Geographical and climatic distribution of lentil-nodulating rhizobia in Iran

Abstract Lentil is one of the most important legumes cultivated in various provinces of Iran. However, there is limited information about the symbiotic rhizobia of lentils in this country. In this study, molecular identification of lentil-nodulating rhizobia was performed based on 16S–23S rRNA intergenic spacer (IGS) and recA, atpD, glnII, and nodC gene sequencing. Using PCR-RFLP analysis of 16S–23S rRNA IGS, a total of 116 rhizobia isolates were classified into 20 groups, leaving seven strains unclustered. Phylogenetic analysis of representative isolates revealed that the rhizobia strains belonged to Rhizobium leguminosarum and Rhizobium laguerreae, and the distribution of the species is partially related to geographical location. Rhizobium leguminosarum was the dominant species in North Khorasan and Zanjan, while R. laguerreae prevailed in Ardabil and East Azerbaijan. The distribution of the species was also influenced by agroecological climates; R. leguminosarum thrived in cold semiarid climates, whereas R. laguerreae adapted to humid continental climates. Both species exhibited equal dominance in the Mediterranean climate, characterized by warm, dry summers and mild, wet winters, in Lorestan and Kohgiluyeh-Boyer Ahmad provinces.


Introduction
Lentil ( Lens culinaris ), belonging to the genus Lens of the Viceae tribe in the Leguminosae family, plays significant roles in the human diet and soil fertility maintenance.It has high protein and micronutrient contents (Abraham 2015 , Ghanem et al. 2015 ) and impr ov es soil fertility thr ough the addition of nitr ogen (N), carbon, and organic matter (Sarker andErskine 2006 , Abraham 2015 ).The world lentil production reached 5.77 million tons in 2021, with Canada being the major producer contributing up to 1.6 million tons (FAOST A T 2021 ).Lentils are essential to the Iranian household diet, meeting the protein needs of the majority (Hashemzadeh and Monirifar 2016 ).Iran possesses significant potential as a lentil-producing country and ranks ninth globally in terms of cultivated area dedicated to lentil production.In 2021, lentil production in Iran amounted to 79 750 tons, contributing to a world share of 1.07% (FAOST A T 2021 ).Due to the increasing population in Iran, the need to boost legume pr oduction, especiall y lentils, is greater than ever.Traditionally, this has been achieved through the use of chemical fertilizers to enhance production per unit area and expand the cultivation area of legumes.Ho w ever, fr equent dr oughts and land degr adation r esulting fr om excessiv e use of chemical fertilizers and intensive tillage operations have led to a decline in arable land area in Iran.Consequently, there is a pressing need to reevaluate methods for enhancing plant performance (Adesemo y e et al. 2009, Abadi et al. 2020, 2021 ).
Lentils have the capability to establish symbiotic relationships with rhizobia, which can fix atmospheric N and meet the nutritional needs of legumes .T his symbiosis has the potential to have environmental and agronomic benefits by reducing the utilization of chemical N fertilizers in a gricultur e (Riah et al. 2014 ).Integrating lentils into crop rotations with cereal crops promotes sustainable systems .T he symbiotic relationship with rhizobacteria enhances the N content of the system, with reported levels of up to 107 kg ha −1 (Abraham 2015 ).The arid and semiarid zones of Iran face significant limitations for cr op pr oduction, including lo w rainfall, w ater deficits, poor soil fertility, and high soil salinity.These factors have adverse effects on symbiotic associations and N fixation.Ther efor e, the a gr onomic and ecological impacts of rhizobia partnerships depend on their symbiotic properties and their ability to adapt to environmental constraints (Riah et al. 2014 ).Similar to other legumes, lentils r el y on effectiv e rhizobia to fix atmospheric N. Accurate identification of symbiotic rhizobia of lentils is a crucial step in producing rhizobia inoculants.It is necessary to confirm the efficiency of rhizobia strains thr ough gr eenhouse and field experiments, ultimately leading to increased lentil production (Anglade et al. 2015 ).16S rRNA has low phylogenetic power at the species le v el, so accur ate identification of closely related rhizobia strains usually requires phylogenetic analyses based on multilocus sequence analysis (MLSA), including different housek ee ping genes or other techniques such as whole genome sequence analysis (Rashid et al. 2014, Aguilar et al. 2018, Sijilmassi et al. 2020 ).
Specificity is crucial for legume-rhizobia symbiotic associations; typically, a specific rhizobia species or biovar infects only a limited number of host plant species (Riah et al. 2014 ).The species Rhizobium leguminosarum curr entl y comprises thr ee symbiov ars, each associated with specific host plants .T hese symbio vars include phaseoli , trifolii , and viciae , which mediate nodulation in the tribes of Phaseolus , Trifolium , and Viciae , r espectiv el y.Additionall y, symbiovar viciae strains have been reported in Rhizobium fabae and Rhizobium pisi (Rashid et al. 2014 ).Lentils form a symbiotic relationship with R. leguminosarum biovar viciae , enabling atmospheric N fixation.Ho w e v er, r eports on lentil symbiotic Rhizobium vary depending on geographical location and specific regional conditions.
Phylogenetic analyses of housek ee ping genes (16S rRNA, recA , atpD , and glnII ) and nodulation genes ( nodC , nodD , and nodA ) were conducted on 36 bacterial isolates to assess bacterial diversity and identify rhizobia nodulating lentil in Bangladesh.The majority of isolates were associated with Rhizobium etli and R. leguminosarum .Results of phylogenetic analyses indicated that the nodulation genes are linked to R. leguminosarum biovar viciae but form a distinct cluster.MLSA, DNA fingerprinting, and phenotypic characterizations suggested the involvement of at least three clades in lentil nodulation.These clades differ ed fr om the R. etli -R.leguminosarum groups (Rashid et al. 2014 ) and were subsequently identified as three new species: Rhizobium bangladeshense , Rhizobium lentis , and Rhizobium binae .Taha et al. ( 2018 ) conducted a study based on the molecular phylogeny of housek ee ping genes from isolates gather ed fr om 40 cultiv ated fields in the primary lentil pr oduction r egions of Mor occo.Their findings demonstr ated that Rhizobium laguerreae serves as the primary symbiont of cultivated lentils in Morocco.The phylogenetic reconstruction of 26 strains nodulating lentils in Ethiopia, based on recA , atpD , and glnII genes, r e v ealed thr ee distinct sublinea ges (Clades I-III).Genospecies I and II were identified as R. etli and R. leguminosarum , r espectiv el y.Ho w e v er, Genospecies III was suggested to potentially represent an unnamed Rhizobium species (Tena et al. 2017 ).Young et al. ( 2021 ) constructed a phylogeny based on concatenated sequences of 120 universal genes .T hey concluded that the R. leguminosarum species complex comprises 18 distinct genospecies, along with se v en unique strains that are not placed in these genospecies.Among these genospecies, five include the type strains of named species: R. laguerreae , R. sophorae , R. ruizarguesonis , R. indicum , and R. leguminosarum .
No studies have yet investigated the genetic diversity and taxonomic status of lentil rhizobia in Iran using the MLSA a ppr oac h.Pr e vious studies hav e mainl y focused on the impact of symbiosis on lentil gr owth.Ther efor e, the objectiv es of this study wer e: (i) to explore the genetic diversity and population structure of rhizobia nodulating lentils adapted to the environmental conditions of Iran; and (ii) to characterize lentil rhizobial isolates through sequencing analysis of 16S rRNA, protein-encoding housek ee ping genes ( recA , atpD , and glnII ), and the nodulation gene ( nodC ).

