As an important leguminous food crop, the seeds of the common bean (Phaseolus vulgaris L.) contain rich nutrients, especially protein and starch. Compared with expensive animal-derived food, common beans are widely consumed in many countries as a cheaper source of protein (Los et al., 2018; Shamseldin and Velázquez, 2020). This legume plant is believed to originate and first domesticated in Mexico and the Andes Mountains, and then, it was widely distributed/planted all over the world (De Ron et al., 2019). During the long history of its cultivation, several secondary diversification centers of the common bean have developed, such as Europe (Rodiño et al., 2003, 2006), Africa (Martin and Adams, 1987; Khairallah et al., 1990), Asia (Zhang et al., 2008), and South America (Maciel et al., 2003). The common bean has been cultivated for more than 400 years in China (Zhang et al., 2008), covering various agricultural areas, including the provinces of Heilongjiang and Guizhou (Yeh et al., 1997). Similar to many other legumes, the common bean can fix nitrogen in symbiosis (root nodules) with diverse rhizobia (Shamseldin and Velázquez, 2020). However, the symbiotic nitrogen fixation capacity of the common bean is lower than other major crop legumes (Herridge et al., 2008), but this capacity may vary depending on the compatibility between the symbiosis partners and their adaption to the local geographic environments (Efstathiadou et al., 2021).

Growing in different geographic regions, common bean plants can nodulate with diverse rhizobia, including species in the genera Rhizobium, Pararhizobium, Sinorhizobium, and Bradyrhizobium in the class of Alphaproteobacteria and in the genera Paraburkholderia and Cupriavidus in the class Betaproteobacteria; so, it was defined as a promiscuous host for rhizobia (Shamseldin and Velázquez, 2020). It has been estimated that the symbiosis promiscuity could be a result of the interactions among the rhizobia, host plants, and the environmental factors in their habitats (Cao et al., 2014; Zhang et al., 2018, 2020) because both the hosts and the microsymbionts have to adapt the local conditions. In this case, common bean plants grown in different regions have to select the most adapted/compatible species and symbiovars of rhizobia for their habitats. Up to now, eight symbiovars have been found among the common bean-nodulating rhizobia, including phaseoli, tropici, gallicum, mimosa, orientale, giardini, mediterranense, and unnamed (Efstathiadou et al., 2021). Symbiovar phaseoli is the most common one found in common bean rhizobia (Shamseldin and Velázquez, 2020), and it is one of the symbiovars mainly contained by rhizobia nodulating with the common bean in its center of origin. Symbiovar phaseoli has been found in Asia (Huo et al., 2019), Europe (García-Fraile et al., 2010), and Africa (Mhamdi et al., 2002). Accompanying the distribution of the common bean in the world, Rhizobium etli in the symbiovar phaseoli also disseminated by attaching to the seeds (Pérez-Ramírez et al., 1998), and then, the symbiosis genes of this symbiovar might have been acquired by local bacterial strains through horizontal gene transfer (HGT) (Rajnovic et al., 2019; Tong et al., 2020). Previously, rhizobia associated with the common bean have been classified in Bradyrhizobium diazoefficiens, four unclassified Bradyrhizobium spp., R. leguminosarum, R. etli complex (with 4 subgroups), R. phaseoli, R. vallis, and four unnamed Rhizobium genospecies in the northern region of China (Cao et al., 2014), with Rhizobium sp. IV (R. chutanense, Huo et al., 2019) as the most abundant genospecies. In Shaanxi province (Loess Plateau), common bean formed nodules with four novel Rhizobium genospecies, R. phaseoli, Pararhizobium giardinii, Sinorhizobium fredii, and S. kummerowiae and two novel genospecies of Sinorhizobium and Bradyrhizobium liaoningense (Wang et al., 2016). These studies demonstrated that different rhizobial communities have developed in association with the common bean in distinct Chinese regions. Meanwhile, these studies and some others also revealed that at least three symbiovars with quite different nodC types existed in these common bean-nodulating rhizobia, in which symbiovar phaseoli was the main one found in various species (genera) (Wang et al., 2011, 2016; Cao et al., 2014; Huo et al., 2019).

Blacklands with “chernozem” soil in the world are known as the most productive regions for agriculture in relation to their high fertility and rich in humus, mainly located in North America, east Europe, and northeastern China. In these three regions, the black soil in China has characteristics distinct from those in the other two regions in relation to the difference in soil formation procedure. The Chinese black soil was developed from steppe meadow and characterized by the absence of CaCO₃ but is rich in Fe-Mn nuclei and SiO₂ powder, while CaCO₃ is rich in the American and European black soils (https://www.sohu.com/a/515314435_100941). Based on these characteristics, the Chinese blackland formed a special ecosystem, which might exist in unique microbial communities adapting to the local conditions. The blackland in China is located in the connecting area between Heilongjiang province, Jilin province, and Inner Mongolia Autonomous Region, where maize, rice, and soybean are the main crops, and the common bean is also widely cultured as a vegetable crop. Previously, several rhizobial strains associated with the common bean in an agricultural-forestry ecosystem in Heilongjiang were classified as B. japonicum, R. etli, and R. leguminosarum, which harbored nodC genes corresponding to those of soybean nodulating Bradyrhizobium and symbiovar phaseoli (Wang et al., 2009). However, the rhizobial community associated with the common bean in the black soil in China has not been studied.

