Assessment of Rhizobia Strains Isolates of Soils around Lake Victoria Basin for their Effectiveness in Nodulation and Symbiotic Efficiency on Soybeans and Bambara Groundnuts

The symbiotic Biological nitrogen fixation (SBNF) is a sustainable and low-cost alternative to expensive and inaccessible inorganic fertilizers. However, SBNF is underutilized in soils of Lake Victoria basin due to insufficient information on local rhizobial strains diversity and their N-fixation efficiency. This study was carried out to assess the effectiveness of rhizobium strains isolates of Kisumu, Port Victoria, Kendu bay and Karungu soils within Lake Victoria basin in nodulation and symbiotic efficiency on soybeans and bambara groundnuts. Two bambara seeds of groundnut landraces; Kakamega Cream (KAKC) and Busia Brown (BUSB) used in this study were collected from farmers in Kakamega and Busia counties respectively. Screen house experiment was performed at Kenya Forestry Research (KEFRI) in plastic pots with four seeds of each cultivar which was later thinned to two plants. Randomized Complete Block Design (RCBD) was used. Experiments data were subjected to analysis of variance (ANOVA) using Genstat 16 Edition and significant means separated using Least Significant Difference at [LSD5%] and Dancun Multiple Range Test (DMRT). Result indicated highly significant (p<0.05) effect of isolate inoculation on number of nodules per plant. Soybean Variety SB19 formed effective nodules with rhizobia in the genera Bradyrhizobium, Rhizobium and Agrobacterium. On the other hand, ‘Safari’ was quite selective and formed very few nodules with isolates identified as Bradyrhizobium. However, both varieties SB19 and ‘Safari’ had better growth under glasshouse inoculation with Bradyrhizobium spp., rhizobia isolates although one Rhizobium isolate (SoyKis1) resulted in good nodulation of both varieties. Seed treatment of the two legumes with some isolates resulted in improved nodulation and better plant growth; in some instances, outperforming the commercial strain Bradyrhizobium japonicum USDA110. In conclusion, Isolates BAMKis12, BAMKis8, BAMKis4, BAMKbay8 and SoyKar2 were found to be potential elite strains and are recommended for more host range tests as viable inoculants sources. Keywords—Rhizobia, Nodulation Effectiveness, Soybeans, Bambara Groundnuts.

ultimately comes from atmospheric nitrogen (N2) and has to been converted into forms available for living organisms [6]. The biosynthetic process of conversion starts with reduction of N2 into NH3 (ammonia) in a process called nitrogen fixation by the rhizobia bacteria [7]. Majority of legumes have a distinctive symbiotic association with Nfixing rhizobia [8]. The symbiosis results in the plant roots being invaded by the bacteria to form organized plant root structures known as root nodules [9]. Effective use of this association requires knowledge of indigenous rhizobia that can effectively nodulate with legumes such as soybeans and bambara groundnuts cultivated in the region [10]. This information which is required for selection of competitive indigenous strains for use as bio-fertilizers in seed inoculation programmes is currently limited. Commercial inoculants available in the market are made from exotic strains most of which may be less adapted to the local soils than indigenous strains. As a result, the potential of local strains to optimize BNF for improved soil N-fertility is not fully documented. The BNF is a sustainable and lowcost alternative to inorganic fertilizers in smallholder farming systems. Nonetheless, it is underutilized in soils of Lake Victoria basin partly due to insufficient information on the diverse populations of rhizobial strains and their Nfixation efficiency. To overcome these drawbacks, identifying resident and environmentally adaptable N-fixing rhizobia with superior symbiotic abilities and utilizing them in local cropping systems is necessary [11]. This is important because symbiotic efficiency of rhizobia is strongly influenced by edaphic and environmental factors [12] [13].  Fig. 1 and site characteristics are shown on Table 1.