Collection of nodules and rhizobia isolation
A total of 80 different lentil farms located in the nine principal provinces of lentil production in Iran (Fars, Lorestan, Kohgiluyeh-Bo y er Ahmad, Qazvin, Ardabil, North Khorasan, Zanjan, East Azerbaijan, and Semnan) were randomly selected.Sampling sites were distributed across humid continental, Mediterranean, cold semiarid, and continental climates (Fig. 1 ).
Nodulated roots and soil samples were collected from lentil farms during the flo w ering sta ge fr om April to Ma y 2018.T hese farms have a history of lentil cultivation and have never been artificially inoculated with rhizobia.The nodulated roots were washed with sterile distilled water, dried with tissue paper, and then pr eserv ed on silica gel until further experiments.Soil samples were utilized for the rhizobia-trapping experiment.T herefore , the lentil rhizobia collection comprised bacteria isolated from nodules collected in the farms and nodules obtained via bacterial tr a pping in the gr eenhouse.
To tr a p rhizobia fr om the soil, lentil seeds (Bilesav ar) underwent surface sterilization by immersion in 96% ethanol for 30 s and then in 0.1% mercury chloride (HgCl 2 ) for 2 min, followed by rinsing 10 times with sterile distilled water.Subsequently, the seeds wer e tr ansferr ed to a gar plates for germination and k e pt at room temper atur e for 48 h.Disinfected plastic pots were filled with various soil samples, and five seedlings were planted in each pot, with three pots allocated for each soil sample .T he plants were cultivated in the greenhouse under controlled conditions at 25 • C under 16 h light/8 h dark cycles for a duration of 2 months .T he plants wer e upr ooted, and the r oots wer e rinsed with sterile distilled water.Nodules were separated from the roots obtained from both the field and the greenhouse culture .T he nodules were sterilized by immersing them in 96% alcohol for 30 s and then in 30% sodium hypochlorite for 2 min, follo w ed b y rinsing ten times with sterile distilled water.For the isolation of rhizobia, a single nodule was crushed in sterile distilled water using a homogenizer under aseptic conditions .T he extr act was str eaked onto yeast-extr act mannitol agar (YEMA) plates containing Congo red (Vincent 1970 ) and incubated at 28 • C for 3-5 da ys .T he isolates were purified by str eaking se v er al times on YEMA plates, and then single colonies were maintained on agar slants at 4 • C until subsequent experiments.

Plant infection assay
To confirm the ability of isolates to form effective nodules on lentils , plant infection assa ys were conducted.Lentil seeds were surface-sterilized and germinated as pr e viousl y described.The seedlings were then transferred to holes created in test tubes (300 mm × 35 mm) filled with Fahraeus agar medium containing 1% (w/v) CaCO 3 .A 1 ml rhizobial suspension ( ∼10 8 cells ml −1 ) was used to inoculate 4-da y-old seedlings .T he tubes were placed in a growth chamber at 25 • C under 16 h light/8 h dark cycles for 4 weeks.During this period, sterile distilled water and Jensen's N-fr ee solution wer e used for plant irrigation.For each isolate, nodulation tests were conducted in three replicates, while negativ e contr ol plants wer e left uninoculated.Finall y, nodule de v elopment was e v aluated.

RFLP analysis of the intergenic spacer gene
Based on the plant infection assay, 116 isolates were obtained, and DNA extraction was performed using the genomic DNA extraction kit (CinnaGen Co., Ltd.Tehran, Iran).The genomic region intergenic spacer (IGS) between the 16S and 23S rRN As w as amplified using primers FGPS1490 and FGPL1320, with the PCR conditions provided in Table 1 .The enzymes BshFI and MspI were used separ atel y to digest the IGS fr a gments at the GG"CC and C"CGG sites, r espectiv el y.Subsequentl y, the r estriction fr a gments wer e separ ated by electr ophor esis in a 2% (w/v) a gar ose gel containing DNA Safe Stain.The IGS restriction patterns were stained, visualized under UV light, and analyzed using Bio-Vision software.The unweighted pair group method with arithmetic av er a ges (UP-GMA) (Nei and Li 1979 ) was utilized for cluster analysis through the NTSYS 2.1 pr ogr am (Rohlf 1998 ).Additionall y, the distribution of Rhizobium populations was statistically compared among and within provinces and four climatic zones by analysis of molecu-Table 1. Primer sequences and PCR conditions used in this study.lar variance (AMOVA) using GenAlex software version 6.5 (Peakall and Smouse 2012).

Amplification and sequencing of genes
T he 16S rRNA gene , along with thr ee c hr omosomal housek ee ping genes ( recA , atpD , and glnII ), and the symbiosis-related gene ( nodC ) of 27 r epr esentativ e str ains fr om the IGS-RFLP gr oups were amplified.PCR conditions and primer sequences used for gene amplification are provided in Table 1 .For sequencing, PCR pr oducts wer e purified using a PCR purification kit (CinnaGen Co., Ltd.) and sequenced by Bioneer Co., South Korea.A total of 135 sequences were generated and deposited in Gen-Bank under the accession numbers: recA (ON454898-ON454924), glnII (ON478264-ON478290), atpD (ON454925-ON454951), nodC (ON454952-ON454978), and 16S rRNA (ON428638-ON428664).The Rhizobial strains were deposited in the Culture Collection of Soil Micr oor ganisms (CCSM) at the Soil and Water Research Institute (SWRI), Iran.