Considering the biogeographic patterns and the great diversity of common bean rhizobia uncovered previously, it could be estimated that a unique rhizobial community might exist in the Chinese black soils. To verify this estimation and investigate the compatibility of the native rhizobia with different common bean varieties, rhizobia isolated from two cultivars of the common bean grown in black soils at two sampling sites in Heilongjiang province were characterized in this study. The results evidenced that rhizobia in black soils formed a unique community characterized by the dominance of three unnamed Rhizobium genospecies belonging to symbiovar phaseoli.

This study was performed in order to learn the diversity of common bean rhizobia in the black soils in China and their symbiosis compatibility with the host varieties. Based on phylogenetic analyses of housekeeping genes (16S rRNA, atpD, and recA) and the symbiosis genes (nodC and nifH), 79 rhizobial strains isolated from two soil samples in association with a local cultivar and a widely cultured cultivar were classified into six genospecies, including three defined species and three novel genospecies (Table 1; Figures 1–3). This classification was verified by the genome analysis for the three novel species. Among these genomic species, the three unclassified genospecies Rhizobium sp. I (63.29%), Rhizobium sp. II (12.66%), and Rhizobium sp. III (15.19%) were dominant, while R. indigoferae, R. anhuiense, and R. croatiense had relative abundances of 3.80, 1.26, and 3.80%, respectively. This community composition was unique for the studied black soils in China compared with those reported in other studies. For example, common bean nodulated (1) with R. etli, R. leguminosarum, and B. japonicum in a forest region in Heilongjiang, which also has acid soil with pH 5.3–5.5 (Wang et al. 2009); (2) with R. leguminosarum bv. phaseoli in neutral and alkaline soils and Bradyrhizobium sp. in acidic soils (pH < 6.0) of the subtropical region of China (Han et al., 2005); (3) with Sinorhizobium meliloti sv. mediterranense and Sinorhizobium americanum in alkaline soils (Mnasri et al., 2007; Verástegui-Valdés et al., 2014), and Rhizobium acidisoli/Rhizobium hidalgoense in acidic soils in other countries (Verástegui-Valdés et al., 2014) and so on. This unique rhizobial community also differed from those mentioned in the introduction (Cao et al., 2014; Wang et al., 2016; Huo et al., 2019). Thus, it could be estimated that common bean plants may have selected some indigenous bacteria as their microsymbionts in the studied black soils. In this study, three dominant unnamed genospecies and R. indigoferae were described as microsymbionts of the common bean for the first time, since their nodulation ability on the common bean was verified by the nodulation test with the representative strains and/or by the nodC gene phylogeny. Previously, the phylogeny of nodC and nodA has been used to determine the host spectra and symbiovar of rhizobia, and all the strains belonging to symbiovar phaseoli could nodulate common bean (Efstathiadou et al., 2021; Gunnabo et al., 2021). However, the findings in this study further enlarged the spectrum of common bean-nodulating rhizobia and added novel information about the interactions among the soil factors, rhizobial species, and host plants.

In this study, 16S rRNA, recA, and atpD genes, as well as the whole genome sequences, were used to define the species, since they have been suggested as valuable biomarkers for differentiating rhizobial species (Zhang et al., 2018, 2020; Tong et al., 2020). In most cases, the grouping results in the phylogenetic trees constructed with the individual genes were similar and were well supported by the whole genome phylogeny (Figures 1–3). However, some discrepancies were observed between the phylogenetic trees of recA and atpD genes. First, the recA gene was more sensitive to estimating the genomic diversity than the atpD gene, since more genotypes were identified by recA, which gave higher taxonomic resolution. For example, R. indigoferae and R. anhuiense shared the same sequences in the 16S rRNA gene but they could be distinguished by recA sequence analysis. Thus, the results in this study and the previous study demonstrated that < 97% of recA gene sequences could be used to primitively distinguish species (Wang et al., 2019). Second, three strains MJ39, MC77, and MC79 in genospecies VI shared identical recA gene sequences with the type strain of R. croatiense, but their atpD genes were identical with that of R. changzhiense type strain. Based on the whole-genome sequencing analysis, these three strains were identified as R. croatiense. Thus, it could be estimated that the atpD gene of these three strains was acquired from R. changzhiense by lateral transfer. This case remained us that it is better to define rhizobial species based on the analysis of multilocus genes.

