Genetic Diversity and Symbiotic Ef fi ciency of Indigenous Common Bean Rhizobia in Croatia

Nodule bacteria (rhizobia) in symbiotic associations with legumes enable considerable entries of biologically fi xed nitrogen into soil. Eff orts are therefore made to intensify the natural process of symbiotic nitrogen fi xation by legume inoculation. Studies of fi eld populations of rhizobia open up the possibility to preserve and probably exploit some indigenous strains with hidden symbiotic or ecological potentials. The main aim of the present study is to determine genetic diversity of common bean rhizobia isolated from diff erent fi eld sites in central Croatia and to evaluate their symbiotic effi ciency and compatibility with host plants. The isolation procedure revealed that most soil samples contained no indigenous common bean rhizobia. The results indicate that the cropping history had a signifi cant impact on the presence of indigenous strains. Although all isolates were found to belong to species Rhizobium leguminosarum, signifi cant genetic diversity at the strain level was determined. Application of both random amplifi cation of polymorphic DNA (RAPD) and enterobacterial repetitive intergenic consensus–polymerase chain reaction (ERIC-PCR) methods resulted in similar grouping of strains. Symbiotic effi ciency of indigenous rhizobia as well as their compatibility with two commonly grown bean varieties were tested in fi eld experiments. Application of indigenous rhizobial strains as inoculants resulted in signifi cantly diff erent values of nodulation, seed yield as well as plant nitrogen and seed protein contents. The most abundant nodulation and the highest plant nitrogen and protein contents were determined in plants inoculated with R. leguminosarum strains S17/2 and S21/6. Although, in general, the inoculation had a positive impact on seed yield, diff erences depending on the applied strain were not determined. The overall results show the high degree of symbiotic effi ciency of the specifi c indigenous strain S21/6. These results indicate diff erent symbiotic potential of indigenous strains and confi rmed the importance of rhizobial strain selection. These are the fi rst studies of indigenous common bean rhizobia in Croatia that provide the basis for further characterization and selection of highly effi cient indigenous strains and their potential use in agricultural practice and future research.


Introduction
Natural process of symbiotic nitrogen fi xation has a unique role in a sustainable legume production.Nodule bacteria, collectively named rhizobia, in symbiotic associations with legumes enable considerable entries of biologically fi xed nitrogen into ecosystems.In order to maxi-ISSN 1330-9862 preliminary communication doi: 10.17113/ft b.54.04.16.4740 mize the use of symbiotic nitrogen fi xation, inoculation of legume seed with effi cient rhizobial strains has become usual procedure in agriculture.The application of rhizobial inoculants enables considerable reduction of mineral nitrogen fertilization, which is of great ecological and economic importance.Nitrogen fi xing capacity in agricultural systems depends on rhizobial strain, host plant and environmental conditions (1).It is well known that rhizobial strains strongly diff er in their properties like symbiotic effi ciency, competitiveness with other rhizobia in the soil, compatibility with host plant, and adaptive ability to stress conditions in the soil.Therefore, the selection of high quality rhizobial strains is one of the main presumptions for successful inoculation.Furthermore, in order to maximize the contribution of biologically fi xed nitrogen in common bean production, it is also very important to select the most eff ective symbiotic associations between the common bean cultivar and the rhizobial strain.However, indigenous rhizobial strains can be present in some soils.In specifi c environmental conditions, these strains usually show higher competitiveness and adaptive ability than commercial strains, while their symbiotic effi ciency is usually unknown.Therefore, in order to improve the benefi cial eff ect of common bean inoculation, it is important to determine the actual composition and the characteristics of rhizobial fi eld population.
Compared to other legumes, common bean can form symbiotic relations with a number of species mostly belonging to the genus Rhizobium (2)(3)(4).Despite the great symbiotic potential of common bean, in comparison with other legumes, it fi xes considerably small amounts of atmospheric nitrogen (5).It has been recognized that symbiosis of common bean with a number of rhizobial species has a negative correlation with the degree of nodulation and nitrogen fi xing capacity (6).
Common bean (Phaseolus vulgaris L.) is the most important grain legume in human diet due to the high content of valuable nutrients that have benefi cial eff ects on human health.Common beans are important source of proteins, complex carbohydrates, dietary fi bre, vitamins and minerals and have very low content of saturated fat and cholesterol (7).Although the common bean originates from the Americas, it is widely distributed in various parts of the world.In Croatia, common bean is a traditional but completely marginalized crop despite very favourable conditions for its cultivation.In order to improve and expand the production of beans in Croatia, different investigations have been carried out, mainly with the most widespread cultivars sown in Croatia called Trešnjevac and Slavonski zeleni, while the studying of common bean rhizobia has been neglected (8).
However, many authors from diff erent countries that are the main common bean producers have studied the distribution and composition of indigenous common bean rhizobia as well as their symbiotic effi ciency (9)(10)(11).
For the fi rst time, this study has provided insight into the composition and diversity of indigenous rhizobial populations that nodulate common bean in diff erent soil types in Croatia.Besides, symbiotic characterization of indigenous strains is a prerequisite for selection of the most effi cient strains that are at the same time well-adapted to the soils of central Croatia.Investigations conducted so far in Croatia involved indigenous rhizobial populations that nodulate soybean (12,13), fi eld pea (14) and alfalfa (15).It is assumed that indigenous rhizobial strains with diff erent symbiotic characteristics are present in the soils of central Croatia.The main aim of the present study is to identify and characterize the indigenous common bean rhizobia isolated from diff erent fi eld sites in central Croatia, to determine genetic diversity within rhizobial fi eld population and to evaluate symbiotic properties of indigenous strains.
This study is the beginning of research of indigenous populations of common bean rhizobia in Croatia.The results will provide the fi rst data on the composition and characteristics of indigenous populations in central Croatia.