Phylogeny and nucleotide polymorphisms analyses
The sequences obtained wer e c hec ked and assembled with the Vector NTI Advance TM 10 software.Subsequently, each gene sequence was compared with the corresponding genes of reference Rhizobium species in the NCBI database using nucleotide BLAST.
Multiple sequence alignments for all isolates and r efer ence Rhizobium species (Table 2 ) were conducted using the CLUSTAL W pr ogr am fr om MEGA v ersion 11.Phylogenetic tr ees wer e constructed using the maximum likelihood method with the Kimura 2-parameter model in MEGA version 11 (Tamura et al. 2021 ).The robustness of the tree topology was assessed through bootstrap analysis with 1000 replications of each sequence using MEGA 11 software .T he phylogenetic tree of concatenated genes was gener ated fr om the sequence alignments of all isolates and the described r efer ence species.Nucleotide pol ymor phisms, suc h as nucleotide di versity, the n umber of ha plotypes, and ha plotype div ersity, were calculated using DNaSP V6 software (Rozas et al. 2017 ).

Average nucleotide identity
Av er a ge nucleotide identity (ANI) analysis was performed using the ANI calculator provided by ChunLab (Seoul, Korea) ( http:// www.ezbiocloud.net/tools/ ani ) and Kostas Lab ( http://en veomics .ce.gatech.edu/ani).ANI w as determined betw een strains of this study and closely related type species of the genus Rhizobium .

Bacterial isolation
In this study, a total of 218 str ains wer e isolated fr om v arious pr ovinces of Ir an (Table 3 ).The number of rhizobial isolates obtained from each province varied, primarily due to differences Based on the plant infection assay, 116 isolates were capable of forming nodules and were subsequently selected for amplification of the 16S-23S rRNA IGS r egion.The highest percenta ge of rhizobial isolates obtained based on the plant infection assay was 78% from Ardabil pro vince , follo w ed b y Lorestan and Kohgiluyeh-Bo y er Ahmad pro vinces , with nodulation rates of 59% and 56%, respectiv el y.In North Khor asan, Zanjan, and Kohgiluy eh-Bo y er Ahmad pro vinces , the predominant species identified was R. leguminosarum , accounting for 83%, 79%, and 53% of isolates, respectiv el y.Conv ersel y, in Ardabil and Lor estan pr o vinces , R. laguerreae was identified as the dominant species, constituting 61% and 56% of isolates, r espectiv el y.The geogr a phical origin and IGS-RFLP groups of the isolates are listed in Table S1 ( Supporting Information ).

PCR-RFLP of 16S-23S rRNA IGS region
The amplified DNA products ranged in size from 1350 to 1400 bp.RFLP analysis of isolates revealed that BshFI produced two to nine bands ranging in size from 100 to 520 bp, while MspI generated three to nine bands ranging from 100 to 530 bp (Fig. 2 ).Based on the combined patterns obtained from the RFLP results, the 116 isolates were classified into 20 groups and seven unclustered strains (Fig. 2 ; Table S1 , Supporting Information ).Ther efor e, one r epr esentative was selected from each group, and the subsequent steps were carried out with 27 selected isolates.
Four climatic zones (humid continental, Mediterranean, cold semiarid, and continental climates) were analyzed by AMOVA analysis to determine genetic variation among and within populations (Table 4 ).An ov er all significant differ entiation was observed among the four climatic zones, with the variation among populations accounting for 12% of the total variation.With the exception of Semnan pro vince , which had only one rhizobial isolate and ther efor e was not included in the analysis, the remaining provinces underwent AMOVA analysis .T he results sho w ed that 86% of the variation was within pro vinces , while 14% was among pro vinces .Notably, no significant differentiation among provinces in the Mediterranean climate and cold semiarid climate was observed.Ho w ever, differentiation among provinces in the humid continental climate was found to be significant (Table 4 ).