Glass house inoculation experiments at KEFRI headquarters
Authentication and symbiotic efficiency glasshouse pot experiments were arranged in Completely Randomized Design (CRD) with the two soybean varieties or the two bambara groundnut landraces as treatments in four replicates. Sixty four and 42 rhizobial colonies isolated from bambara groundnuts and soybeans respectively were evaluated by 16S rRNA gene sequences and grouped according to the maximum identity of the genus with those at the NCBI gene bank.  [15]. Seeds were prepared as previously described and spread on water-agar (4:1) plates. The plates were inverted and incubated in dark cabinets at 25 o C for germination and were considered to be successfully germinated when their radicles reached the same length as the seed or longer. Improvised Leonard jars were prepared by sterilizing in 5% sodium hypochlorite solution for 15 minutes. Vermiculite obtained from a commercial supplier in Nairobi's industrial area was sterilized at 121 o C for 1 hr and the pH normalized to near neutral using 5% CaCO3 solution and then filled in the jars.
Each filled jar bottom was covered with absorbent sterile cotton wool to allow free flow of solutions while limiting contamination. Two seedlings of each variety/landrace were aseptically transferred into each vermiculite pot at a depth of 5 mm using sterile pipette tips. The inoculum was prepared by streaking each isolate on freshly prepared YEMA plates. 1 ml of the pure culture was aseptically picked, re-suspended into sterile bottles containing 10 ml of Yeast Extract Mannitol broth and left on a shaker for 2 days. Eight replicates per isolate were inoculated directly around seedling hypocotyls 3 days after establishment using sterile disposable pipette tips. On each occasion, 1 ml (about 10 8 cells) of bacterial suspension was used to inoculate each seedling. Immediately after inoculation, vermiculite surfaces were covered with steam sterilized sand to inhibit contamination.

Growing conditions, data collection and harvesting
The experiments to evaluate symbiotic status of selected isolates were conducted in glasshouse-controlled conditions at 16/8 hr light/dark cycle, 25/18 °C day/night temperature and 70% relative humidity. The test seedlings and the negative controls (-N) were watered twice each week using 20 ml of sterile N-free nutrient solution [16] but the solution was increased to 40 ml after 2 weeks of growth. The positive controls (+N) were supplied with 10 ml of 0.5 M KNO3 solution each week to supply the plants with nitrogen. After 45 days, the growth media was gently shaken off the plants, the roots carefully washed and the number of well-formed nodules counted. Nodules (when available), roots and shoots from the same pot were considered to belong to the same unit and were put in separately marked 2×1.75 cm zip-lock polythene bags for laboratory analysis. Fresh weights in grams per plant was taken in four replicates and recorded on an analytical electronic balance (The Lab Depot Inc, USA). Nodules, shoots and roots were enclosed in aluminium foil and inserted in an oven set at 70 o C for 1-2 days and dry weights obtained on an analytical electronic balance (The Lab Depot Inc, USA). Representative shoot samples were ground using a mortar and pestle and submitted to Kenya Industrial Research and Development laboratories in Nairobi for analysis of Ncontent using the Kjeldahl method. The procedure used was as follows; Approximately 1 g of the sample was placed in a digestion flask and 1.5 g of potassium sulfate, 0.04 g anhydrous copper sulfate and 1.0 g alundum granules was added followed by 2.0 ml of sulfuric acid. The flask was placed on preheated burner and heated for 90 min until white fumes appeared while swirling gently. The mixture was cooled to room temperature while adding 250 ml of distilled water. Three drops of tributyl citrate was added then 45% sodium hydroxide solution drop by drop. The flask was immediately connected to a distillation apparatus and distilled at 25 o C until at least 150 ml distillate was collected in titrating flask and N concentration determined.

III. RESULTS Glasshouse Authentication Experiments of Rhizobial isolates
The population of rhizobia from the soils based on the MPN results were; 1.7×10 3 , 2.0×10 3 , 1.9×10 3 and 1.5×10 3 for Port Victoria, Kisumu, Kendu bay and Karungu respectively. Host-strain interaction test results are shown in table 2. Out of the 64 bambara groundnut isolates, only 24 formed nodules with landrace KAKC out of which 7 were highly effective, 9 were partly effective while 8 were ineffective. Eighteen isolates were able to form nodules on landrace BUSB, but only 6 highly effective nodules were observed. In particular, soybean Variety 'Safari' was quite selective in nodulation potential failing to form nodules with 36 out of the 42 isolates. Only 2 isolates formed highly effective nodules on 'Safari' while effective nodules were observed in plants inoculated with 4 isolates. A distinction was observed on Variety SB19 which produced nodules with 22 isolates in which 9 were effective and 4 formed highly effective nodules. Majority of effective isolates on bambara groundnuts and soybeans were derived from Kisumu and Karungu soils.