Materials and Methods
Twenty-three soil samples were collected for isolation of indigenous common bean rhizobia from diff erent sites in Zagreb County in central Croatia.Soil chemical analyses included determination of soil acidity in water and in 1 M KCl, humus content by Tjurin's method (16), total nitrogen content by the modifi ed Kjeldahl method (17), and available phosphorus and potassium by the method described by Egner et al. (18).
In greenhouse pot experiment, surface-sterilized common bean seeds were sown directly into soil samples collected from diff erent fi eld sites.Each experimental pot contained about 3 kg of soil sample.Rhizobial strains were isolated from surface-sterilized nodules following a standard protocol (19).
Genomic DNA from all strains was extracted by using a NucleoSpin ® tissue kit (Macherey-Nagel, Duren, Germany) in accordance with the manufacturer's protocols.Isolates were identifi ed at the species level by 16S rDNA genotyping using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP), while random amplifi cation of polymorphic DNA (RAPD) and enterobacterial repetitive intergenic consensus-poly me rase chain reaction (ERIC-PCR) were used for strain diff erentiation as previously described (20).Identifi cation at species level (16S rDNA PCR-RFLP) included rhizobial strains as follows: Rhizobium tropici 11418, R. phaseoli 30137, R. etli 11541, R. leguminosarum 6039 (Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), Braunschweig, Germany) and the reference strain R. leguminosarum 3622 (Institute of Grassland and Environmental Research, Aberystwyth, UK).
Field trials were set up to investigate the symbiotic properties of common bean rhizobia under the agroecological conditions of central Croatia.The two-year fi eld trials (2008 and 2009) were conducted at two sites (Velika Gorica and Maksimir, Zagreb) positioned 26 km apart.Two-factorial fi eld trials were laid out according to the randomized block design with four replications.Trial factors included: (i) rhizobial strains (R. leguminosarum isolates: S 7/3 , S 8/1 , S 12/5 , S 17/2 and S 21/6 , reference strain R. leguminosarum 3622 and uninoculated control S B0 ), and (ii) common bean cultivars (Trešnjevac and Slavonski zeleni).The size of the experimental plot was 12 m 2 with six rows, leaving 0.50 m between them.The indigenous R. leguminosarum strains and reference strain R. leguminosarum 3622 were used for inoculant preparation.Strains were grown on yeast mannitol broth (YMB) (19) at 28 °C.Cultures were harvested at the late exponential phase of growth and used for preparation of peat-based inoculants.Inoculants were applied to common bean seeds to obtain approx.10 6 cells per seed.The following parameters were recorded to evaluate the symbiotic effi ciency of indigenous and reference strains: nodule number and nodule dry mass (g), total nitrogen content (%) in the above-ground part of plants (21), seed protein content (%) and seed yield (t/ha).All data obtained were processed by analysis of variance (ANOVA) using MSTAT-C statistical program (Michigan State University, East Lansing, MI, USA).