Analysis of 16S rRNA, protein-encoding housekeeping, and nodulation genes
The recA , glnII , atpD , nodC , and 16S rRNA sequences were compared to those of the most closely related bacterial species using the NCBI BLAST pr ogr am and the EzTaxon Database.Subsequently, the sequences were submitted to the GeneBank/NCBI database through BankIt.Accession numbers for 27 isolates se-

Source of v aria tion % Variation a Nm b
Among climates 12 * * 3 .77Among provinces c 14 * 3 .01Among provinces in humid continental climates quenced in this r esearc h ar e av ailable in Table S2 ( Supporting Information ).
Aligned sequences from the recA , glnII , and atpD genes were concatenated, resulting in 1644 bp positions .T he phylogenetic analysis began by comparing the concatenated sequences and the nodC gene obtained in this study with those pr e viousl y published in GenBank.All sequence accession numbers used in phylogenetic anal yses ar e shown in par entheses in Figs 3 and 4 .
The concatenated tree based on recA -glnII-a tpD gene sequences displays two distinct clades (Fig. 3 ).Clade 1 consists of 13 isolates, whic h shar e similarity with str ains suc h as R. laguerreae FB206T, R. laguerreae FB14022, and R. laguerreae FB310 from Tunisia, as well as R. laguerreae WSM1455 fr om Gr eece .Con v ersel y, Clade 2 encompasses 14 isolates that exhibit similarity to related genes found in R. leguminosarum bv.viciae 248 from UK, R. leguminosarum GLR17 fr om German y, R. leguminosarum CCBAU 65264 fr om China, and R. leguminosarum GLR19 from Bangladesh.
The ANI values of the concatenated sequences of rhizobial strains in Clade 1 and R. laguerreae FB206 T ranged from 95.4% to 96%.Ho w e v er, the ANI percenta ges between str ains in clade 2 and R. laguerreae FB206 T w ere betw een 91.3% and 92.6%.Conv ersel y, Figur e 3. T he concatenated dendr ogr am was constructed based on the recA , glnII , and atpD gene sequences of the selected rhizobia strains .T he bootstr a p test was performed with 1000 replications, and values higher than 70% were written on the relevant branches.
The ANI values of the concatenated sequences of rhizobial strains in Clade 2 and R. leguminosarum bv.viciae 248, as well as R. leguminosarum GLR17 exceeded 95%.Ho w e v er, the ANI percenta ges between strains in Clade 1 and these two species w ere lo w er than 95%.
Based on the nodC gene sequences, the studied strains are categorized in three distinct clades (Fig. 4 ).Clade 1 comprises 14 isolates, which cluster together with similar strains of R. leguminosarum from Syria.Clade 2 consists three isolates with a high Bootstr a p v alue (91%), gr ouping with similar str ains of R.
fabae CCBAU 33202 T and R. laguerreae FB206 T from China and Tunisia, r espectiv el y.Clade 3 encompasses 10 isolates with a robust Bootstr a p v alue (99%) along with similar strains of R. lentis BLR27 T and R. bangladeshense BLR175 T from Bangladesh, R. laguerreae PEPV11 and R. laguerreae MLSC04 from Spain, and R. laguerreae LMR614 fr om Mor occo.Based on the nodC gene sequence, strains belonging to the same species were not consistently grouped within the same clade .T his observation suggests that the nodC gene region may not have the capability to effectively differentiate between R. leguminosarum and R. laguerreae , possibly due Figur e 4. T he maximum-likelihood tree based on the nodC gene sequences of the selected rhizobia strains .T he bootstrap test was performed with 1000 replications, and values higher than 70% were written on the relevant branches.
to its location on plasmids and susceptibility to horizontal gene transfer.

