Analysis of 1,000+ Type-Strain Genomes Substantially Improves Taxonomic Classification of Alphaproteobacteria

Abstract

The class Alphaproteobacteria is comprised of a diverse assemblage of Gram-negative bacteria that includes organisms of varying morphologies, physiologies and habitat preferences many of which are of clinical and ecological importance. Alphaproteobacteria classification has proved to be difficult, not least when taxonomic decisions rested heavily on a limited number of phenotypic features and interpretation of poorly resolved 16S rRNA gene trees. Despite progress in recent years regarding the classification of bacteria assigned to the class, there remains a need to further clarify taxonomic relationships. Here, draft genome sequences of a collection of genomes of more than 1000 Alphaproteobacteria and outgroup type strains were used to infer phylogenetic trees from genome-scale data using the principles drawn from phylogenetic systematics. The majority of taxa were found to be monophyletic but several orders, families and genera, including taxa recognized as problematic long ago but also quite recent taxa, as well as a few species were shown to be in need of revision. According proposals are made for the recognition of new orders, families and genera, as well as the transfer of a variety of species to other genera and of a variety of genera to other families. In addition, emended descriptions are given for many species mainly involving information on DNA G+C content and (approximate) genome size, both of which are confirmed as valuable taxonomic markers. Similarly, analysis of the gene content was shown to provide valuable taxonomic insights in the class. Significant incongruities between 16S rRNA gene and whole genome trees were not found in the class. The incongruities that became obvious when comparing the results of the present study with existing classifications appeared to be caused mainly by insufficiently resolved 16S rRNA gene trees or incomplete taxon sampling. Another probable cause of misclassifications in the past is the partially low overall fit of phenotypic characters to the sequence-based tree. Even though a significant degree of phylogenetic conservation was detected in all characters investigated, the overall fit to the tree varied considerably.

Introduction

The class Alphaproteobacteria is a diverse group of bacteria that is taxonomically assigned to the phylum Proteobacteria (Garrity et al., 2005a). At the time of writing the class comprises more than a dozen orders with validly published names. Alphaproteobacteria are cosmopolitan and colonize a wide range of habitats including soil, pelagic and benthic regions of the ocean, fresh water, and lichens. Frequently Alphaproteobacteria account for one of the most active and numerically dominant taxon of microbial communities (Brinkhoff et al., 2008; Bates et al., 2011; Schiaffino et al., 2016). The variety of habitats is illustrated by Rhodobacteraceae which is predominantly marine as for genera such as Oceanicella (Albuquerque et al., 2012) but also includes genera such as Pannonibacter (Borsodi et al., 2003; Biebl et al., 2007), which is found in lakes, and Ketogulonigenium (Urbance et al., 2001; Simon et al., 2017), found in soil.

Although the vast majority of Alphaproteobacteria are free-living, this class does include representatives associated with a broad range of hosts. Rhizobium, for example, establishes endosymbiotic nitrogen-fixing associations with roots of legumes (Pini et al., 2011). These bacteria are key players in the nitrogen turnover and have an important role in agriculture because they act as a natural fertilizer for plants (Fox et al., 2007) and for bioremediation and mineralization of industrial pollutants (Siddavattam et al., 2011). Other kinds of symbiosis are also established, such as the one between Silicibacter and marine phytoplankton (Belas et al., 2009). Wolbachia includes endosymbionts of arthropods (Hedges et al., 2008). Their host interactions are often complex and in some cases have evolved into a mutualistic rather than parasitic relationship (Hosokawa et al., 2010; Nikoh et al., 2014) giving their hosts resistance to viral infections (Teixeira et al., 2008). Other Alphaproteobacteria, such as Bartonella (Strong et al., 1915; Brenner et al., 1993; Birtles et al., 1995) and Brucella (Verger et al., 1985; Meyer and Shaw, 1920), are obligate intracellular parasites. Genera like Rickettsia can trigger serious diseases in plants, animals and humans (Fournier et al., 2000; Luis-Pantoja et al., 2015; Maina et al., 2016). Alphaproteobacteria also harbours opportunistic human pathogens such as Roseomonas (Rihs et al., 1993; Sánchez-Porro et al., 2009; Venkata Ramana et al., 2010).

Alphaproteobacteria are metabolically diverse, too. Most representatives of the class are chemoorganoheterotrophs but many others perform anoxygenic photosynthesis (Brinkmann et al., 2018), including the families Rhodobacteraceae or Rhodospirillaceae, the so-called purple non-sulfur bacteria (Imhoff et al., 1998). The phototrophic genera include Porphyrobacter (Fuerst et al., 1993; Coil et al., 2015), Roseobacter (Shiba, 1991; Martens et al., 2006), and Rhodobacter (Imhoff et al., 1984; Srinivas et al., 2007b; Wang et al., 2014). bacteriochlorophyll α and carotenoids are mostly present in phototrophic bacteria but can also be found in non-phototrophic bacteria like Roseibium (Zhong et al., 2014). While photoorganoheterotrophy is found in Rhodovulum (Hiraishi and Ueda, 1994a) and Phaeospirillum (Imhoff et al., 1998), chemolithoorganotrophy is present in Elioraea (Albuquerque et al., 2008) and facultative methylotrophy in Methylarcula (Doronina et al., 2000). Magnetospirillum (Schleifer et al., 1991), Magnetococcus (Bazylinski et al., 2013a) and Magnetovibrio (Bazylinski et al., 2013b) contain tiny chains of magnetite which support magnetotaxis (Schleifer et al., 1991). Alphaproteobacteria include obligate aerobic bacteria such as Maribius (Choi et al., 2007) as well as facultative aerobes, facultative anaerobes like Pannonibacter (Borsodi et al., 2003), and obligate anaerobes such as Phaeobacterium (Borsodi et al., 2003; Choi et al., 2007; Nupur et al., 2015). Yet the vast majority of Alphaproteobacteria are aerobes and to a lesser extent facultative anaerobes.

As for chemotaxonomy, the presence of sphingolipids is remarkable within Alphaproteobacteria since it appears to be restricted to Sphingomonadales (Kosako and Yabuuchi, 2005). Morphologically, Alphaproteobacteria are mostly found to be rod-, coccus- or ovoid-shaped. Yet some taxa deviate from this pattern, such as the spirilla-shaped Rhodospirillaceae (Pfennig and Trüper, 1971) including Magnetospirillum (Schleifer et al., 1991) and Thalassospira (López-López et al., 2002; Liu et al., 2007; Tsubouchi et al., 2014). Caulobacter (Henrici and Johnson, 1935; Abraham et al., 1999) and Brevundimonas (Segers et al., 1994; Abraham et al., 1999) of Caulobacteraceae (Henrici and Johnson, 1935), as well as Litorimonas (Jung et al., 2011; Nedashkovskaya et al., 2013), Hellea (Alain et al., 2008) and Oceanibulbus (Wagner-Döbler et al., 2004) of Rhodobacteraceae (Garrity et al., 2005b), are also unique as they form stalks. Many Alphaproteobacteria are motile by means of flagella, as exemplified by Caulobacterales (Henrici and Johnson, 1935) which mostly display flagella. Periplasmic flagella are present in some species, particularly in Salinispira (Ben Hania et al., 2015). Gliding motility has rarely been reported; examples are Pacificimonas (Liu K. et al., 2014) and Acuticoccus (Hou et al., 2015).

The class Alphaproteobacteria was proposed relatively recently (Garrity et al., 2005a) even though the first representatives of the group were isolated as early as 1898 (Beijerinck, 1898). As in other groups of bacteria the initial classification of Alphaproteobacteria into orders, families and genera was based on morphological and physiological characteristics, whereas advances in molecular systematics led to the view that taxonomic classification should be based on the integrated use of genotypic and phenotypic data (Wayne et al., 1987; Stackebrandt, 1992), an approach known as polyphasic taxonomy (Colwell, 1970; Vandamme et al., 1996; Gillis et al., 2005; Kämpfer and Glaeser, 2012). In particular, 16S rRNA gene sequences have been routinely applied to infer phylogenetic trees or in conjunction with simpler approaches such as pairwise distance or similarities (Meier-Kolthoff et al., 2013b; Kim and Chun, 2014; Yarza and Munoz, 2014). The technique named Multilocus Sequence Analysis or MLSA (Glaeser and Kämpfer, 2015) has widely been used to resolve the phylogeny of different taxonomic groups of Alphaproteobacteria like Ensifer (Martens et al., 2008) and Bradyrhizobium (Rivas et al., 2009). However, trees based on a few thousand nucleotides such as those based on a single phylogenetic marker (1400–1500 nucleotides in the case of the 16S rRNA gene), or even a few concatenated housekeeping genes as in the case of MLSA tend to have branches with low bootstrap values (Klenk and Göker, 2010).

Better resolved phylogenies based on the hundreds of housekeeping genes or even the core-genome has been used to elucidate the phylogenetic relationships among selected groups of closely related taxa (Williams et al., 2007; Wirth and Whitman, 2018). Given the rapid and ongoing progress in sequencing technologies (Mavromatis et al., 2012), classifications based on whole genome sequences and associated bioinformatic tools can be based on millions of characters. This provides a step change in reliability, as evidenced by high average bootstrap support in phylogenomic trees (Breider et al., 2014; Meier-Kolthoff et al., 2014a), even though the ordinary bootstrap is not necessarily the most reliable approach when dealing with supermatrices potentially comprised of genes with distinct histories (Siddall, 2010; Simon et al., 2017). Reclassifications at all levels of the taxonomic hierarchy can result from such approaches (Hahnke et al., 2016; Nouioui et al., 2018). It was also shown that DNA G+C composition values directly calculated from genome sequences have a significantly better fit to the phylogeny than the experimentally determined ones cited in many species descriptions (Hahnke et al., 2016). This is in line with the observation that within-species variation is at most 1% when G+C content is calculated from genome sequences (Meier-Kolthoff et al., 2014c) and that claims in the literature that the variation in G+C content within bacterial species is at most 3 mol% (Mesbah et al., 1989) or even 5% (Rosselló-Mora and Amann, 2001) can be attributed to experimental error in traditional methods (Mesbah et al., 1989; Moreira et al., 2011). Recent studies based on complete genomes also confirm that the distribution of the G+C content is phylogenetically conserved. While this also holds to a somewhat lesser degree for genome size (Nouioui et al., 2018), phylogenetic inertia of these features has not yet been measured for Alphaproteobacteria. Likewise, it is as yet unknown to which degree gene-content phylogenies (Huson and Steel, 2004) are in concordance with standard genome-scale phylogenies even though both approaches showed high agreement in subgroups of Alphaproteobacteria (Breider et al., 2014) and because the gene content is of relevance as it conveys phenotypic features (Zhu et al., 2015).

The aim of the present study is an improved phylogenetic framework for the classification of Alphaproteobacteria. Genome-scale phylogenetic trees were inferred for genome-sequenced type strains and augmented by analyses of a comprehensive collection of type-strain 16S rRNA gene sequences to address the following questions: (i) to what extent are phylogenies calculated from whole genome sequences still in conflict with the current classification of Alphaproteobacteria and with their 16S rRNA gene phylogenies? (ii) Which taxa need to be revised because they are evidently non-monophyletic? (iii) Which taxon descriptions should be modified because of inaccurate or missing G+C values? and (iv) How do G+C content, genome size, genomic gene content and routinely recorded phenotypic features of Alphaproteobacteria relate to their phylogeny and to which degree can they serve as a taxonomic marker?

Materials and Methods

The approach to taxon sampling and analysis was in almost all respects the same as previously described (Hahnke et al., 2016; Nouioui et al., 2018). A total of 1104 annotated type-strain genome sequences (Supplementary Table S1) for Alphaproteobacteria (ingroup) and Spirochaetes (outgroup) were collected. While some originated from GenBank the majority was obtained de novo in the course of the KMG projects phase II (Mukherjee et al., 2017) and phase IV and deposited in the Integrated Microbial Genomes platform (Chen et al., 2019) and in the Type-Strain Genome Server database (Meier-Kolthoff and Göker, 2019). Among Alphaproteobacteria KMG-II mainly targeted Rhodobacteraceae but also representatives of other families. All newly generated KMG sequences underwent standard quality control at DSMZ and JGI documented on the respective web pages and had < 100 contigs. All accepted genome sequences had < 500 contigs and matched the 16S rRNA gene reference database described below. Structural annotation at JGI and DSMZ was done using Prodigal v. 2.6.2 (Hyatt et al., 2010). The features of all genome sequences that entered these analyses are provided in Supplementary Table S1. These annotated genome sequences were processed further as in our previous study using the high-throughput version of the Genome BLAST Distance Phylogeny (GBDP) approach in conjunction with BLAST+ v2.2.30 in blastp mode (Auch et al., 2006; Camacho et al., 2009; Meier-Kolthoff et al., 2014a) and FastME version 2.1.6.1 using the improved neighbor-joining algorithm BioNJ for obtaining starting trees followed by branch swapping under the balanced minimum evolution criterion (Desper and Gascuel, 2004) using the subtree-pruning-and-regrafting algorithm (Desper and Gascuel, 2006; Lefort et al., 2015). One hundred pseudo-bootstrap replicates (Meier-Kolthoff et al., 2013a, 2014a) were used to obtain branch-support values for these genome-scale phylogenies.