Results and Discussion
The results of greenhouse pot experiment revealed that indigenous common bean rhizobia were not widespread in the studied part of central Croatia.Out of 23 collected soil samples, indigenous rhizobial populations were present only in fi ve.In total, eighteen rhizobial strains were isolated from all soil samples.Soil samples containing common bean rhizobia were collected from diff erent fi eld sites.Previous cultivation of diff erent crops had a signifi cant impact on the occurrence of indigenous strains.Bean was previously grown at three out of fi ve fi eld sites at which the presence of common bean rhizobia was recorded (Table 1).Andrade et al. (22) also deter-mined higher Phaseolus-nodulating rhizobial populations in soils where common bean had previously been grown.Indigenous rhizobial strains were isolated from soil samples that were characterized by neutral or almost neutral pH values.Besides, all soil samples were well supplied with nitrogen, while available phosphorus, potassium and humus content varied signifi cantly (Table 2).Chemical characteristics were determined in all collected soil samples (data not shown).Most soil samples without indigenous rhizobia were also characterized by neutral pH values, while only three soil samples had low pH values.Therefore, our results indicate that soil acidity could not be the most relevant factor in indigenous strain distribution.
Genetic diversity of indigenous rhizobial populations was studied by using diff erent PCR fi ngerprinting methods such as 16S rDNA PCR-RFLP, ERIC-PCR and RAPD.The 16S rRNA gene fulfi ls all of the requirements for a good phylogenetic marker, such as functional conservation, ubiquitous distribution, conserved size and information content, and the presence of evolutionary conserved regions alongside highly variable structural elements (23).
Eighteen rhizobial strains isolated from common bean nodules were compared with reference and/or type strains representing Rhizobium leguminosarum, Rhizobium etli, Rhizobium tropici and Rhizobium phaseoli.Dendrogram derived from PCR-RFLP of 16S rDNA revealed that all common bean isolates diff er signifi cantly from R. tropici, R. etli and R. phaseoli type strains, while they were identical to R. leguminosarum type strain (data not shown).These results suggest that all common bean isolates can be regarded as R. leguminosarum.Low genetic diversity of bean-nodulating rhizobia was also determined in the central region of Chile where R. etli was found to be the dominant species (24).On the contrary, considerably higher diversity within common bean rhizobial populations was determined in other countries, such as Ethiopia, China and Nepal (25)(26)(27).
Further characterization was performed by RAPD and ERIC-PCR, which enable intraspecies diff erentiation.ERIC sequences are widespread in the genome of Gram--negative soil bacteria.This technique is considered a powerful tool for strain diff erentiation in bacterial taxonomy (28).RAPD fi ngerprinting method was also successfully used for assessing genetic diversity of rhizobia (29)(30)(31).
Characteristic RAPD profi les were obtained for each particular rhizobial strain.Dendrogram derived from RAPD fi ngerprints revealed that one isolate (S 21/6 ) diff ered signifi cantly from all the other strains, which were divided into two main clusters (Fig. 1).
Similar results were obtained by the ERIC-PCR method.All isolates were divided into two main clusters, with strains S 12/5 , S 21/4 and S 21/6 being signifi cantly diff erent from all other common bean rhizobia studied (Fig. 2).Comparison of the data obtained by RAPD and ERIC-PCR fi ngerprinting allows the conclusion that the application of both methods showed similar grouping of strains.The obtained results revealed that indigenous rhizobial strains were mostly grouped according to the site of their origin.The only exception was determined for the indigenous strains isolated from the location Sv.Ivan Zelina (S 12 ) and particularly for the isolates from the location Bapča (S 21 ), which were not grouped within the same cluster.However, besides isolates obtained from these two locations, both dendrograms indicate considerable infl uence of biogeographic factors on genetic diversity within indigenous rhizobial population.The results of some other studies also suggest biogeographic structuring of rhizobial populations (32)(33)(34).Two-year fi eld trials were carried out to evaluate the symbiotic properties of indigenous rhizobial strains under the agroecological conditions of central Croatia.The sparse nodulation was determined in uninoculated plants, indicating the presence of some indigenous common bean rhizobia as was also noted in previous reports (35).Rhizobia belong to the group of soil bacteria, and therefore their presence in natural environment is quite common.However, since only sporadic nodulation was noticed at control plots, it is diffi cult to assume that it will have a signifi cant impact on inoculation response.The results showed that the most abundant nodulation was achieved by inoculation with indigenous strains S 17/2 and S 21/6 (Table 3).In most cases, signifi cant diff erences in nodule number and nodule dry mass were not determined.
Poorer nodulation and lower yields were recorded at location Velika Gorica, which was characterized by acidic soil reaction, than at Maksimir, Zagreb.These results confi rm the potential adverse eff ect of soil acidity on nodulation and symbiotic effi ciency of some strains, as described in other papers (36).
In the fi rst year of investigation ( 2008), the highest total nitrogen content was obtained in the above-ground plant parts at both sites by inoculation with indigenous strains S 21/6 and S 17/2 .In 2009, signifi cantly lower nitrogen content was recorded in uninoculated treatments, while statistically signifi cant diff erences between strains were not determined.Inoculation with indigenous rhizobial strains S 17/2 and S 21/6 also resulted in signifi cantly higher protein contents in bean seed than with other indigenous strains and the reference strain.In most cases, no signi fi cant differences in seed yields were determined in depen dence on the strain used.In the second year of investigation, signifi cantly lower yields were obtained in uninoculated treatments at location Maksimir, Zagreb, while signifi cant diff erences between strains were not determined.At location Velika Gorica, in the fi rst year of investigation significantly higher seed yields were achieved with the application of isolate S 8/1 , while in the second year no statistically signifi cant eff ect of inoculation on seed yields was determined.However, the lowest yields were always recorded in the control, uninoculated trial treatments.According to the assessment of the symbiotic effi ciency of indigenous strains, the best results were achieved with the application of strain S 17/2 and the genetically specifi c strain S 21/6 .
Besides, the overall results of the fi eld trials indicate a considerable infl uence of environmental conditions, particularly soil properties, on inoculation response.The weaker eff ect of inoculation at the location Velika Gorica was largely infl uenced by soil acidity.This also confi rms the inhibitory eff ect of high soil acidity on the effi ciency of the symbiotic nitrogen fi xation process.Although some acid-tolerant rhizobia were identifi ed in several studies (37,38), in most cases soil acidity greatly infl uences nodulation and persistence of rhizobia (39).Nodulation was in most cases signifi cantly higher when indigenous rhizobial strains were applied than when they were inoculated with the reference strain.This was particularly expressed under the unfavourable conditions of high soil acidity at the location Velika Gorica, where inoculation with indigenous strains resulted in bett er nodulation than the inoculation with the reference strain.This result confi rms that indigenous rhizobial strains are bett er adaptable to adverse environmental conditions.Benefi cial eff ect of inoculation with indigenous common bean rhizobia was also determined in diff erent studies performed in Brazil (40,41), Tunisia (42) and Spain (43).
Since symbiotic effi ciency of rhizobia greatly depends on the host plant (44), these investigations included two common bean cultivars, Trešnjevac and Slavonski zeleni.Compatibility of both cultivars with all the tested strains was determined.Signifi cant diff erences between common bean cultivars in all parameters were not found, with the ex ception of seed yield.During both years of investigation, signifi cantly higher seed yield of Trešnjevac cultivar was recorded at both locations.
To our knowledge, these are the fi rst studies of indigenous common bean rhizobial populations in Croatia, and the beginning of characterization and selection of indigenous strains.Further studies are needed in order to obtain more data on the composition of rhizobial populations in diff erent regions and to perform more detailed characterization of indigenous common bean rhizobia.