Performance of Selected Isolates on Bambara
Groundnuts Glasshouse inoculation experiments on two bambara groundnut landraces with 18 isolates from this study and two reference strains (Bradyrhizobium spp. strains KFR259 and USDA110) resulted in variation in nodule, shoot, root and total plant biomass factors evaluated as shown in Table 3. There was a highly significant (p<0.05) effect of isolate inoculation on number of nodules per plant, although no clear landrace difference (p>0.05) was observed on this factor. Plants inoculated with isolates BAMKis1, BAMKis8, BAMKis4 and BAMKar3 which were genetically identified as Bradyrhizobium spp. produced highly effective and significantly (p<0.05) the highest mean nodule number per plant (above 20) followed by two isolates BAMKbay8 and BAMsp3 identified as Burkholderia spp. also giving above 20 nodules per plant. Results of mean nodule number per plant obtained from the six isolates compared favorably with those obtained from the reference strains Bradyrhizobium spp. strain KFR259 and Bradyrhizobium japonicum strain USDA110, which produced an average of above 30 nodules per plant on landrace KAKC with the highest value of 49 nodules occurring on plants inoculated with strain USDA110. One isolate from Kisumu soils BAMKis1 genetically identified as Rhizobium spp. produced highly effective nodules on both landraces which numbered above 30, outperforming some of the Bradyrhizobia spp. (BAMKis8 and BAMKar3) that produced below 25 nodules per plant. Nodules classified as effective based on internal nodule colour were mostly found in plants inoculated with isolates genetically identified as Rhizobium (BAMKar2, BAMKar1 and BAMKis2each giving below 20 nodules per plant).Inoculation with isolates identified as Agrobacterium spp. resulted in no nodules (BAMKar9, BAMsp5 and BAMKbay9) except one isolate BAMKbay1 which produced above 10 nodules per plant most of which were effective. Results on mean nodule fresh and dry weights followed a similar trend and posted highly significant response due to isolate inoculation (p<0.05), but there was no significant landrace effect on the two variables (p>0.05). The greatest mean nodule fresh weight of 0.477 g per plant was obtained from plants inoculated with isolate BAMKis1, producing slightly higher values (p<0.05) than both the reference strains, followed by plants inoculated with isolates BAMKis4 and BAMKis12 giving 0.449 g. Nodules retrieved from plants inoculated with isolate BAMKis1 had the highest mean dry weight of 0.073 g per plant followed by plants inoculated with BAMKis4 and BAMKis12 which ranged between 0.064 g and 0.065 g, values which were nearly identical to the mean dry weight of 0.067 g obtained from landrace KAKC plants inoculated with the reference strain USDA110 (p<0.05). The mean root and shoot dry weights did not give a significant (p>0.05) difference due to inoculation but numerically the highest root dry weight of 0.302 g per plant was surprisingly obtained from non-nodulated plants of landrace KAKC inoculated with BAMKar9 -(identified as Agrobacterium sp.), which was much higher than the results obtained from the +N un-inoculated positive control (averagely 0.25 g per plant for both landraces). Although not significantly different (p>0.05) numerically the highest shoot dry weights of 2.32 g and 1.59 g per plant were obtained from the +N control plants with all the plant appearing tall and exhibiting large leaves. The total plant biomass also showed no significant difference (p>0.05) due to inoculation having been obtained from both the root and shoot dry weights.