Trees were visualized using Interactive Tree Of Life (Letunic and Bork, 2019) in conjunction with the script deposited at https://github.com/mgoeker/table2itol. Outgroup-based rooting was compared with rooting using least-squares dating as implemented in LSD version 0.2 (To et al., 2016) after removing the outgroup taxa and inferring an accordingly reduced tree with FastME. Species and subspecies boundaries were investigated using digital DNA:DNA hybridization (dDDH) as implemented in the Genome-To-Genome Distance Calculator (GGDC) version 2.1 (Meier-Kolthoff et al., 2013a) and in TYGS, the Type (Strain) Genome Server (Meier-Kolthoff and Göker, 2019).

In addition to GBDP formula d₅, which explores sequence (dis-)similarity and is the recommended one for phylogenetic inference (Auch et al., 2006; Meier-Kolthoff et al., 2014a) we here used formula d₃, which compares the gene content of the investigated genomes after correcting for reduction in genome size (Henz et al., 2005). While this analysis was also done using the GBDP software, for consistency with previous work we will refer to the d₅ phylogeny as GBDP tree and to the d₃ tree as gene-content analysis. There are various reasons why a gene-content phylogeny may fail to recover the true tree, as detailed below, hence the gene-content analysis is not intended to lend phylogenetic support. However, it may nevertheless be of taxonomic interest whether or not a certain branch is supported by gene-content data, particularly since the gene content conveys metabolic capabilities (Zhu et al., 2015) and yield independent evidence for conclusions from standard genome-scale phylogenies (Breider et al., 2014).

Full-length 16S rRNA gene sequences were extracted from the genomes using RNAmmer version 1.2 (Lagesen et al., 2007) and compared with the 16S rRNA gene reference database using BLAST and phylogenetic trees to verify the taxonomic affiliation of genomes. Non-matching genome sequences were discarded from further analyses. A comprehensive sequence alignment was generated with MAFFT version 7.271 with the “localpair” option (Katoh et al., 2005) using either the sequences extracted from the genome sequences or the previously published 16S rRNA gene sequences, depending on the length and number of ambiguous bases. Trees were inferred from the alignment with RAxML (Stamatakis, 2014) under the maximum-likelihood (ML) criterion and with TNT (Goloboff et al., 2008) under the maximum-parsimony (MP). In addition to unconstrained, comprehensive 16S rRNA gene trees (UCT), constrained comprehensive trees (CCT) were inferred with ML and MP using the bipartitions of the GBDP tree with ≥95% support as backbone constraint, as previously described (Hahnke et al., 2016; Nouioui et al., 2018).

Taxa were analyzed to determine whether they were monophyletic, paraphyletic or polyphyletic (Farris, 1974; Wood, 1994) Taxa non-monophyletic according to the GBDP tree were tested for evidence for their monophyly in the UCT and the 16S rRNA gene trees, if any, in the original publication. In the case of a significant conflict (i.e., high support values for contradicting bipartitions) between trees or low support in the GBDP tree, additional phylogenomic analyses of selected taxa were conducted. To this end, protein sequences of those taxa with the reciprocal best hits from GBDP/BLAST were clustered with MCL (Markov Chain Clustering) version 14-137 (Enright et al., 2002) under default settings and an e-value filter of 10^–5 in analogy to OrthoMCL (Li et al., 2003). The resulting sets of orthologous proteins were aligned with MAFFT and concatenated to form a supermatrix after discarding the few clusters that still contained more than a single protein for at least one genome. Comprehensive supermatrices were compiled from all the orthologs that occurred in at least four genomes, whereas core-genome supermatrices were constructed for the orthologs that occurred in all of the genomes. Supermatrices were analyzed with TNT, and with RAxML under the “PROTCATLGF” model, in conjunction with 100 partition bootstrap replicates (Siddall, 2010; Simon et al., 2017)

Additionally, selected phenotypic features relevant for the taxonomic classification of Alphaproteobacteria were as comprehensively as possible collected from the taxonomic literature: motility by flagella, absence or presence of carotenoids, absence or presence of bacteriochlorophyll α, absence or presence of sphingolipids, average number of isoprene residues of the major ubiquinones, and relationship to oxygen. To avoid circular reasoning, missing features of a species were only inferred from features of its genus when species and genus were described in the same publication or when the species description had explicitly been declared as adding to the features of the genus. For the binary chemotaxonomic characters an alternative coding was also investigated that treated all missing values as indicating absence. Ubiquinone percentages would be more informative than just statements about being “major” but mostly only the latter are provided in the literature. Oxygen conditions were coded as ordered multi-state character: (1) strictly anaerobic; (2) facultatively aerobic, facultatively anaerobic, or microaerophilic; (3) strictly aerobic. Among all nine coding options tested, this yielded the highest fit to the tree (Supplementary Table S1) but the differences between the coding options were not pronounced. Phylogenetic conservation of selected phenotypic and genomic characters with respect to the GBDP tree (reduced to represent each set of equivalent strains by only a single genome) was evaluated using a tip-permutation test in conjunction with the calculation of maximum-parsimony scores with TNT as previously described (Simon et al., 2017; Carro et al., 2018) and 10,000 permutations. TNT input files were generated with opm (Vaas et al., 2013). The proportion of times the score of a permuted tree was at least as low as the score of the original tree yielded the p-value. Maximum-parsimony retention indices (Farris, 1989; Wiley and Lieberman, 2011) were calculated to further differentiate between the fit of each character to the tree.

Taxa that were unambiguously non-monophyletic according to the genome-scale analyses were screened for published evidence of their monophyly. The published evidence was judged as inconclusive when based on unsupported branches in phylogenetic trees, based on probably homoplastic characters or on probable plesiomorphic character states. Plesiomorphies might well be “diagnostic” but just for paraphyletic groups (Hennig, 1965; Wiley and Lieberman, 2011; Montero-Calasanz et al., 2017) hence “diagnostic” features alone are insufficient in phylogenetic systematics.

For fixing the obviously non-monophyletic taxa taxonomic consequences were proposed if new taxon delineations could be determined that were sufficiently supported by the CCT. In these cases, the uncertain phylogenetic placement of taxa whose genome sequences were not available at the time of writing would not affect the new proposals. Where necessary taxa were tentatively place in newly delineated groups.

Results

The presentation of the results is organized as follows. After a brief overview on the figures and tables the outcome of the tests for the phylogenetic conservation are illustrated. Next, the phylogenetic results for the outgroup taxa are described and put in the context of their current taxonomic classification. Finally, the hierarchical classification of the class Alphaproteobacteria itself, arranged according to the orders in which it is currently subdivided and then according to the taxonomic categories, is compared to the phylogenomic trees. These sections motivate the need for a variety of reclassifications, whereas the actual taxonomic consequences are listed at the end of the section “Discussion.” Finally, the outcome of the tests for the phylogenetic conservation are illustrated.

The GBDP tree is shown in Figures 1–8; Figure 1 provides an overview and explains which specific sections of the same tree are displayed in greater detail in Figures 2–8. Table 1 shows dDDH results for pairs of type strains of interest, while Table 2 displays the results of the tests for phylogenetic inertia. Phenotypic information for groups of taxa whose taxonomic classification is treated in detail below is summarized in Supplementary Table S1. This Supplementary Table S1 also contains the complete list of genome sequences used in this study, including their GenBank and IMG accession numbers. Additional phylogenetic trees, including the GBDP tree in a single figure and with phenotypic annotation and the results from the gene-content analysis, are found in Supplementary File S2.

FIGURE 1

FIGURE 2

FIGURE 3

FIGURE 4

FIGURE 5

FIGURE 6

FIGURE 7

FIGURE 8

Classes and Orders

The taxon sampling used in the present study was not mainly intended to provide support for or against the monophyly of the class Alphaproteobacteria, or of the phylum Proteobacteria in general. The choice of the outgroup in the present study was not intended to indicate that the phylum Spirochaetes represents the sister group of Alphaproteobacteria but was motivated by uncertainty regarding the monophyly of Proteobacteria (Yarza et al., 2014). Inferring the tree depicted in Figure 1 again after removing the outgroup and rooting this reduced tree using least-squares dating yielded the same branching order for the ingroup, i.e., the root was located between the clade formed by Magnetococcus and Mariprofundus on the one hand and the remainder of the tree on the other hand. Spirochaetes may thus not be the ideal outgroup for Alphaproteobacteria phylogeny but the alternative rooting confirmed the depicted branching order.

Only a single issue regarding the classes became apparent in this study, and most of the orders of the class Alphaproteobacteria appeared as monophyletic in our analysis (Figure 1). The exceptions were mainly caused by specific genera taxonomically assigned to Rhodospirillales and particularly genera assigned to Rhodobacterales that were phylogenetically intermixed with the order currently called Rhizobiales.

Alphaproteobacteria appeared as paraphyletic in the GBDP tree (Figures 1, 2) since Mariprofundus ferrooxydans of Zetaproteobacteria (Emerson et al., 2007) formed a strongly supported clade together with the alphaproteobacterium Magnetococcus marinus (Bazylinski et al., 2013a). The clade even obtained reasonable support in the gene-content analysis (Supplementary File S2) and its two representatives displayed almost the same G+C content. In the original description of M. ferrooxydans (Emerson et al., 2007) a new class (Zetaproteobacteria), order (Mariprofundales) and family (Mariprofundaceae) were proposed in the Supplementary Material only. None of these names became validly published so far even though a corrected name, Mariprofundia, was suggested for Zetaproteobacteria in the meantime (Oren, 2017a).

In the originally presented 16S rRNA gene trees the placement of Mariprofundus has no strong support and Magnetococcus marinus could not yet be considered. The additionally presented protein phylogenies (RecA, GyrB) only partially showed support for the placement of M. ferrooxydans separate from Alphaproteobacteria. In the CCT we did not find strong support for the placement of M. ferrooxydans branching first within the ingroup (Supplementary File S2). Phenotypically, alphaproteobacterial taxa such as Magnetococcus marinus are capable of forming iron oxides much like M. ferrooxydans. Even though the filamentous iron oxyhydroxide and branched-chain fatty acids produced by this species may differentiate it from taxa with a similar ecology, this alone provides no evidence for a separate class. It thus makes sense to again propose a separate order and family for Mariprofundus but to tentatively assign it to the class Alphaproteobacteria; an alternative arrangement is to remove Magnetococcus from Alphaproteobacteria.

Within Alphaproteobacteria, Rhodospirillales appeared as paraphyletic in the GBDP tree and in the CCT (Figure 2 and Supplementary File S2) because Kiloniella (Wiese et al., 2009; Yang S.-H. et al., 2015) of Kiloniellales as well as Roseospirillum (Glaeser and Overmann, 1999) and Terasakiella (Satomi et al., 2002; Han et al., 2016) of the order currently called Rhizobiales were nested within Rhodospirillales with high support. As this also affects the monophyly of the family Rhodospirillaceae, we will below propose the reclassification of these three genera into Rhodospirillales as the preferred way to restore a monophyletic order and family. Moreover, the distant position of Geminicoccaceae (Proença et al., 2018) also conflicts with the monophyly of Rhodospirillales. Because this conflict was only poorly supported, we do not propose taxonomic changes for Geminicoccaceae based on the here examined data.