Discussion
16S rRNA PCR amplification is one of the best techniques for the taxonomic determination of large databases of bacteria.Ho w e v er, it is inadequate for effective identification and discrimination of Rhizobium species.Contrary to the slow rate of evolution of the 16S rRNA gene, which does not allow for proper identification of closel y r elated species, housek ee ping genes pr ovide a better r esolution to differentiate Rhizobium species.
Based on the sequence alignment of the 16S rRNA gene and ANI values, it was observed that the sequences of our strains and the R. leguminosarum species complex (Rlc) wer e highl y similar, with ANI values ranging between 99.5% and 100%.Therefor the 16S rRNA gene was not a suitable region for identifying and distinguishing the genospecies.Pr e vious studies hav e also demonstrated that identical 16S rRNA sequences can be shared by Rhizobium species like R. laguerreae and R. leguminosarum (Taha et al. 2018 ).Young et al. ( 2021 ) further noted that the 16S ribosomal RNA sequence tends to be ov erl y conserv ed, limiting its discriminatory po w er among closely related species.Notably, the full-length 16S sequences of all five type strains within the Rlc-R.leguminosarum , R. laguerreae , R. sophorae , R. ruizarguesonis , and R. indicum -were found to be identical.Even the sequence of R. anhuiense , and that of the mor e distantl y r elated R. acidisoli , pr ov ed identical.Ther e-fore, recA , atpD , and glnII genes were utilized for molecular identification of lentil rhizobia.
In this study, 116 isolates were obtained for PCR-RFLP analysis of the 16S-23S rRNA IGS based on the plant infection assa y.T his technique has been widely utilized to explore the diversity within rhizobia strains.In our study, 116 isolates from nine pr ovinces wer e classified into 20 groups , lea ving se v en str ains unclustered utilizing BshFI and MspI restriction enzymes.Our study unveiled an intermediate level of chromosomal diversity.Palmer and Young ( 2000 ) identified 25 genotypes out of 285 isolates using HaeIII, considering it within the typical v ariation r ange for R. leguminosarum .They conclude that rhizobial diversity can be influenced by differences between two mana gement r egimens in arable and grass sites .Con versely, Depret and Laguerre ( 2008 ) detected 28 haplotypes among 1100 isolates from a single site in France using HaeIII.Lu et al. ( 2009 ) reported 33 clusters or single strains among 174 rhizobia strains in three ecological regions of China using MspI, HhaI, and HaeIII.In contr ast, Mutc h et al. ( 2003) found only nine different types among 625 isolates using TaqI.
Based on the AMOVA a ppr oac h, our study found significant genetic variation among provinces and four climatic zones in rhizobia strains associated with lentils .T his finding is consistent with the observations of Riah et al. ( 2014 ), who conducted a similar study using HaeIII (an isoschizomer of BshF I) for IGS RFLP analysis of rhizobia populations associated with lentils and peas in Eastern Algeria.In their study, significant differentiation in the distribution of IGS haplotypes was observed between populations fr om differ ent ecoclimatic zones, namel y subhumid and semiarid zones.Ad ditionally, the y found significant differentiation between sites within each ecoclimatic zone.In contrast to the findings of Riah et al. ( 2014 ), our study r e v ealed significant differ entiation among provinces in the humid continental climate, while differentiation among provinces in the Mediterranean climate and cold semiarid climate was not significant.These variations in findings may reflect differences in the specific environmental conditions, a gricultur al pr actices, and genetic div ersity of rhizobia populations across different geographic regions and ecoclimatic zones.