Performance of Selected Isolates on Soybeans
Glasshouse inoculation experiments of two soybean varieties (SB19 and 'Safari') with 16 isolates and two reference strains resulted in variation in plant growth as shown in Table 4. There was a significant (p<0.05) effect of isolate inoculation on the mean number of nodules obtained from both soybean varieties although there was no significant (p>0.05) inter-varietal difference for the same variable. Plants inoculated with isolates SoyKis3 and SoyKar2 identified as Bradyrhizobium elkanii had significantly (p<0.05) the highest nodule number per plant, giving over 20 highly effective nodules from SB19 and 'Safari'. Inoculation of SB19 with isolates SoyKar2 and SoyKar4 (Bradyrhizobium spp.) from Karungu soils resulted in effective nodules which ranged from 13.8 to 16.5 nodules per plant, significantly higher (p<0.05) than plants inoculated with isolates from Kendu bay soils which had less than 10 nodules per plant. It was evident that most isolates from Port Victoria soils (Soysp2, Soysp3 and Soysp4within Rhizobium and Agrobacterium genus) were incompatible with both soybean varieties used in this study as the isolates did not form any nodules; with only Soysp1 producing six nodules on SB19. A comparison of mean nodule fresh weight revealed no significant (p>0.05) varietal differences between noninoculated SB19 and 'Safari', but there was a significant response to inoculation in both varieties (p<0.05). The greatest mean nodule fresh weight of 0.350 g was obtained from varieties SB19 and 'Safari' inoculated with SoyKar1 while the least nodule fresh weight was obtained from plants inoculated with SoyKbay2 and SoyKbay1 both giving a similar value of 0.005 g for SB19. Plants inoculated with SoyKis1 and the reference strain KFR259 produced higher nodule dry weight (p<0.05) giving a similar value of 0.086 g. A significant (p<0.05) variation in mean shoot dry weight occurred with +N positive control plants producing the highest value of 1.434 g on the specific 'Safari' variety, followed by the same treatment on Variety SB19 with a value of 1.159 g(p<0.05). Inoculation with the reference strain USDA110 and one isolate SoyKis1 resulted in above 1 g shoot dry weight, significantly higher than the values obtained from all the other isolates. No significant difference (p>0.05) due to treatment effects occurred between the varieties with regard to root dry weight. The mean total plant biomass was significantly (p<0.05) affected by inoculation with the Variety 'Safari' posting values of above 1.0 g with most of the isolates which produced nodules. However, +N plants produced the highest mean plant biomass of 1.704 g.     [29]. Variety SB19 had a wider host range forming partially effective nodules with some non-Bradyrhizobia groups including SoyKis1 which was identified as Rhizobium sp. and SoyKar3 identified as Mesorhizobium spp. further confirming its promiscuous status. However, the symbiotic efficiency of the two strains (SoyKis1 and SoyKar3) was lower than that of strains identified as Bradyrhizobia. These findings indicate that SB19 may have a greater competition for nodule occupancy in cultivated soils with a wider range of rhizobial populations, which may negatively affect Nfixation. This challenge could be addressed by seed treatment of the promiscuous host with competent strains. This may provide a competitive advantage and result in higher nodulation [30].

International Journal of Environment, Agriculture and Biotechnology (IJEAB)
V. CONCLUSIONS Soybean Variety SB19 showed 'promiscuous' tendencies and formed effective nodules with rhizobia in the genera Bradyrhizobium, Rhizobium and Agrobacterium while 'Safari' was quite selective and formed very few nodules with isolates identified as Bradyrhizobium, which were mostly obtained from Kisumu soils. Both varieties SB19 and 'Safari' had better growth under glasshouse inoculation with isolates identified as Bradyrhizobium spp., although one Rhizobium isolate (SoyKis1) resulted in good nodulation of both varieties.

VI.
RECOMMENDATIONS Seed treatment of the two legumes with some isolates resulted in improved nodulation and better plant growth; in some instances, outperforming the commercial strain Bradyrhizobium japonicum USDA110. Isolates BAMKis12, BAMKis8, BAMKis4, BAMKbay8 and SoyKar2 are promising elite strains and are recommended for more host range tests as potential inoculants sources.