Rhodobacterales were shown as non-monophyletic in various ways, most of which also affect families and will thus be treated below. Rhizobiales (Kuykendall, 2005) appeared as paraphyletic in GBDP tree (Figures 3–5 and Supplementary File S2) because Acuticoccus (Hou et al., 2015) Ahrensia (Uchino et al., 1998; Liu J. et al., 2016) Labrenzia (Biebl et al., 2007;Bibi et al., 2014) Nesiotobacter (Donachie et al., 2006) Pannonibacter (Borsodi et al., 2003; Biebl et al., 2007) Pseudovibrio (Shieh et al., 2004) Roseibium (Suzuki et al., 2000) and Stappia (Uchino et al., 1998; Biebl et al., 2007) all of which are currently classified in Rhodobacterales, were nested within Rhizobiales. According taxonomic solutions are suggested below for the affected families. Hartmannibacter (Suarez et al., 2014) which was not explicitly assigned to an order in its original description, is also treated below.

It should also be noted that Rhizobiales (Kuykendall, 2005) is validly published but illegitimate (i.e., not in accordance with the rules of the International Code of Nomenclature of Prokaryotes) as this order includes Hyphomicrobium of Hyphomonadaceae, type genus of Hyphomicrobiales (Douglas, 1957) which has priority. Our analyses do not call for placing Rhizobium and Hyphomicrobium in distinct orders (Figure 4 and Supplementary File S2), hence we will below propose an emended description of Hyphomicrobiales to replace the illegitimate Rhizobiales. The following description of the results will nevertheless use the name Rhizobiales throughout because this is the name used in our taxonomic input data although the name will be marked as in need of a replacement.

Neomegalonema (Oren, 2017b) of Rhizobiales (Hyphomicrobiales) was phylogenetically placed as sister group of Rhodobacteraceae with strong support to the exclusion of Rhizobiales (Figure 7), an arrangement that also obtained reasonable support in the gene-content analysis (Supplementary File S2). Neomegalonema was originally proposed as Meganema (Thomsen et al., 2006) which was later on regarded as an illegitimate name, and supposed to be related to the “Methylobacterium/Xanthobacter group” within Alphaproteobacteria based on a 16S rRNA gene analysis with a reduced taxon sampling and without calculating branch support. Given the phylogenetic evidence presented here and the lack of evidence to the contrary, the genus should be transferred from Rhizobiales (Hyphomicrobiales) to Rhodobacterales. Instead of assigning the genus to the already highly divergent and mainly marine family Rhodobacteraceae we suggest to place Neomegalonema into a family of its own within Rhodobacterales. This proposal is in accordance with the major phenotypic features of the genus (Supplementary Table S1).

Outgroup Families and Genera

A revision of the phylum Spirochaetes is beyond the scope of the current study but some cautionary remarks on its taxonomic classification seem appropriate.

Spirochaetaceae (Swellengrebel, 1907; Abt et al., 2012; Gupta et al., 2013) were shown as rather heterogeneous assemblage not particularly well supported as monophyletic in the GBDP tree and in the CCT (Figure 2 and Supplementary File S2). While Borreliaceae (Gupta et al., 2013) were already separated from Spirochaetaceae, the overall genomic and phenotypic divergence of the group still calls for splitting Spirochaetaceae. The current taxonomic placement of Sphaerochaeta (Abt et al., 2012; Ritalahti et al., 2012; Miyazaki et al., 2014a; Arroua et al., 2017) which does not form spiral-shaped cells (Supplementary Table S1), makes Spirochaetaceae phenotypically heterogeneous. Spiral-shaped cells are apparently an apomorphy of the phylum and thus plesiomorphic within the phylum, hence this feature does not provide evidence for grouping Treponema within Spirochaetaceae.

An earlier taxonomic study already proposed the family Treponemataceae (Robinson, 1948) to accommodate the type genus Treponema (Schaudinn, 1905; Abt et al., 2013). Hence, we suggest to reuse the validly published name Treponemataceae and to place Sphaerochaeta in a phenotypically homogeneous family of its own. According to the CCT and UCT (Supplementary File S2), Pleomorphochaeta (Arroua et al., 2017) which like Sphaerochaeta displays spherical cells (Supplementary File S1), should be assigned to Sphaerochaetaceae fam. nov., too, whereas Rectinema (Koelschbach et al., 2017) should be placed in Treponemataceae. Rectinema produces non-motile rods but, according to its phylogenetic position (Supplementary File S2), the development of non-helical cells occurred independently of Pleomorphochaeta and Sphaerochaeta. In contrast to Spirochaetaceae sensu stricto as proposed here, Sphaerochaetaceae fam. nov. and Treponemataceae were strongly supported by the gene-content analysis (Supplementary File S2).

As for the outgroup genera, Spirochaeta (Pikuta et al., 2009; Miyazaki et al., 2014b; Shivani et al., 2015) of Spirochaetaceae was shown as paraphyletic in the GBDP tree and in the CCT (Figure 2 and Supplementary File S2) because Alkalispirochaeta (Sravanthi et al., 2016) and Salinispira (Ben Hania et al., 2015) were nested within Spirochaeta. These problems were already visible in the original literature sources whose 16S rRNA gene analyses indicated that the proposal of these new genera would render Spirochaeta paraphyletic. A major obstacle in treating the genus with modern taxonomic methods is that the type species of Spirochaeta, Spirochaeta plicatilis, is not represented by a type strain (Skerman et al., 1980). One potential albeit radical solution is to place all other Spirochaeta species into genera of their own. Because of overall insufficient resolution in the 16S rRNA gene trees (Supplementary File S2), we refrain from a taxonomic revisions of the genus because genome sequences of relevance are not yet available. The classification of Spirochaetaceae sensu stricto should be reconsidered once more genome sequences become available.

Rickettsiales Families

Rickettsiales is an order of Alphaproteobacteria that comprises pathogens with reduced genomes and a reduced G+C content (Figure 2). Our analysis did not call the monophyly of the order into question, but a nomenclatural issue became apparent in the course of this study.

We found that Anaplasmataceae (Philip, 1957; Dumler et al., 2001) is illegitimate if one takes its last emendation into account as this family now includes Ehrlichia (Moshkovski, 1945) which is the type genus of Ehrlichiaceae (Moshkovski, 1945) which has priority. Our analyses do not argue for placing the two genera, which form a clade strongly supported even by the gene-content analysis, into distinct families (Figure 1 and Supplementary File S2), hence we will propose below an emended description of Ehrlichiaceae as the correct name for the family of the genera currently included in Anaplasmataceae.

Rhodospirillales Families

A variety of families of Rhodospirillales appeared in need of a taxonomic revision in the light of our analyses (Figure 2). These discrepancies were not only caused by Rhodospirillales appearing to be intermixed with Kiloniellales.

Acetobacteraceae (Gillis and De Ley, 1980) were shown as non-monophyletic in the GBDP tree and in the CCT (Figure 2 and Supplementary File S2) because of the relatively isolated position of Stella (Vasilyeva, 1985) and the even more distinct position of Zavarzinia (Meyer et al., 1993). In the original description of Zavarzinia, a phylogenetic analysis could not yet be conducted, and the genus was not assigned to any family. Later the genus was assigned to Acetobacteraceae (Boone et al., 2001b) but the reasoning behind this remained unclear. Here Zavarzinia was shown as branching first within Rhodospirillales. While the backbone support within the order was low, strong support indicates that Zavarzinia does not belong to core Acetobacteraceae. Given its isolated position in the GBDP and 16S rRNA gene trees, the genus is best assigned to a family of its own. This taxonomic consequence did not appear in conflict with the major phenotypic features of Zavarzinia (Supplementary Table S1).

Stella appeared as sister group of the equally deviating genus of Rhodospirillaceae, Reyranella (Pagnier et al., 2011) with moderate support, and separated from core Acetobacteraceae by long branches. The original description of Stella did not explicitly assign the genus to a family. Stella was placed in Acetobacteraceae in Bergey’s manual (Boone et al., 2001a); whether this placement was based on a phylogenetic assessment is unclear. Phenotypically the genus is rather unique because of its star-like morphology (Supplementary Table S1). All results thus suggest placing Stella into a family of its own. The same holds for Reyranella, which is treated below; the gene-content analysis does not support the sister-group relationship of the two genera.

When Constrictibacter (Yamada et al., 2011) was proposed it was placed in Rhodospirillaceae in a 16S rRNA gene tree with low support values. In the CCT (Supplementary File S2) with its much broader taxon sampling Constrictibacter formed a clade together with Stella albeit with low support. We thus tentatively include Constrictibacter in the family newly proposed to accommodate Stella (see above). Although phenotypic differences in morphology and respiration might suggest to alternatively place Constrictibacter into a family of its own this solution should be postponed until a Constrictibacter genome sequence is available.

Rhodospirillaceae (Pfennig and Trüper, 1971) appeared as paraphyletic in GBDP tree and in the CCT (Figure 2 and Supplementary File S2) for a variety of reasons. For instance, Roseospirillum (Glaeser and Overmann, 1999) of Rhodobiaceae was nested within Rhodospirillaceae. Roseospirillum was placed as sister group of Rhodospira (Pfennig et al., 1997) with high support (Figure 2), a clade that in turn formed, with high support, the sister group of a clade containing the type species of the type genus of the family, Rhodospirillum rubrum (Molisch, 1907) these arrangements are even supported by the gene-content analysis (Supplementary File S2). For these reasons, we propose to include Roseospirillum into Rhodospirillaceae, which is also supported by the high phenotypic agreement between them (Supplementary Table S1). When Roseospirillum was proposed, bootstrapping was not conducted, and only few species could be included in the phylogenetic analysis at that time.

Furthermore, Ferrovibrio (Sorokina et al., 2012) Taonella (Xi et al., 2013) and Marinibaculum (Yu et al., 2016) formed an isolated but strongly supported clade with Sneathiellaceae (Kurahashi et al., 2008) in the CCT (Supplementary File S2). When Ferrovibrio was proposed, it was already placed in the order Sneathiellales but not explicitly in Sneathiellaceae. When Taonella and Marinibaculum were proposed both publications lacked strong support for the positioning of the respective taxa and both publications did not sample Sneathiella (Jordan et al., 2007). Therefore, we propose to tentatively assign Ferrovibrio, Taonella, and Marinibaculum to Sneathiellaceae. These taxonomic consequences are not precluded by the phenotypic features of the involved genera (Supplementary Table S1).

Similarly, Terasakiella (Satomi et al., 2002; Han et al., 2016) of Methylocystaceae and Kiloniella (Wiese et al., 2009; Yang S.-H. et al., 2015) of Kiloniellaceae were nested within Rhodospirillaceae (Figure 2). However, even if these genera were included in Rhodospirillaceae, the family would lack any support. For this reason, to address the remaining causes of the non-monophyly of Rhodospirillaceae other measures are advisable. Rhodospirillaceae would not appear monophyletic even if these three genera were included. Rhodospirillaceae showed high genomic divergence (Figure 2) and unsupported by the gene-content analysis (Supplementary File S2). The family is phenotypically heterogeneous, as, e.g., most of its representatives are phototrophic whereas some genera are chemoorganoheterotrophs (Supplementary Table S1). Splitting Rhodospirillaceae into several families corresponding to well-supported clades can also solve this issue.

Azospirillaceae fam. nov. is thus proposed in agreement with the GBDP tree (Figure 2 and Supplementary File S2) to accommodate Niveispirillum (Lin et al., 2014) Azospirillum (Tarrand et al., 1978; Falk et al., 1985) Rhodocista (Kawasaki et al., 1993) and Skermanella (Sly and Stackebrandt, 1999; Luo et al., 2012; Weon et al., 2007) Nitrospirillum (Lin et al., 2014) and Desertibacter (Liu et al., 2011) are also assigned to this family based on the 16S rRNA gene analyses (Supplementary File S2). Even though the gene-content analysis does not support the clade but only its two major subclades, this conclusion fits well to the major phenotypic features of these genera (Supplementary Table S1), most of which also show larger genomes than the other genera currently included in Rhodospirillaceae. Azospirillum and Rhodocista were placed in Rhodospirillaceae in Bergey’s manual (Boone et al., 2001a) but it remained unclear to us whether this was based on a phylogenetic assessment. The proposals of these genera directly assigned them to Rhodospirillaceae but phylogenetic support for the monophyly of the family was not presented. The situation regarding genera such as Elstera (Rahalkar et al., 2012) and Inquilinus (Coenye et al., 2002) is less clear in the GBDP tree and particularly in the 16S rRNA gene trees, even the constrained ones. This holds also for Lacibacterium (Sheu et al., 2013) which lacks a published genome but was found to be the sister taxon of Elstera according to 16S rRNA gene analyses when it was proposed (Sheu et al., 2013) and in our findings (Supplementary File S2). These genera are tentatively assigned to the new family Azospirillaceae, too. Further reclassifications should be attempted once more genome sequences become available.