As mentioned above, the symbiosis promiscuous property of the common bean could be seen from both the species level and the symbiovar level. Previously, eight symbiovars have been reported across all the common bean-nodulating rhizobia in different species and genera (Efstathiadou et al., 2021), while the horizontal transfer of symbiosis genes across species and genera has been evidenced or estimated (Tong et al., 2020). In this study, all of them were identified into one symbiovar, e.g., sv. phaseoli (Figures 4 and Supplementary Figure S4), which might demonstrate that (1) horizontal transfer of symbiosis genes might have happened in the local rhizobial communities; and (2) the symbiosis genes of sv. phaseoli might have been introduced from the center of origin of the common bean, along with the plant seeds (Pérez-Ramírez et al., 1998; Tong et al., 2020). Moreover, this may support the hypothesis that the common bean was introduced into China directly from America, and China may be the secondary diversity center of the common bean (Zhang et al., 2008).

In this study, nifH and nodC genes were amplified from most of the strains, but three of them failed for amplification of both genes or nodC, which might be due to primer mismatching or poor DNA quality (De Castro et al., 2018). It is also possible that there are no symbiotic endophytes in nodules since no symbiotic rhizobial strains have been found in soils (Soenens and Imperial, 2018) and root nodules (Moyano et al., 2017). In this experiment, the genotypes of the nifH gene and nodC gene were inconsistent, and the number of nifH genotypes was more than that of nodC genotypes. The phenomenon that the nifH gene was not synchronized with the nodC gene and evolves independently also conforms to the hypothesis that the origin of nitrogen fixation is earlier than nodulation, and the distribution of the nifH gene is more extensive (Dos Santos et al., 2012).

Another interesting finding in this study is that the compatibility between common bean cultivars and the rhizobial species might be a feature related to the soil conditions. It seems both the varieties used in this experiment are capable to nodulate all six rhizobial species, but they showed different affiliations with Rhizobium sp. I and Rhizobium sp. II strains in soils from Binxian but not in the soil of Muling (Table 1). In Binxian soil, the local cv. Cuican Rhizobium sp. I did not show preference between the two rhizobial genospecies (10:8), but the widely cultured cv. Jiadouwang preferred sp. I more than sp. II (16:2). Therefore, the symbiosis compatibility between legume and rhizobia might be determined by the genetic backgrounds of the host plant and rhizobial species (Wu et al., 2020; Jiao et al., 2022) but also regulated by the soil conditions, including both the physiochemical features and the local microbiome in soil (Zhang et al., 2018, 2020). In this case, the symbiosis compatibility might also be related to the biogeography or adaptation (Zhang et al., 2018, 2020) and nodulation competition (Jia et al., 2013) of rhizobia.

Previously, soil pH has been reported as a key factor to determine the geographic distribution or nodule occupation of rhizobial species, such as soybean, which mainly nodulate with Bradyrhizobium species in acidic soils and with Sinorhizobium in alkaline–saline soils (Zhang et al., 2011). A similar situation was also observed in common bean rhizobia (Han et al., 2005; Mnasri et al., 2007; Wang et al., 2009, 2016; Cao et al., 2014; Verástegui-Valdés et al., 2014; Huo et al., 2019). In this study, a clear difference in the geographic distribution of rhizobia was observed on Rhizobium sp. II and Rhizobium sp. III: the former was only in Binxian soil as the second abundant group; while the latter was only in Muling soil as the second abundant group. It was possible that Rhizobium sp. II was more adapted to acidic soil with lower P, while Rhizobium sp. III was contrast (Table 1). The soils' pH values in the studies of Wang et al. (2009) and Verástegui-Valdés et al. (2014) were similar to that of the soil from Binxian in this study (around pH 5.5), but the rhizobial communities were quite different from each other; therefore, it can be estimated that soil pH was not the only determinant for rhizobial biogeography (Zhang et al., 2018). The adaptation of rhizobial species to the whole environment and their interaction with other native microbes might also play an important role in the biogeography and compatibility with the host legumes.

In conclusion, a unique rhizobial community associated with the common bean exists in the studied black soils, which was dominated by three unnamed Rhizobium genospecies and three minor species corresponding to R. indigoferae, R. anhuiense, and R. croatiense. All of them belonged to the symbiovar phaseoli, evidencing that they have coevolved with the host plant under the double selection from soil conditions (for species) and host specificity (for symbiosis genes). In addition, symbiosis compatibility between common bean cultivars and Rhizobium species might be altered by soil characteristics. These findings further enlarged the spectrum of common bean-nodulating rhizobia and added novel information about the interactions among the soil factors, rhizobial species, and host plants in the symbiosis.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

HW was responsible for the design of the experimental scheme. ZW was responsible for the isolation of strains, gene amplification, and paper writing. LL was responsible for the sequencing of the genome. DH was responsible for the potted cultivation of samples. EW was responsible for the modification of the paper. CG was responsible for the analysis of data. All authors contributed to the article and approved the submitted version.