated control were produced in Dursitu. Dursitu at Haramaya, Babile and Fedis sites and Kufanzik at Hirna site induced the highest mean NDW across the treatments. This indicates the presence of more infectiveness by inoculated Rhizobium and background rhizobia with Dursitu rather than other tested genotypes. This might be attributed to the high promiscuity of Dursitu with several rhizobial species (Cardoso et al. 2012) apparently resulting from the capacity of the host plant to perceive a genotype of rhizobial molecular signals (Michiels et al. 1998). Significant environment by inoculant interaction effect on nodule dry weight was reported by Nleya et al. (2009). Therefore, the current work found the presence of Rhizobium isolate-genotype specificity in nodule production in a different location.

The result of the present study indicated the highest mean total plant N accumulation across treatments including uninoculated control was recorded in Dursitu as it was found in nodulation. Similar results have been previously reported by lentil and pea (Abi-Ghanem et al. 2011). This implies that improving nodulation is important traits to enhance the total N in plant tissue. Variation in plant N accumulation among genotypes could be due to the presence of variability in SNF among common bean genotypes (Hardarson et al. 1993; Nleya et al. 2002). Yadegari et al. (2010) found that Cultivar ‘Akhtar’ demonstrated the highest potential for nodulation, nitrogen fixation, and seed yield production compared to cultivars ‘Sayyad’ and ‘Goli’. Buttery et al. (1997) also compared 17 common bean genotypes inoculated with various Rhizobium strains for N fixation and they found differences among genotypes in acetylene reduction activity and seed N content.

In contrast to the finding in nodulation, the mean TBY and GY across locations were the highest in Gofta. This genotype also produced the highest mean TBY and GY across treatments including uninoculated control. The highest biomass and grain production in all experimental locations was also recorded with Gofta. This finding is consistent with the observation of Tsai et al. (1993) who found that Mexico-309 was superior for nodulation parameters but poor for seed yield, while Preto Caruaru produced high seed yield, but was inferior in nodulation traits. The yield advantage of Gofta could be attributed to its delayed maturity when compared to other tested common bean genotypes. Due to genetic makeup difference among common bean genotypes, it may record high production though induced low nodulation (Pereira et al. 1984). Conversely, Rodiño et al. (2011) found that genotypes with a big nodule phenotype showed a good plant response and more beneficial for plant growth and seed yield. In contrast to the current study, Farid and Navabi (2015) found the common bean genotypes × environment interaction for grain yield production.

Regardless of the tested genotypes, the highest TBY at Babile site was recorded with inorganic N treatment. Similarly, Hungria et al. (2003) found further increase of common bean production on average by 132 kg ha−1 with a supplement of 15 kg N ha−1 over the inoculated plants. In other experimental sites, a significant increase in TBY was obtained with Rhizobium inoculation. Similarly, Huntington et al. (1986) found that Rhizobium inoculation increased the yield by 30–80% in common bean using when compared to N fertilizer plant. In contrast to the current finding, Ruiz Diaz et al. (2009) found that the non-significant effect of inorganic N application with and without inoculation on the yield of soybean n though plant N accumulation was improved. This result could be attributed to high N2 derived from the atmosphere by soybean when compared to common bean.

It has been shown that none of the inoculated Rhizobium significantly improved the plant accumulated N at Babile when compared to the uninoculated control. This result could be attributed to dry condition and low soil moisture availability in Babile (Saito et al. 1984; Smith et al. 1985, 1988) and this cause early nodules senescence and decline in nitrogenase activity (Becana and Sprent 1987) and low N2 fixation. On the other hand, the Rhizobium inoculation at other experimental locations significantly increased plant N accumulation. This result could be attributed to the fact that more than 50% of its plant N accumulated was derived from biological N2 fixation when inoculated with effective Rhizobium under favorable condition (Pena-cabriales et al. 1993).

Some of the isolates inoculated to Dursitu accumulated significantly higher plant N than the uninoculated control but this result was not observed with the remaining genotypes. Previously investigations under field conditions (Hobbs and Mahon 1982; Rengel 2002; Young et al. 1982) have shown that some Rhizobium isolates are more efficient when inoculated on some genotypes than on others. Huntington et al. (1986) concluded from their greenhouse study that the host/endophyte combination forms a relatively ineffective symbiotic association being primarily inherent in the host plant rather than the endophyte or the environment. The current result is also consistent with the findings of Hungria and Neves (1987); Hardarson et al. (1993) and Neves et al. (1987) who found that plant N concentration in different pulse crops is influenced by the host plant cultivar as well as by Rhizobium strain. Graham (1981) and Amarger (1986) found that nitrogen fixation depends on rhizobia × line interaction and that the process of selection of efficient rhizobia should be developed with adequate lines.

Conclusion

The result of this experiment showed the presence of Rhizobium strain × locations specificity. Besides, the result exhibited the need for different Rhizobium isolate for tested common bean genotypes. The result indicated the similar performance of all common beans varieties in most of the investigated traits, except nodulation, regardless of the experimental locations. This suggests the need for specific Rhizobium strain development for biofertilizer production for different locations. Hence, we recommend the development of location-based Rhizobium isolates for inoculants production.