When Terasakiella was originally described, it was not assigned to any family. Terasakiella was placed in Methylocystaceae later on (Garrity et al., 2003b) but this proposal did not appear to be based on a phylogenetic analysis. The last emendation of the genus (Yoon and Kang, 2018) still assigned Terasakiella to Methylocystaceae but did not include the type species of the family in the phylogenetic analysis. We conclude that the taxonomic literature contains no phylogenetic evidence for an affiliation of the genus to the family. Similarly, Magnetovibrio (Williams et al., 2012) Thalassospira (López-López et al., 2002), and Varunaivibrio (Patwardhan and Vetriani, 2016) were placed in Rhodospirillaceae based on largely unresolved 16S rRNA gene trees with a reduced taxon sampling. Given the phylogenetic evidence presented here (Figure 2 and Supplementary File S2), we propose Thalassospiraceae fam. nov. to accommodate Magnetovibrio (Bazylinski et al., 2013b) Terasakiella, Thalassospira (López-López et al., 2002; Liu et al., 2007; Tsubouchi et al., 2014) and Varunaivibrio. Magnetospira (Williams et al., 2012) is also tentatively assigned to this family based on to the 16S rRNA gene analyses (Supplementary File S2). These taxonomic consequences are not precluded by the phenotypic features of the involved genera (Supplementary Table S1). The gene-content analysis does not support the newly proposed family but genome size and G+C content are rather homogeneous (Figure 2 and Supplementary File S2).

When Kiloniella, Kiloniellaceae and Kiloniellales were proposed (Wiese et al., 2009) the phylogenetic analysis was characterized by low support at the backbone, and the monophyly of the resulting orders and families was not immediately evident. The most conservative solution for the observed discrepancy (Figure 2) is to retain Kiloniellaceae but to place it in Rhodospirillales. Aestuariispira (Park et al., 2014e) should according to our phylogenetic results also be placed in the family. Aestuariispira was placed in Rhodospirillaceae when the genus was originally proposed but this was based on a partially rather unresolved 16S rRNA gene tree that lacked a representative sampling of Alphaproteobacteria. Among the known major phenotypic features, it differs from Kiloniella only regarding motility (Supplementary Table S1). The clade comprising Kiloniella and Aestuariispira also receives strong support in the gene-content analysis (Supplementary File S2).

Marivibrio (Chen S. et al., 2017b) was originally placed in Rhodospirillaceae based on a 16S rRNA gene analysis with low taxon sampling in which Marivibrio was grouped together with Pelagibius of Rhodospirillaceae. In the CCT Marivibrio was isolated from Rhodospirillaceae but formed a clade together with Kiloniella of Kiloniellaceae and Aestuariispira as its sister taxon. Because of low support in the CCT we propose to tentatively include Marivibrio into Kiloniellaceae. Although phenotypic differences might suggest the alternative of assigning Marivibrio to its own family we here refrain from this solution due to a lack of genomic data.

When Thalassocola (Lin et al., 2015) was proposed it was placed in Phyllobacteriaceae in a 16S rRNA gene tree with low support. In the CCT (Supplementary File S2) Kiloniella of Kiloniellaceae appeared as the sister taxon of Thalassocola albeit with low support. We thus propose to tentatively include Thalassocola into Kiloniellaceae which is taxonomically more favorable to its current inclusion in Phyllobacteriaceae. Although phenotypic differences might suggest the alternative of assigning Thalassocola to its own family we refrain from this solution because of the low resolution and because it would be taxonomically less conservative.

Given that the non-monophyly of Rhodospirillaceae should be solved by splitting rather than by merging, the trees inferred from genome-scale data (Figure 2 and Supplementary File S2) also call for the proposal of Rhodovibrionaceae fam. nov. to accommodate Fodinicurvata (Wang Y. et al., 2009) Limimonas (Amoozegar et al., 2013) Rhodovibrio (Imhoff et al., 1998) and Tistlia (Díaz-Cárdenas et al., 2010) Limibacillus (Kim et al., 2015) and Pelagibius (Choi et al., 2009) are also tentatively assigned to this family based on the 16S rRNA gene analyses (Supplementary File S2). The creation of this family is neither contradicted by the gene-content analysis (even though it does not provide support) nor by major phenotypic features of the involved genera (Supplementary Table S1). For instance, while the resulting family is heterogeneous regarding the occurrence of phototrophy and related pigments, these character states appear relatively scattered across the phylogeny (Figure 2 and Supplementary File S2). Rhodovibrio was placed in Rhodospirillaceae in Bergey’s manual (Boone et al., 2001a) but it remained unclear to us whether this was based on a phylogenetic assessment. As in the case of the deviating genera of Rhodospirillaceae treated above, the original proposals of Fodinicurvata, Limibacillus, Limimonas, and Tistlia assigned these genera to the family, each time based on 16S rRNA gene trees that were either largely unresolved or lacked a representative taxon sampling, i.e., strong phylogenetic evidence for the monophyly of Rhodospirillaceae including these genera was not presented. In contrast, at least moderate bootstrap support was obtained in the taxonomic literature (Amoozegar et al., 2013) for a clade comprising Fodinicurvata, Limibacillus, Limimonas, Rhodovibrio, and Tistlia, in accordance with our results.

Based on the phylogenetic results (Figure 2 and Supplementary File S2), Thalassobaculaceae fam. nov. is proposed to contain Nisaea (Urios et al., 2008) Oceanibaculum (Lai et al., 2009; Dong et al., 2010) and Thalassobaculum (Zhang G.I. et al., 2008; Urios et al., 2010). These genera were placed in Rhodospirillaceae when they were originally proposed, each time based on 16S rRNA gene trees that were either largely unresolved or lacked a representative taxon sampling, i.e., there was no strong phylogenetic evidence for the monophyly of Rhodospirillaceae including these genera in the taxonomic literature. We did not find any significant phenotypic differences between these genera (Supplementary Table S1). The monophyly of this family obtained no support in the gene-content analysis but moderate support in the 16S rRNA gene analyses (Supplementary File S2).

While taxonomically assigned to Rhodospirillaceae, Reyranella massiliensis (Pagnier et al., 2011) displayed an isolated position in the phylogenomic tree (Figure 2), distant to the type genus of the family. When Reyranella was proposed, it appeared as sister group of Magnetospirillum magnetotacticum (Schleifer et al., 1991) of Rhodospirillaceae but with low support only. More importantly, the low taxon sampling in that study did not allow for safely assigning Reyranella to a family; the same holds for the subsequent emendations of the genus (Kim et al., 2013; Lee H. et al., 2017) Reyranella is best assigned to a family of its own, which is not contradicted by the phenotype (Supplementary Table S1).

For several genera of Rhodospirillaceae, namely Aliidongia (Chen et al., 2017a), Dongia (Liu Y. et al., 2010), Constrictibacter (Yamada et al., 2011), Defluviicoccus (Maszenan et al., 2005), Marivibrio (Chen S. et al., 2017b), Tagaea (Jean et al., 2016), and Tistrella (Shi et al., 2002) our CCT results (Supplementary File S2) questioned their assignment to Rhodospirillaceae. Aliidongia was placed in Rhodospirillaceae when the genus was originally proposed but this was based on a 16S rRNA gene analysis with low taxon sampling in which Aliidongia was grouped together with Inquilinus and Dongia of Rhodospirillaceae with low support. When Dongia was originally proposed, it grouped together in a 16S rRNA gene tree with Rhodospirillaceae genera including Azospirillum, Rhodocista and Skermanella, but with overall low taxon sampling. When Tagaea was originally proposed it was placed in a 16S rRNA gene tree in a well supported clade together with Oceanibaculum, Nisaea, and Thalassobaculum of Rhodospirillaceae. Yet the placement of the clade itself showed low support and overall taxon sampling lacked the type genus of Rhodospirillaceae. None of these genera could safely be placed in a family in the CCT (Supplementary File S2) hence we recommend to regard Aliidongia, Dongia, and Tagaea as genera incertae sedis until their phylogenetic position can be clarified once more genome sequences become available.

Defluviicoccus (Maszenan et al., 2005) was placed in Rhodospirillaceae (Garrity et al., 2007) after its original description but based on the original 16S rRNA gene analysis wherein Defluviicoccus was placed together with Rhodospirillaceae genera such as Rhodospirillum, Azospirillum, and Magnetospirillum yet with low taxon sampling. Tistrella was originally not placed in any family but was placed in Rhodospirillaceae later on in Bergey’s Manual (Garrity et al., 2003b) which cited the original description of Tistrella even though it had cautioned against an assignment to a family because of low support in 16S rRNA gene analysis. Defluviicoccus and Tistrella were isolated from Rhodospirillaceae in the far better sampled CCT and formed a clade together with Geminicoccus and Arboricoccus of Geminicoccaceae. Support for this arrangement was also low, hence we suggest the tentative inclusion of Defluviicoccus and Tistrella in Geminicoccaceae. Although phenotypic differences might suggest to alternatively assign both genera to their own family we refrain from this solution due to a lack of genomic data.

Rhodospirillales Genera

Some genera of Rhodospirillales also appeared in need of a taxonomic revision in the light of our analyses (Figure 2) although to a lesser extent than the families of the order.

Within Rhodospirillales, Azospirillum (Tarrand et al., 1978; Falk et al., 1985) was shown as paraphyletic in the GBDP and in the 16S trees (Figure 2 and Supplementary File S2) because Azospirillum irakense (Khammas et al., 1989) was placed as sister group of Niveispirillum (Cai et al., 2015, 2018b) with high support. While its genome sequence was lacking at the time of writing, the CCT showed that the type species of Niveispirillum, N. fermenti (Lin et al., 2014) was also placed in this clade, whereas the type species of Azospirillum, A. lipoferum, was placed in a clade together with the four other Azospirillum species represented in the GBDP tree. These arrangements also obtained support in the gene-content analysis (Supplementary File S2). It was already proposed to reclassify A. irakense as N. irakense (Lin et al., 2014) but this name was not validly published, and hence Niveispirillum remained non-monophyletic. While A. irakense differs from Niveispirillum by its microaerophilic lifestyle (Supplementary Table S1), this alone is not a sufficient reason to separate two genera. We accordingly propose to place A. irakense in Niveispirillum.

Magnetospirillum (Schleifer et al., 1991) appeared as paraphyletic in the GBDP tree (Figure 2) because Phaeospirillum molischianum (Giesberger, 1947; Imhoff et al., 1998) was nested within Magnetospirillum. Telmatospirillum siberiense (Sizova et al., 2007) formed the sister group of this clade. In addition, in the CCT, P. chandramohanii (Anil Kumar et al., 2009), P. fulvum (van Niel, 1944; Imhoff et al., 1998), P. oryzae (Lakshmi et al., 2011), and P. tilakii (Raj et al., 2012) were also nested within Magnetospirillum (Supplementary File S2). When Magnetospirillum was proposed, the phylogenetic analysis was characterized by a low number of included taxa and by a lack of support values. When Phaeospirillum was proposed, a well resolved tree was presented that included only few species; in particular, Magnetospirillum was not considered (Imhoff et al., 1998). Hence, phylogenetic evidence for the separation of the two genera is lacking. P. chandramohanii, P. fulvum, P. molischianum, P. oryzae, and P. tilakii display a phenotype similar to the one of Magnetospirillum (Supplementary Table S1). Consequently, it is proposed that P. chandramohanii, P. fulvum, P. molischianum, P. oryzae and P. tilakii be classified in Magnetospirillum.