The comparison of genetic diversity within and among populations across different climates in our study revealed that the majority (88%) of the diversity exists within populations, while the amount of genetic diversity among populations is relatively lo w.This lo w le v el of genetic div ersity among populations can be attributed to the high degree of gene flo w, which w as calculated to be 3.77.Wright ( 1978 ) classified gene flow (Nm) into three categories: high ( ≥1.0), medium (0.250-0.99), and low (0.0-0.249).According to this classification, gene flow between populations is considered to occur when Nm ≥ 1.In our study, the data indicate a high le v el of gene flow among the populations .T he observed high gene flow among populations could be due to the absence of physical barriers that limit gene flow, facilitating the movement of rhizobia str ains acr oss differ ent r egions (Muthini et al. 2014 ).Human activities, such as the transfer of plants , soils , and the circulation of lentil seeds through trading within the region, may have also contributed to the maintenance of genetic diversity among rhizobia populations in Iran.These findings align with those of Elboutahiri et al. ( 2010 ), who reported a higher proportion of significant genetic variation distributed within regions (89%) than among regions (11%) in Sinorhizobium meliloti and S. medicae obtained from drought-and salt-affected regions of Morocco.
In this study, we observed that R. leguminosarum and R. laguerreae were present in most provinces of Iran, including North Khorasan, Ardabil, Zanjan, Kohgiluyeh-Boyer Ahmad, Lorestan, and Fars, albeit with varying frequencies.For instance, in North Khorasan and Zanjan pro vinces , 83% and 79% of the isolates, respectiv el y, belonged to R. leguminosarum .Conv ersel y, in Ardabil and East Azerbaijan pro vinces , the dominant species was R. laguerreae , with 61% and 100% of the isolates belonging to this species, r espectiv el y.In some provinces such as Lorestan, Fars, and Kohgiluy eh-Bo y er Ahmad, both species w er e observ ed with almost equal fr equency.Geogr a phical location a ppears to play a role in the distribution of these species.For example, North Khor asan pr o vince , wher e most isolates wer e r elated to R. leguminosarum , is situated in the northeast of Ir an.Conv ersel y, R. laguerreae was dominant in East Azerbaijan and Ardabil pro vinces , located in the northwest of Iran.In the western and southwestern parts of Iran, including Lorestan and Kohgiluy eh-Bo y er Ahmad pro vinces , both species were present almost equally.It is noteworthy that in Lorestan and Ardabil pro vinces , both species were isolated from the same soil, indicating their coexistence within these regions.
Pr e vious studies have revealed variability in the main symbiotic bacteria of L. culinaris (lentils) across different regions.In several countries such as West Asia-North Africa, Algeria, Canada, and France, R. leguminosarum has been identified as the primary symbiont of lentils (Hynes and O'Connell 1990, Moawad and Beck 1991, La guerr e et al. 1992, Tian et al. 2010 ).Ho w e v er, r esearc h conducted in Morocco by Taha et al. ( 2018 ) demonstrated that the dominant species associated with lentils was R. laguerreae , indicating r egional v ariation in symbiotic partners.Similarl y, inv es-tigations in Bangladesh, utilizing multiple genes including 16S rRNA, glnII , atpD , recA , and nodulation genes nodA , nodC , and nodD , r e v ealed genetic similarity with R. etli and R. leguminosarum species (Rashid et al. 2012 ).Subsequent studies by Rashid et al. ( 2015 ) identified additional symbiotic species including R. lentis , R. bangladeshense , and R. binae in Bangladesh.Mor eov er, Dhaoui et al. ( 2016 ) reported the presence of various endosymbiotic bacteria of L. culinaris in Tunisia, including R. leguminosarum symbiovar trifolii , Ensifer numidicus , and Mesorhizobium amorphae symbiovar cicero , highlighting the diversity of symbiotic associations in lentil crops across different geographic regions .T hese findings underscore the importance of understanding r egional v ariations in symbiotic relationships for effective lentil cultivation and crop mana gement str ategies.
In our study, the incongruence observed in the phylogenetic tr ees constructed fr om individual pr otein-coding housek ee ping genes ( recA and atpD ) and the concatenated sequences of the recA -glnII-atpD gene sequences underscores the complexity of evolutionary relationships within lentil rhizobia populations .T his phenomenon is not uncommon in phylogenetic studies, as different genes may e volv e at differ ent r ates or be subject to different selectiv e pr essur es, leading to discordant tr ee topologies.Ho w e v er, the consistent tree topology observed between the glnII gene sequences and concatenated sequences (data not shown).This is in line with pr e vious findings by Ribeiro et al. ( 2013 ) and