Conclusions
Common bean rhizobia were isolated only from fi ve out of 23 investigated soil samples, showing low abundance of indigenous rhizobia.Presence of only one species (Rhizobium leguminosarum) was confi rmed, although Factor level mean values with diff erent lett ers in superscript are signifi cantly diff erent with an error of p<0.05.N=nodule number per plant; T=Trešnjevac, SZ=Slavonski zeleni considerable level of genetic diversity was found at intraspecies level.The positive impact of inoculation was determined on all studied parameters in comparison with uninoculated controls.The overall results of fi eld trials revealed that two indigenous isolates (S 21/6 and S 17/2 ) were the most effi cient ones and therefore they can be used in further selection programs.Results of fi eld trials clearly confi rm the importance of rhizobial strain selection since signifi cant diff erences in all tested parameters were determined depending on the strain used for inoculation.The obtained data can contribute to a bett er understanding of rhizobial ecology and effi cient selection of rhizobial strains for a particular agroecological area.

Table 1 .
Origin and designation of Rhizobium leguminosarum isolates used in the study

Table 2 .
Chemical properties of soil samples used in the study

Table 3 .
Mean values of nodule number, nodule dry mass and nitrogen content in aerial parts of common bean, seed yield and seed protein content and signifi cance of their diff erences at the locations Maksimir, Zagreb and Velika Gorica