Roseomonas (Rihs et al., 1993; Sánchez-Porro et al., 2009; Venkata Ramana et al., 2010) was shown as polyphyletic in the GBDP tree (Figure 2) because Roseomonas stagni (Furuhata et al., 2008) and Roseomonas lacus (Jiang et al., 2006; Sánchez-Porro et al., 2009) were placed as sister group of Humitalea rosea (Margesin and Zhang, 2013) and Rubritepida flocculans (Alarico et al., 2002) respectively, albeit with low support. In the CCT, species from further genera, such as Rhodovarius lipocyclicus (Kämpfer et al., 2004) were nested within the main Roseomonas clade while Roseomonas fauriae (Rihs et al., 1993) was placed in a remote position as sister group of Azospirillum formosense (Lin et al., 2012) (Supplementary File S2). A previous study concluded that R. fauriae is a later heterotypic synonym of A. formosense (Helsel et al., 2006). While a genome sequence of the type strain of the type species of Roseomonas, R. gilardii, was lacking at the time of writing, it is represented by R. gilardii subsp. rosea in the GBDP tree and safely placed in the CCT. It would be premature, however, to propose new genera for Roseomonas lacus and R. stagni since the low resolution in even the constrained 16S rRNA gene trees currently hinders the assignment of those species not represented by a genome sequence to the resulting set of genera.

Gluconacetobacter (Yamada et al., 1997) appeared as polyphyletic in the GBDP tree (Figure 2) because G. entanii (Schüller et al., 2000) was nested within Komagataeibacter (Yamada et al., 2012) with high support both in the GBDP and the 16S rRNA gene trees (Figure 2 and Supplementary File S2); even the gene-content analysis provided support. When Komagataeibacter was proposed to harbour species formerly placed in Gluconacetobacter (Yamada et al., 2012), G. entanii could not be transferred to the new genus because the type strain seemed unavailable from any culture collection. As long as this problem remains unsolved a new combination for G. entanii cannot be proposed (Parker et al., 2019).

Rhizobiales (Hyphomicrobiales) Families

Many families of Rhizobiales (Hyphomicrobiales) appeared to be in need of a taxonomic revision according to our results (Figures 3–5), even more so than in Rhodospirillales. The discrepancies were on the one hand caused by Rhizobiales (Hyphomicrobiales) families which appeared intermixed and on the other hand caused by genera taxonomically assigned to Rhodobacteraceae within Rhodobacterales but phylogenetically placed within Rhizobiales (Hyphomicrobiales). Both kinds of cases are treated in this section.

Within Rhizobiales (Hyphomicrobiales), Rhodobiaceae (Garrity et al., 2005f) were shown as non-monophyletic in the GBDP tree and in the CCT (Figures 3–5 and Supplementary File S2) because representatives of this family were placed into phylogenetically quite distant clades. Parvibaculum (Schleheck et al., 2004) was placed together with Tepidicaulis (Takeuchi et al., 2015) in a clade that appeared as sister group (Figure 3) of the remaining Rhizobiales (Hyphomicrobiales). When Parvibaculum was proposed, a phylogenetic tree was not presented, and the genus was not assigned to any family. The original description of Tepidicaulis showed it as sister group of Parvibaculum with strong support. The gene-content analysis also strongly supported their sister-group relationship (Supplementary File S2). Thus placing this clade into a family separate from Rhodobiaceae appeared as most appropriate solution. This is not precluded by the phenotype of the involved genera (Supplementary Table S1). Additional genera may need to be added to the newly proposed family once genome sequences provide sufficient resolution.

Similarly, Lutibaculum (Anil Kumar et al., 2012) and Tepidamorphus (Albuquerque et al., 2010) formed a strongly supported clade of their own (Figure 3) within Rhizobiales (Hyphomicrobiales), without any obvious affiliation to an already proposed family. The phylogenies presented in the original descriptions of Lutibaculum and Tepidamorphus already suffered from low support at the backbone; our analyses of the 16S rRNA gene did not show any significant conflict with the phylogenomic tree either. Thus placing these two genera into a family separate from Rhodobiaceae appears as most appropriate solution. According to the 16S rRNA gene analyses (Supplementary File S2), Butyratibacter (Wang et al., 2017) and Microbaculum (Su et al., 2017) should also be placed in this family. This is not precluded by the phenotype of the involved genera (Supplementary Table S1). The gene-content analysis did not support the sister-group relationship of Lutibaculum and Tepidamorphus but they display quite similar G+C content values and genome sizes (Supplementary File S2).

The type genus of the family, Rhodobium (Hiraishi et al., 1995; Urdiain et al., 2008) was placed in an isolated position relative to the rest of the family (Figure 4), only showing a weakly supported sister-group relationship to Afifella (Urdiain et al., 2008). The original description of Afifella was characterized by the lack of branch support and low taxon sampling of the phylogenetic analysis; an assignment of the genus to a family was not proposed. Later on (Su et al., 2017) Afifella was regarded as affiliated to Rhodobiaceae. A sister-group relationship of Afifella and Rhodobium is possible but only supported by the gene-content analysis, which is not the phylogenetic method of choice, while unsupported in the GBDP and 16S rRNA gene trees (Figure 4 and Supplementary File S2). Additional supermatrix analyses indicated that Afifella and Rhodobium do not form a clade (Supplementary File S2). For this reason, a new family is proposed to accommodate Afifella, which is not in disagreement with its phenotype (Supplementary Table S1).

The CCT called the assignment of Anderseniella (Brettar et al., 2007) and Rhodoligotrophos (Fukuda et al., 2012) to Rhodobiaceae into question (Supplementary File S2). Anderseniella was regarded as affiliated to Rhodobiaceae only after its original description (Su et al., 2017) Rhodoligotrophos was placed in Rhodobiaceae when it was originally proposed, which was based on a 16S rRNA gene tree that grouped Rhodoligotrophos together with Parvibaculum of Rhodobiaceae. Yet this tree showed a non-monophyletic Rhodobiaceae. In the CCT Anderseniella and Rhodoligotrophos formed a clade together with Parvibaculum (Schleheck et al., 2004) and Tepidicaulis (Takeuchi et al., 2015) which we propose to transfer to the new family Parvibaculaceae (see above). Anderseniella and Rhodoligotrophos may only tentatively be assigned to Parvibaculaceae because of low support in the CCT. As Pyruvatibacter (Wang G. et al., 2016) has not been assigned to a family yet and is located in the same clade we would also tentatively assign Pyruvatibacter to Parvibaculaceae.

Rhizobiaceae (Conn, 1938) were also shown as non-monophyletic in the GBDP tree and in the CCT (Figure 3 and Supplementary File S2) for several reasons. The genus Kaistia (Im et al., 2004) formed a well-supported subtree distinct from core Rhizobiaceae and close to Bauldia (Yee et al., 2010) which was not yet assigned to a family. The phylogenetic analysis of the 16S rRNA gene used for the proposal of Bauldia was characterized by low support at the backbone of the tree. In its original description Kaistia was not assigned to a family (Im et al., 2004); the genus was placed in Rhizobiaceae only later on (Garrity et al., 2007) but the rationale behind this decision remained obscure. The phenotype of the two genera is largely in agreement (Supplementary Table S1) and there is a certain amount of support for the clade in the gene-content analysis (Supplementary File S2). We accordingly propose a new family Kaistiaceae, fam. nov., to accommodate Kaistia and Bauldia.

As for Hyphomicrobiaceae, it should be mentioned that although Gemmiger (Gossling and Moore, 1975) was listed in Bergey’s Manual (Boone et al., 2001a) as a genus of Hyphomicrobiaceae it was not considered in our analysis because a later study (Yarza et al., 2013) showed that Gemmiger does phylogenetically not even belong to the phylum Proteobacteria. Its taxonomic classification is beyond the scope of the current study.

Even apart from that genus, Hyphomicrobiaceae (Babudieri, 1950) appeared as non-monophyletic in the GBDP and in 16S gene rRNA trees (Figure 3 and Supplementary File S2) in other respects because Aquabacter (Irgens et al., 1991), Blastochloris (Hiraishi, 1997) and Prosthecomicrobium (Staley, 1984) as well as a clade comprising the genera Cucumibacter (Hwang and Cho, 2008b), Devosia (Nakagawa and Yokotat, 1996; Rivas et al., 2003), Maritalea (Hwang et al., 2009b) and Pelagibacterium (Xu et al., 2011) were placed apart from the core Hyphomicrobiaceae clade that comprised the type genus of Hyphomicrobiaceae. Although the genome sequence of the type strain of the type species of the type genus of the family, Hyphomicrobium vulgare (Stutzer and Hartleb, 1899), was not available at the time of writing, the CCT and UCT placed the species with strong support within the clade comprising H. nitrativorans and H. zavarzini (Supplementary File S2). Hyphomicrobiaceae also appeared as the taxonomic home for Methyloceanibacter (Vekeman et al., 2016), which was originally not assigned to a family but was here located within core Hyphomicrobiaceae in Figure 3.

When Aquabacter was proposed (Irgens et al., 1991), a phylogenetic tree was not presented, and the publication did not assign the genus to a family. Aquabacter was placed in Hyphomicrobiaceae in Bergey’s manual (Boone et al., 2001a) but the reason for doing so was not obvious to us. Including Aquabacter into Xanthobacteraceae appears to be the most conservative solution, which is not precluded by the phenotype (Supplementary Table S1). The Aquabacter-Azorhizobium-Xanthobacter clade obtained high support in the gene-content analysis while the 16S rRNA gene analyses even indicated that Aquabacter, Xanthobacter and Azorhizobium are difficult to discern as currently circumscribed (Supplementary File S2). This problem should be addressed once more type-strain genome sequences from the group become available.

The original description of Prosthecomicrobium and its type species, P. hirschii, did not include a 16S rRNA gene analysis (Staley, 1968, 1984). The current assignment of this genus to Hyphomicrobiaceae was called into question in the literature (Lee et al., 2005) and could not be confirmed by our study either. Prosthecomicrobium was first placed in Hyphomicrobiaceae in Bergey’s manual (Boone et al., 2001b) but this assignment may not be based on a phylogenetic analysis. In the CCT (Supplementary File S2) Prosthecomicrobium forms a strongly supported clade with the genera Ancalomicrobium and Pinisolibacter (Staley, 1968; Dahal et al., 2018). Therefore we propose to place Prosthecomicrobium in Ancalomicrobiaceae (Dahal et al., 2018) which is not precluded by the phenotypic characteristics of these genera (Supplementary Table S1).

Phreatobacter (Tóth et al., 2014; Lee S.D. et al., 2017) was as yet not assigned to a family, whereas Blastochloris (Hiraishi, 1997) was assigned by some authors to Hyphomicrobiaceae (Garrity et al., 2003b), although other studies concluded that this genus cannot safely be assigned to a family based on 16S rRNA gene data (Lee et al., 2005). The GBDP tree showed both genera located within a highly supported clade also comprising Bradyrhizobiaceae and Xanthobacteraceae but the interrelationships between these four subclades remained unresolved (Figure 3). Similarly, the 16S rRNA gene analyses did not resolve the placement of these two isolated genera and did not indicate the affiliation to any existing family (Supplementary File S2). For this reason, we propose to assign each of the two genera to a family of its own. This solution is not precluded by the phenotype of the two genera (Supplementary Table S1).

The CCT (Supplementary File S2) called the assignment of Rhodoplanes (Hiraishi and Ueda, 1994b) to Hyphomicrobiaceae into question. Rhodoplanes was placed in Hyphomicrobiaceae by Bergey’s Manual (Boone et al., 2001b) but the reason behind this decision remained unclear. In the CCT Rhodoplanes formed a clade together with genera of Bradyrhizobiaceae (Garrity et al., 2005e) such as Variibacter (Kim et al., 2014), Afipia (Brenner et al., 1991) and Nitrobacter (Winogradsky, 1892) Pseudolabrys (Kämpfer et al., 2006) and Pseudorhodoplanes (Tirandaz et al., 2015) appeared as closest relatives in the CCT. Because of low clade support we propose to tentatively assign Rhodoplanes to the same family as these genera; the nomenclature of Bradyrhizobiaceae is treated in detail below.

In the CCT (Supplementary File S2) Angulomicrobium (Vasil’eva et al., 1980) and Methylorhabdus (Doronina et al., 1995) of Hyphomicrobiaceae formed a strongly supported clade within Xanthobacteraceae. When Angulomicrobium was proposed it was not assigned to a family but later on Bergey’s manual (Boone et al., 2001a) listed Angulomicrobium as a genus of Hyphomicrobiaceae. This was confirmed by 16S rRNA gene and lipid-composition analysis (Fritz et al., 2004) but branch support in the presented 16S rRNA gene tree was not shown and the analysis did not consider Xanthobacter (Wiegel et al., 1978), which is now the type genus of Xanthobacteraceae. When Methylorhabdus was proposed it was not assigned to a family but later on Bergey’s manual (Boone et al., 2001a) listed Methylorhabdus as a genus of Hyphomicrobiaceae. A more recent edition of Bergeys manual (Brenner et al., 2005) mentioned that the highest similarity in DNA:DNA hybridization of Methylorhabdus was observed with Xanthobacter and only 10% with Hyphomicrobium. Due to their position in the CCT we propose to assign Angulomicrobium and Methylorhabdus to Xanthobacteraceae. This solution is not in conflict with their phenotype (Supplementary Table S1).