highlights the importance of considering multiple genes for accurate phylogenetic inference .T he study by Young et al. ( 2021 ) further supports the notion that partial sequences of housek ee ping genes ma y not alwa ys reflect the true phylogenetic relationships among str ains, especiall y when alleles within a genospecies do not form distinct clades .T her efor e, r el ying solel y on single genes for phylogenetic analysis may lead to incomplete or inaccurate interpretations of evolutionary history.The phylogenetic analysis of the concatenated recA -glnII-atpD gene sequences r e v ealed distinct clades corresponding to different species.Isolates within clade 1 sho w ed similarity to R. laguerreae isolates from Tunisia and Greece, suggesting a close evolutionary relationship.Conversely, isolates belonging to R. leguminosarum formed a distinct clade along with isolates from UK and Germany, indicating distinct species boundaries and evolutionary lineages.
The genomic evidence presented by Young et al. ( 2021 ) highlights the considerable diversity within the Rlc, indicating that it comprises multiple distinct species.By constructing a phylogeny based on concatenated sequences of 120 universal genes and calculating pairwise ANI between all genomes, they delineated 18 distinct genospecies within the Rlc.Ad ditionally, the y identified se v en unique strains that do not fit into any of these genospecies.Among the 18 distinct genospecies identified, eight of them include the type strains of named species: R. laguerreae , R. sophorae , R. ruizarguesonis , R. indicum , R. brockwellii , R. johnstonii , R. beringeri , and R. leguminosarum itself (Young et al. 2023 ).
The Genome Taxonomy Database employs a 95% ANI threshold to define species and aims to incor por ate and classify all available genomes (Parks et al. 2020 ).In our study, ANI analysis revealed that our str ains ar e r elated to two species within the Rhizobium genus: R. leguminosarum and R .laguerreae .The ANI values obtained for the concatenated sequences of rhizobial strains in Clade 1 and R. laguerreae FB206 T r anged fr om 95.4% to 96%, indicating a close relationship between these str ains.Similarl y, the ANI v alues of the concatenated sequences of rhizobial strains in Clade 2 and R. leguminosarum bv.viciae 248, as well as R. leguminosarum GLR17, exceeded 95%, suggesting their affiliation with R. leguminosarum species.Comparing ANI values between our strains and thr ee ne wl y intr oduced species ( R. broc kwellii CC275e T , R. johnstonii 3841 T , and R. beringeri SM51 T ) r e v ealed ANI v alues r anging between 73% and 79%, indicating our strains are distinct from these species .Moreo ver, the ANI values of our strains with R. ruizarguesonis UPM1133 T , which is closely related to Clade 1 in Fig. 3 , ranged between 91% and 93%.
In conclusion, the findings of this study pr ovide v aluable insights into the nodulation patterns and genetic diversity of lentilnodulating rhizobia across various provinces of Iran.It was observed that L. culinaris could nodulate with both R. leguminosarum and R. laguerreae in different climatic zones.Specifically, R. leguminosarum a ppear ed to be more compatible in semiarid climates, such as those found in North Khorasan and Ghazvin pro vinces , while R. laguerreae sho w ed a higher compatibility in humid continental climates, as seen in Ardabil and East Azerbaijan provinces.Inter estingl y, Lor estan and Kohgiluy eh-Bo y er Ahmad provinces, c har acterized by a Mediterranean climate, exhibited the presence of both species with almost equal frequency.This suggests that the climatic conditions in these provinces may support the growth and nodulation of both R. leguminosarum and R. laguerreae .Howe v er, the study also underscores the need for further research to explore the genetic diversity of lentil-nodulating rhizobia across a wider range of climatic conditions.Understanding the genetic diversity and adaptation of rhizobia to different environments is crucial for optimizing lentil production and enhancing symbiotic N-fixation in a gricultur al systems.In summary, the present study sheds light on the nodulation dynamics of lentil with R. leguminosarum and R. laguerreae across diverse climatic zones in Iran and highlights avenues for future research to deepen our understanding of rhizobial diversity and adaptation in legume crops.