Xanthobacteraceae (Lee et al., 2005) appeared as both paraphyletic and polyphyletic in the GBDP tree and the CCT (Figure 3 and Supplementary File S2) because Pseudoxanthobacter (Arun et al., 2008) was placed apart from the remaining Xanthobacteraceae while genera of other families, such as Aquabacter spiritensis of Hyphomicrobiaceae, were placed within Xanthobacteraceae. The original description of Pseudoxanthobacter did not explicitly assign the genus to a family, and the shown phylogeny suffered from low support at the backbone. The assignment of Pseudoxanthobacter to Xanthobacteraceae occurred only later (Ueki et al., 2010) based on mere sequence similarity search. According to the relatively isolated and not maximally supported position of Pseudoxanthobacter in the phylogenomic trees (Figure 3 and Supplementary File S2) and the lack of evidence supporting its inclusion in Xanthobacteraceae it is proposed to assign it to a family of its own. This solution is not in conflict with the phenotype (Supplementary Table S1) and not in conflict with the gene-content analysis (Supplementary File S2).

While Pleomorphomonas (Xie and Yokota, 2005) was taxonomically assigned to Methylocystaceae, it here appeared as only distantly related to the type genus of the family (Figure 3). In contrast, a highly supported clade included both Pleomorphomonas and a set of genera not yet assigned to family, Hartmannibacter (Suarez et al., 2014), Methylobrevis (Poroshina et al., 2015), Mongoliimonas (Xi et al., 2017) and Oharaeibacter (Lv et al., 2017). The 16S rRNA gene analyses (Supplementary File S2) also provided support for this clade, additionally including Chthonobacter (Kim et al., 2017). In line with the taxonomic consequences proposed above, the phylogenetic results call for suggesting a new family for these genera. Although the gene-content analysis did not provide support for the group, the establishment of this new family did not appear to be in conflict with the phenotype of the involved genera (Supplementary Table S1).

Bradyrhizobiaceae (Garrity et al., 2005e) and Xanthobacteraceae (Lee et al., 2005) were shown as non-monophyletic in the GBDP tree (Figure 3) due to the position of Pseudolabrys (Kämpfer et al., 2006) of Xanthobacteraceae, which was placed within Bradyrhizobiaceae together with Pseudorhodoplanes (Tirandaz et al., 2015), a genus that was as yet not assigned to a family. Bradyrhizobiaceae is actually illegitimate as this family includes Nitrobacter (Buchanan, 1917), which is the type genus of Nitrobacteraceae (Buchanan, 1917), which has priority (Tindall, 2019a). Our analyses do not call for placing Nitrobacter and Bradyrhizobium in distinct families (Figure 3 and Supplementary File S2), hence we will below propose an emended description of Nitrobacteraceae as the correct name for the family of the genera currently assigned to the illegitimate Bradyrhizobiaceae. The following description of the results will nevertheless use the name Bradyrhizobiaceae throughout because this is the name used in our input data but the name will be marked as in need of a replacement.

Bradyrhizobiaceae (Nitrobacteraceae) also appeared as non-monophyletic in the GBDP tree (Figure 3) because the genera Rhodoblastus (Imhoff, 2001) and Bosea (Das et al., 1996; La Scola et al., 2003) were placed apart from core Bradyrhizobiaceae (Nitrobacteraceae). Rhodoblastus formed the sister group of Roseiarcus (Kulichevskaya et al., 2014) which is currently taxonomically assigned to its own family. This clade in turn formed the sister group of core Methylocystaceae including its type genus Methylocystis (Bowman et al., 1993), whereas the subsequent sister taxon was Beijerinckiaceae (Garrity et al., 2005d). In the CCT, Rhodoblastus was shown as closely related to Roseiarcus with strong support whereas the relationship of this clade to Alsobacter (Bao et al., 2014) was only poorly supported. While Roseiarcus was taxonomically placed in Roseiarcaceae when it was originally proposed, in its original description Rhodoblastus was not assigned to a family, and a phylogenetic analysis was not performed. Rhodoblastus was placed in Bradyrhizobiaceae (Nitrobacteraceae) in a later study (Garrity et al., 2003b) but the rationale behind this decision remained obscure. When Roseiarcaceae was proposed, Rhodoblastus was considered but the backbone of the 16S rRNA gene tree was only partially resolved. Roseiarcaceae has priority over Alsobacteraceae (Sun et al., 2018), hence the possibility of a later unification of the family does not preclude the assignment of Rhodoblastus to Roseiarcaceae, which we propose below. This solution is not conflict with the phenotype either (Supplementary File S2).

Bradyrhizobiaceae (Nitrobacteraceae) also appeared in the CCT as the taxonomic home for Pseudorhodoplanes (Tirandaz et al., 2015), which was originally not assigned to a family, as well as for Pseudolabrys (Kämpfer et al., 2006) of Xanthobacteraceae. When the latter was proposed the 16S rRNA tree presented showed poor support for the specific placement and the genus was only assigned to the class Alphaproteobacteria in general. Pseudolabrys was later on placed in Xanthobacteraceae (Lin et al., 2015) based on a 16S rRNA gene analysis with low support. The CCT (Supplementary File S2) indicated that Pseudolabrys forms a clade with Bradyrhizobiaceae (Nitrobacteraceae) genera (Garrity et al., 2005e) namely Variibacter (Kim et al., 2014), Afipia (Brenner et al., 1991) and Nitrobacter (Winogradsky, 1892) Pseudorhodoplanes (Tirandaz et al., 2015) appeared as the sister genus of Pseudolabrys with strong support, even in the gene-content analysis. We thus propose to include Pseudolabrys in Bradyrhizobiaceae (Nitrobacteraceae), which is not precluded by its phenotype (Supplementary Table S1).

Bosea as well as Salinarimonas (Liu J.-H. et al., 2010) which was originally (Cai et al., 2011b) also assigned to Bradyrhizobiaceae (Nitrobacteraceae) but later on to a family of its own (Cole et al., 2018), were placed within a strongly supported clade (Figure 3 and Supplementary File S2) containing Methylobacteriaceae, Camelimonas (Kämpfer et al., 2010b) Chelatococcus asaccharovorans (Auling et al., 1993), and Chelatococcus sambhunathii (Panday and Das, 2010). A similar arrangement was observed in the CCT with high support (Supplementary File S2). In the original description of Bosea, bootstrapping was not conducted, and only few species could be considered at that time. Neither its original description nor its emendation explicitly assigned Bosea to a family. The genus was placed in Bradyrhizobiaceae (Nitrobacteraceae) in Bergey’s manual (Boone et al., 2001a) but the rationale was unclear. Other studies emphasized the uncertain placement of Bosea in 16S rRNA gene trees, much like the placement of Chelatococcus (Lee et al., 2005). When Chelatococcus was proposed and emended (Auling et al., 1993; Yoon et al., 2008) it already appeared as closely related to Methylobacteriaceae with moderate support. Chelatococcus was placed in Beijerinckiaceae in Bergey’s manual but the reason for this decision remained unclear. Camelimonas was regarded as belonging to Beijerinckiaceae when Thalassocola was proposed (Lin et al., 2015) but the included 16S rRNA gene tree did not show a monophyletic Beijerinckiaceae. In contrast, the recent proposal to place Chelatococcus and Camelimonas in a separate family Chelatococcaceae (Dedysh et al., 2016) is in agreement with our results. Since the phylogenetic placement of Bosea does not allow for an inclusion of the genus into any of the families within the clade, the taxonomically most conservative solution is to create a new family, Boseaceae fam. nov.

When Enterovirga (Chen et al., 2017d) was proposed it was acknowledged that this genus shows high 16S rRNA similarities with several Methylobacteriaceae genera but also with Chelatococcus and Pseudochelatococcus of Beijerinckiaceae and Bosea of Bradyrhizobiaceae (Nitrobacteraceae). However, the presented 16S rRNA gene tree had insufficient support for assigning Enterovirga to a family. In the CCT (Supplementary File S2) Enterovirga was nested within Methylobacteriaceae (Garrity et al., 2005c) in a moderately supported clade. Given its phylogenetic position, it is thus proposed to tentatively include Enterovirga in Methylobacteriaceae, which is not precluded by its phenotype (Supplementary Table S1).

The original descriptions of Cucumibacter, Maritalea and Pelagibacterium (Figure 4) provided strong support for a clade comprising these genera and Devosia. In these studies, the overall taxon sampling was insufficient to safely assign these genera to a family, even though the affiliation to Hyphomicrobiaceae was not called into question. When Devosia was originally proposed it was not assigned to a family, and even its later emendations (Rivas et al., 2003; Yoo et al., 2006; Yoon et al., 2007c; Zhang D.C. et al., 2012) hesitated to explicitly suggest an affiliation of the genus to a family. Previous studies already concluded that Devosia and Prosthecomicrobium cannot safely be assigned to a family based on 16S rRNA gene data (Lee et al., 2005). The CCT and UCT provided strong support for a clade comprising Arsenicitalea (Mu et al., 2016) Cucumibacter, Devosia, Maritalea, Methyloterrigena (Kim H.S. et al., 2016), Paradevosia (Geng et al., 2014), Pelagibacterium and Youhaiella (Wang Y.X. et al., 2015). It makes phylogenetic sense to taxonomically remove these genera from Hyphomicrobiaceae and assign them to a new family, which is not precluded by their phenotype (Supplementary Table S1).

Rhodobacteraceae (Garrity et al., 2005b) of Rhodobacterales were shown as non-monophyletic in the GBDP tree because the genera Acuticoccus (Hou et al., 2015), Ahrensia (Uchino et al., 1998; Liu J. et al., 2016), Labrenzia (Biebl et al., 2007; Bibi et al., 2014), Nesiotobacter (Donachie et al., 2006; Garrity et al., 2007), Pannonibacter (Borsodi et al., 2003; Biebl et al., 2007), Pseudovibrio (Shieh et al., 2004), Roseibium (Suzuki et al., 2000) and Stappia (Uchino et al., 1998; Biebl et al., 2007) were placed apart from the remaining Rhodobacteraceae and within Rhizobiales (Hyphomicrobiales) instead where they also caused the non-monophyly of some families (Figure 4). It should be noted that Rhodobacteraceae as originally proposed is illegitimate because the family included Hyphomonas (Moore et al., 1984), the type genus of Hyphomonadaceae (Lee et al., 2005), which has priority.

Acuticoccus showed a relatively isolated position (Figure 4) but also strong support for its sister-group relationship to Amorphus (Zeevi Ben Yosef et al., 2008), which was as yet not assigned to a family. It is questionable whether the restricted taxon sampling in the phylogenetic analysis that accompanied the original description of Acuticoccus really allowed for an assignment to a family but the genus was taxonomically placed in Rhodobacteraceae (Hou et al., 2015). Phenotypically, the features of Acuticoccus may be rather rare in Alphaproteobacteria as gliding motility was reported for the genus, whereas Amorphus was described as not-motile (Supplementary Table S1). For reasons of taxonomic conservatism, the two genera are best assigned to the same family, which is supported by the gene-content analysis (Supplementary File S2).

The original description of Stappia did not assign it to a family; this was done in an edition of Bergey’s manual (Boone et al., 2001a) but it is not obvious whether this was based on a phylogenetic analysis. The same holds for the assignment of Labrenzia (Cai et al., 2011a), Pannonibacter (Garrity et al., 2003b), and Pseudovibrio (Garrity et al., 2007) to Rhodobacteraceae. Later studies, such as the proposal of Nesiotobacter (Donachie et al., 2006), also failed to conduct a comparison with the type genus of Rhodobacteraceae. Rather, these studies assigned newly proposed genera to Rhodobacteraceae based on the closeness of these genera to genera that now appeared to phylogenetically not belong to the family. Labrenzia, Nesiotobacter, Pannonibacter, and Roseibium formed a moderately to well supported clade even in the unconstrained 16S rRNA gene analyses (Supplementary File S2). The GBDP topology (Figure 4) differs regarding the positioning of Breoghania (Gallego et al., 2010) but this discrepancy was not well supported. When Breoghania was proposed and assigned to Cohaesibacteraceae, none of the genera with which Cohaesibacteraceae now appeared intermixed were included in the taxon sampling. Additional supermatrix analyses (Supplementary File S2) indicated that Breoghania forms the sister group of a clade comprising the six genera, whereas Cohaesibacter (Hwang and Cho, 2008a) branches first. Given the results from the analysis of the 16S rRNA gene and genome-scale data it seems advisable to propose a new family to accommodate Breoghania as well as another new family to harbour Labrenzia, Pannonibacter, Pseudovibrio (including Nesiotobacter as proposed below), Roseibium and Stappia. These solutions are not in conflict with the phenotype of the involved genera (Supplementary Table S1).

Among the misplaced Rhodobacteraceae, Ahrensia formed an independent lineage with relatively uncertain affiliations to other families (Figure 5). The original description of Ahrensia did not assign it to a family, whereas its last emendation assigned it to Phyllobacteriaceae (Liu J. et al., 2016) Bergey’s manual placed Ahrensia in Rhodobacteraceae (Boone et al., 2001a) but whether this was based on a phylogenetic assessment remained unclear to us. Similarly, the assignment of Nesiotobacter to Rhodobacteraceae (Garrity et al., 2007) may not have been based on phylogenetic argumentation. Given the phylogenomic results presented here, Ahrensia is best assigned to a separate family, which is not precluded by its phenotype (Supplementary Table S1). According to the 16S rRNA gene analyses, Pseudahrensia (Jung et al., 2012c) should also be assigned to this family (Supplementary File S2).

Phyllobacteriaceae (Mergaert and Swings, 2005) appeared as paraphyletic in the GBDP tree (Figure 5) because the genera Aminobacter (Urakami et al., 1992; Kämpfer et al., 2002a), Nitratireductor (Labbé et al., 2004; Jang et al., 2011), Pseudaminobacter (Kämpfer et al., 1999), Aquamicrobium (Bambauer et al., 1998; Lipski and Kämpfer, 2012; Wu Z.-G. et al., 2014), Mesorhizobium (Jarvis et al., 1997), Hoeflea (Peix et al., 2005; Rahul et al., 2015) and Zhengella (Liao et al., 2018) were placed apart from the clade containing the type genus of the family, Phyllobacterium. Zhengella formed the sister group of Notoacmeibacter (Huang et al., 2017) of Notoacmeibacteraceae with reasonable support, even in the gene-content analysis (Supplementary File S2). The taxonomically most conservative solution is to also assign Zhengella to this family, which is not precluded by the major phenotypic features found in the genus (Supplementary Table S1).

The original descriptions of Mesorhizobium and Pseudaminobacter (Figure 5) did not explicitly assign these genera to a family. The recent emendations of Aquamicrobium (Figure 5) were accompanied by phylogenetic trees with low support at the backbone and a taxon sampling which we tend to regard as insufficient for safely assigning a genus to a family. The three genera were placed in Phyllobacteriaceae in an edition of Bergey’s manual (Boone et al., 2001a) but it remained unclear to us whether this was based on a phylogenetic assessment. Nitratireductor (Figure 5) was assigned to Phyllobacteriaceae when it was originally proposed (Labbé et al., 2004) but this was based on a poorly resolved 16S rRNA gene tree with reduced taxon sampling. When Aquamicrobium (Bambauer et al., 1998) was proposed, bootstrapping was not conducted, and at that time only few species could be considered in the phylogenetic analysis. Insufficient taxon sampling was also present in the original description of Nitratireductor, in addition to poor branch support; the most recent emendation of the genus did not include the type species of the type genus of the family either. In the CCT, Nitratireductor, Pseudaminobacter, Aminobacter, Aquamicrobium, and Mesorhizobium formed a clade together with further genera such as Chelativorans (Doronina et al., 2010; Kämpfer et al., 2015a) and Carbophilus (Meyer et al., 1993) but branch support was extraordinarily low throughout (Supplementary File S2). For this reason, we refrain from taxonomic conclusions for these genera. The issue should be tackled once more type-strain genome sequences from the clade become available.

During our investigation we noticed that Chelativorans intermedius (Kämpfer et al., 2015a) (Supplementary File S2) was described as Gram-positive although the opposite is mentioned in the abstract of the same publication. Furthermore the description for Chelativorans (Doronina et al., 2010; Kämpfer et al., 2015a) defines the genus as Gram-negative. Gram-positivity would be unlikely for the whole family of Phyllobacteriaceae (Brenner et al., 2005) in general. We suspect this was an oversight of the authors and Chelativorans intermedius is most likely Gram-negative.

When Hoeflea was proposed, it was included in the family Phyllobacteriaceae even though this conclusion remained phylogenetically unsupported (Peix et al., 2005). Our phylogenomic analysis suggests the inclusion of Hoeflea in Rhizobiaceae (Figure 5 and Supplementary File S2). While the marine bacterium Hoeflea differs from Rhizobiaceae also regarding the presence of photosynthetic pigments, this character is highly homoplastic within the class Alphaproteobacteria. Assigning Hoeflea to Rhizobiaceae is not in conflict with its phenotype (Supplementary Table S1) and more conservative than generating a new family to accommodate the genus.

In the CCT (Supplementary File S2) Lentilitoribacter (Park et al., 2013b) of Phyllobacteriaceae (Mergaert and Swings, 2005) was nested within Hoeflea (Peix et al., 2005; Rahul et al., 2015) with reasonable support. When Lentilitoribacter was originally proposed it was placed in Phyllobacteriaceae in a 16S rRNA gene tree which lacked support and did not include H. phototrophica, which formed the sister group of Lentilitoribacter in the CCT. Although we refrain from resolving the paraphyly of Hoeflea until additional genomic data become available we propose to include Lentilitoribacter in Rhizobiaceae, which is not precluded by its phenotype (Supplementary File S1).

Aurantimonadaceae (Garrity et al., 2003a) were shown as polyphyletic in the GBDP tree (Figures 4, 5) because Martelella endophytica (Bibi et al., 2013) and Martelella mediterranea (Rivas et al., 2005) were placed apart from core Aurantimonadaceae (Figure 4) and within a strongly supported clade (Figure 5) that included Rhizobiaceae and Mycoplana of Brucellaceae. When Martelella and its type species, M. mediterranea, were proposed, they were not assigned to any family and phylogenetically not placed within Aurantimonadaceae but rather within Rhizobiaceae. Martelella was placed in Aurantimonadaceae later on (Garrity et al., 2007) but the rationale behind this decision remained obscure. The CCT did not indicate the monophyly of Aurantimonadaceae either (Supplementary File S2). Consequently, it is proposed to include Martelella into Rhizobiaceae. This is not contradicted by the phenotypic characteristics (Supplementary Table S1). Aurantimonadaceae was originally proposed without a formal description and is not a validly published name. Therefore we newly propose Aurantimonadaceae, fam. nov., with a revised (reduced) set of genera to belong to this family.

The position of Mycoplana dimorpha (Urakami et al., 1990) which appeared nested within Rhizobiaceae (Conn, 1938) in Figure 5, also rendered Brucellaceae (Breed et al., 1957) polyphyletic in the GBDP tree; core Brucellaceae is shown in Figure 4. In the CCT, in addition to Mycoplana dimorpha, Mycoplana ramosa (Urakami et al., 1990) was also placed within the clade containing Rhizobiaceae (Supplementary File S2). Mycoplana (Gray and Thornton, 1928; Urakami et al., 1990) was originally proposed on basis of phenotypic characteristics without taking into account the 16S rRNA gene sequence as phylogenetic marker; it was even supposed to belong to Mycobacteriaceae (Actinobacteria). Later on Mycoplana was placed in Brucellaceae (Boone et al., 2001a) but the rationale behind this assignment remained unclear. Mycoplana displays phenotypic characteristics similar to those of Rhizobiaceae (Supplementary Table S1). Given its phylogenetic positioning, it is thus proposed to include Mycoplana in Rhizobiaceae.

Notoacmeibacter (Huang et al., 2017) not only formed a strongly supported clade together with Mabikibacter of Mabikibacteraceae (Choi et al., 2017) in the CCT but the distance between the two 16S rRNA gene sequences was exceptionally low (Supplementary File S2). The monotypic family Mabikibacteraceae was proposed briefly after Notoacmeibacteraceae hence neither publication was able to account for the respective other taxon. The high similarity of 99.9% in the 16S rRNA gene sequences (Meier-Kolthoff et al., 2013b) and the highly similar phenotypic characteristics (Supplementary File S1) could even indicate identity at the species level but solving this issue required genome sequencing of the type strain of Mabikibacter. We thus here only propose to transfer Mabikibacter to Notoacmeibacteraceae, which has priority. This transfer would also avoid the non-monophyly of Notoacmeibacteraceae that resulted if only Zhengella were to be added to this family (as treated above).

Rhizobiales (Hyphomicrobiales) Genera

In addition to some families of Rhizobiales (Hyphomicrobiales), a variety of genera assigned to the order were found in need of a taxonomic revision, while the majority of the genera was shown as monophyletic (Figures 3–5). This section also treats the Rhodobacterales genera that needed to be taxonomically assigned to Rhizobiales (Hyphomicrobiales) as explained above in case their genus boundaries also needed to be reconsidered.

Within Rhizobiales (Hyphomicrobiales), Afipia (Brenner et al., 1991; La Scola et al., 2002) was shown as paraphyletic in the GBDP tree (Figure 3) because Oligotropha carboxidovorans (Meyer et al., 1993) formed the sister group of the type species of Afipia, Afipia felis (Brenner et al., 1991). O. carboxidovorans was originally proposed as a new species in a new genus on the basis of morphological, physiological and biochemical characteristics, as well as 16S rRNA signature oligonucleotides. However, in the original publication no phylogenetic tree was inferred and it was not clarified which of the character states represent apomorphies. The known phenotypic differences between the two genera are not pronounced (Supplementary Table S1). To include O. carboxidovorans in Afipia, which has priority, seems to be the taxonomically most conservative solution.

Methylobacterium (Patt et al., 1976) appeared non-monophyletic in the GBDP tree (Figure 3) because Methylorubrum (Green and Ardley, 2018) species such as Methylorubrum salsuginis, M. populi and M. chloromethanicum were nested within Methylobacterium. All species were only recently transferred from Methylobacterium to Methylorubrum (Green and Ardley, 2018). The purpose of splitting Methylobacterium was to obtain phenotypically more homogeneous genera. Yet if such an approach yields apparently non-monophyletic groups it is in conflict with the principles of phylogenetic systematics (Hennig, 1965; Klenk and Göker, 2010; Wiley and Lieberman, 2011). In particular, the cited study emphasized that only certain clades of Methylobacterium sensu lato should be assigned to either Methylobacterium sensu stricto or to Methylorubrum but new genera and new combinations for the remaining clades were not proposed. Since species names cannot lack a genus name, it would come as no surprise if this solution rendered Methylobacterium non-monophyletic. The utilization of methylamine as sole carbon source in Methylorubrum was suggested as the main feature for distinguishing it from Methylobacterium sensu stricto (Green and Ardley, 2018). But a single binary character cannot properly separate two taxa according to the principles of phylogenetic systematics because only one of the two character states is an apomorphy (Nouioui et al., 2018). Because of the low resolution in the 16S rRNA gene trees (Supplementary File S2) it seems more appropriate to treat the Methylorubrum species as species of Methylobacterium until more type-strain genome sequences from the group become available and the splitting of the genus can be completed.

Pseudovibrio (Shieh et al., 2004) was shown as paraphyletic in the GBDP tree (Figure 4) because Nesiotobacter exalbescens (Donachie et al., 2006) was nested within Pseudovibrio with moderate support. When Nesiotobacter was proposed, the genus Pseudovibrio was not considered for the phylogenetic analysis of the 16S rRNA gene. While our 16S gene analyses (Supplementary File S2) showed Nesiotobacter as sister group of Pseudovibrio, Nesiotobacter exalbescens and Pseudovibrio display similar phenotypic characteristics (Supplementary Table S1). The gene-content analysis provided support for the entire clade but not for all of the subclades (Supplementary File S2). Therefore, we propose to include N. exalbescens into Pseudovibrio because the rationale of separating the two genera was not obvious.