SequenceFigure 1 .
Figure 1.Map of Iran showing sampling sites and climatic regions.

Figure 2 .
Figure 2. Genetic relationships among the rhizobia isolated from L. culinaris by the IGS PCR-RFLP analysis .T he dendrogram was constructed using the UPGMA method.

Table 4 .
Genetic differentiation of the rhizobial populations according to their geogr a phical origins and climates by AMOVA based on PCR-RFLP of 16S-23S rRNA IGS.

Table 2 .
Refer ence/type str ains used in this study.

Table 3 .
The number of isolated rhizobia in each pro vince . of soil samples collected.For instance, Ardabil pro vince , which boasts the largest area of lentil cultivation in Iran, had a greater number of soil samples collected, consequently resulting in a higher number of rhizobial isolates obtained from this region.In terms of climate, Ardabil, East Azerbaijan, and Semnan pro vinces , characterized by a humid continental climate, yielded 67 isolates.On the other hand, Lorestan, Kohgiluyeh-Bo y er Ahmad, and Fars pro vinces , featuring a Mediterranean climate with warm, dry summers and mild, wet winters, yielded 74 isolates.
L: length of the sequences; h: number of haplotypes, Hd: haplotype (gene) diversity; S: number of polymorphic (segregating) per sites; and π: nucleotide diversity.