Labrenzia (Biebl et al., 2007; Bibi et al., 2014) appeared as paraphyletic in the GBDP tree (Figure 4) because Roseibium denhamense (Suzuki et al., 2000; Biebl et al., 2007) and R. hamelinenseas (Suzuki et al., 2000; Biebl et al., 2007) were nested within Labrenzia, with L. suaedae (Bibi et al., 2014) branching first. Pannonibacter (Borsodi et al., 2003; Biebl et al., 2007) formed the sister group of the entire clade. In the CCT, Roseibium aquae (Zhong et al., 2014) and R. sediminis (Liu et al., 2017) were also placed within Labrenzia (Supplementary File S2). While the original proposals of Roseibium and L. suaedae showed phylogenetic trees with 83–93% support for the monophyly of Roseibium, the species now included in Labrenzia could not be considered when Roseibium was introduced. Support was already weak for the monophyly of Labrenzia when it was originally proposed. The paraphyly of Labrenzia was already evident in a more recent study (Camacho et al., 2016), which also showed an uncertain position of L. suaedae relative to the remaining genus. The Roseibium species display phenotypic characteristics similar to those of Labrenzia (Supplementary Table S1) while the gene-content analysis did not yield well-supported subgroups that could be proposed as genera (Supplementary File S2). For this reason, the best solution appears to be to merge the two genera, which is also the taxonomically most conservative approach.

Ochrobactrum (Holmes et al., 1988) was shown as paraphyletic in the GBDP tree (Figure 4) because the entire genus Brucella (Meyer and Shaw, 1920) was nested within Ochrobactrum. O. thiophenivorans (Kämpfer et al., 2008), O. grignonense (Lebuhn et al., 2000), O. pituitosum (Huber et al., 2010), O. rhizosphaerae (Kämpfer et al., 2008), O. pseudogrignonense (Kämpfer et al., 2007a) formed a clade that branched first, whereas the type species of Ochrobactrum appeared more closely related to Brucella than to this clade. In the CCT, O. haematophilum (Kämpfer et al., 2007a) and O. pecoris (Kämpfer et al., 2011) also belonged to the clade that branched first. The 16S rRNA gene data neither provided evidence for other Ochrobactrum species to form a clade with the type species to the exclusion of Brucella, including species such as O. pseudintermedium (Teyssier et al., 2007), O. oryzae (Tripathi et al., 2006), O. gallinifaecis (Kämpfer et al., 2003) and O. endophyticum (Li L. et al., 2016) (Supplementary File S2). When Ochrobactrum was proposed, an analysis of the 16S rRNA gene could not yet be conducted. When Brucella was proposed, a phylogenetic analysis was not conducted; it was speculated that the genus may belong to Enterobacteriaceae but an assignment to a family was not explicitly suggested. Among those taxonomic studies available to us that showed a phylogenetic analysis with a sufficient number of Brucella and Ochrobactrum species, a paraphyletic Ochrobactrum was always evident, particularly in the recent proposals of new species (Kämpfer et al., 2007a, b, 2008, 2009a, 2010a, 2013b, 2014; Lebuhn et al., 2000; Trujillo et al., 2005; Teyssier et al., 2007; Romanenko et al., 2008; Xu H.Y. et al., 2009; Huber et al., 2010; Imran et al., 2010; Woo et al., 2011; Li L. et al., 2016; Liu B.B. et al., 2016).

It is well known that in terms of DDH similarity, Brucella is only a single species (Verger et al., 1985); this is also reflected in the newly calculated dDDH values (Table 1). The overall genomic divergence of the Brucella-Ochrobactrum clade was lower than in many clades harboring a single genus only. Brucella differs from Ochrobactrum regarding its pathogenic lifestyle, which may be reflected in the lower genome size of Brucella (Figure 4). However, Ochrobactrum species are also known from clinical specimens, including its type species (Holmes et al., 1988), and a more pronounced genome-size reduction of pathogenic species nested within a partially non-pathogenic genus was observed elsewhere, as, e.g., in Mycobacterium leprae (Akinola, 2013). Mycobacterium can also serve as an example for a genus that harbours distinct risk groups (Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, 2015), much like Burkholderia and Yersinia. Hence the difference between Brucella and Ochrobactrum regarding their risk-group assignment could hardly be used as an argument against their inclusion in the same genus. Known phenotypic differences, if any, appeared to be restricted to autapomorphies of Brucella that may well be linked to its evolutionary adaptation to pathogenesis (Supplementary Table S1). Despite the differences in genome size, the gene-content analysis provided more support for the combined Brucella–Ochrobactrum clade than for the subclades (Supplementary File S2). For these reasons, we propose to include Ochrobactrum in Brucella, which has priority. According to the available phenotypic information (Supplementary Table S1), the inclusion of the entire genus Ochrobactrum does not cause a need for the emendation of Brucella (Meyer and Shaw, 1920; Verger et al., 1985).

Mesorhizobium (Jarvis et al., 1997) appeared as non-monophyletic in the GBDP tree and in the CCT (Figure 5 and Supplementary File S2) since several Mesorhizobium species were intermixed with genera such as Aquamicrobium (Bambauer et al., 1998; Lipski and Kämpfer, 2012; Wu Z.-G. et al., 2014) Nitratireductor (Labbé et al., 2004) and Pseudaminobacter (Kämpfer et al., 1999) which were also shown as non-monophyletic. However, due to the low support of the branches, particularly in the comprehensive 16S rRNA gene trees, it is difficult to propose taxonomic conclusions for Aquamicrobium, Mesorhizobium, Nitratireductor, and Pseudaminobacter. These genera belong to a deviating branch of Phyllobacteriaceae that even proved to be difficult to classify at the family level, as explained above. We suppose that a more satisfying classification of these four genera can only be obtained once more genome sequences become available.

Allorhizobium (de Lajudie et al., 1998) and Pararhizobium (Mousavi et al., 2015) appeared as polyphyletic in GBDP and 16S rRNA gene trees (Figure 5 and Supplementary File S2) because several species are mixed up with representatives of Rhizobium (Frank, 1889). When Allorhizobium was proposed, a tree with rather low support at backbone was presented which did not yield well-defined relationships between the genera. Later studies of the group presented much better resolved multi-gene analyses but restricted the taxon sampling to rhizobial and agrobacterial strains (Mousavi et al., 2015). The CCT included a well-supported clade including the type species of Allorhizobium, A. undicola, the other Allorhizobium species except for A. oryzae, as well as Rhizobium oryziradicis (Zhao et al., 2017), and R. taibaishanense (Yao et al., 2012). Rhizobium (Frank, 1889; Young et al., 2001) was shown as non-monophyletic for various reasons in the GBDP tree (Figure 5) and the CCT, including distantly placed species such as R. arenae (Zhang S. et al., 2017), R. flavum (Gu et al., 2014), R. gei (Shi et al., 2016), R. marinum (Liu Y. et al., 2015), R. selenitireducens (Hunter et al., 2007), R. naphthalenivorans (Kaiya et al., 2012), R. subbaraonis (Ramana C.V. et al., 2013) which were placed apart from the well-supported core Rhizobium clade that contained the type species of the genus.

The taxonomic literature disagreed on whether a broad concept of Rhizobium, i.e., including genera such as Allorhizobium and Agrobacterium (Conn, 1942), should be preferred (Young et al., 2001, 2003) or whether Rhizobium should be split into various genera (Farrand et al., 2003; Mousavi et al., 2015). Apparently distinct authors of new species names followed either one or the other of the two competing concepts (Willems, 2006), thus leading to the considerable mix-up of the species of Rhizobium (Figure 5). Given the overall genomic divergence of the group revealed here it appeared advisable to follow a narrow concept for Rhizobium. Further dissecting Rhizobium does not imply introducing many new genus names because names for the respective clades were already proposed in the literature. For instance, R. oryziradicis and R. taibaishanense can be assigned to the already existing genus Allorhizobium (Figure 5 and Supplementary File S2), which is not in disagreement with their phenotype (Supplementary Table S1) while the group is even supported in the gene-content analysis (Supplementary File S2). The new combination Allorhizobium taibaishanense was already proposed (Mousavi et al., 2015) but the name does not appear to be validly published.

On a similar vein, R. subbaraonis (Ramana C.V. et al., 2013) should be assigned to Mycoplana (Urakami et al., 1990; Gray and Thornton, 1928) whereas the phylogenetically relatively isolated Pararhizobium haloflavum (Shen et al., 2018) should be placed into a genus of its own (Figure 5, Supplementary File S2, and Supplementary Table S1). The inclusion of Rhizobium subbaraonis does not cause a need for the emendation of Mycoplana (Supplementary Table S1). The original description of R. subbaraonis did not include Mycoplana, whereas only a limited total number of species was considered when P. haloflavum was originally proposed.

The taxonomy of deviating species of Allorhizobium such as A. borbori, A. oryzae and A. pseudoryzae as well as of the deviating Pararhizobium species P. capsulatum (Mousavi et al., 2015) cannot currently be improved because of low resolution in the comprehensive 16S rRNA gene trees (Supplementary File S2). Similarly, while Rhizobium species such as R. smilacinae (Zhang et al., 2014b), R. cellulosilyticum (García-Fraile et al., 2007), R. zeae (Celador-lera et al., 2017), R. wenxiniae (Gao et al., 2017) and R. yantingense (Chen W. et al., 2015) were nested within Neorhizobium (Mousavi et al., 2014), the clade containing both Neorhizobium and Pseudorhizobium (Kimes et al., 2015) was particularly poorly supported (Supplementary File S2). While Rhizobium species such as R. naphthalenivorans and R. selenitireducens could possibly be placed in Ciceribacter (Kathiravan et al., 2013) and others such as R. arenae and R. gei could potentially be included in Pararhizobium, these arrangements were also poorly resolved. Therefore it would currently be difficult to infer taxonomic conclusions. We suppose that a more satisfying classification of these genera can be obtained once more genome sequences become available. As an exception, Rhizobium vignae (Ren et al., 2011), which was placed as sister group of N. galegae (Mousavi et al., 2014) with strong support, can be assigned to Neorhizobium. While they are closely related, our dDDH results (41.4% similarity) values indicate that R. vignae and N. galegae are independent species. Therefore, we propose to include R. vignae in Neorhizobium. This proposal is not contradicted by phenotypic characteristics and the transfer of R. vignae does not cause a need for the emendation of Neorhizobium (Supplementary Table S1). The removal of R. oryziradicis, R. taibaishanense, R. subbaraonis and R. vignae does neither cause a need for the emendation of Rhizobium (Frank, 1889).

Rhodobacterales Genera

Rhodobacterales has so far been a monotypic order that only included Rhodobacteraceae. As mentioned above, this family as originally proposed is illegitimate because the family included Hyphomonas (Moore et al., 1984), the type genus of Hyphomonadaceae (Lee et al., 2005), which has priority. Except for the genera taxonomically assigned to Rhodobacterales but phylogenetically placed in Rhizobiales (Hyphomicrobiales), which were treated above, Rhodobacteraceae was monophyletic. However, whereas most of its genera were also shown as monophyletic in our analyses (Figures 7–9), some were shown to be in need of a taxonomic revision.

Roseivivax (Suzuki et al., 1999a; Park et al., 2010; Chen M.-H. et al., 2012) was shown as non-monophyletic in the GBDP tree and the CCT (Figures 7, 9 and Supplementary File S2) because R. roseus (Zhang et al., 2014c) formed a clade (Figure 7) together with Tranquillimonas alkanivorans (Harwati et al., 2008) while R. pacificus (Wu et al., 2013) appeared (Figure 9) as sister group of Citreimonas salinaria (Choi and Cho, 2006). When R. roseus was originally proposed, it formed a group with R. pacificus which received only poor support in 16S rRNA gene trees, much like its relationship to core Roseivivax. Similarly, the phylogenetic position of T. alkanivorans and C. salinaria remained unresolved in these trees. The isolated position of the species in the 16S rRNA gene and genomic trees calls for the proposal of a new genus to accommodate R. pacificus. In addition, we propose to include R. roseus in Tranquillimonas. This is not contradicted by the phenotypic characteristics of these taxa (Supplementary Table S1) even though the gene-content analysis does not lend support (while core Roseivivax is reasonably supported; Supplementary File S2). The removal of R. roseus and R. pacificus does not cause a need for the emendation of Roseivivax (Supplementary Table S1).

FIGURE 9

Defluviimonas (Foesel et al., 2011; Math et al., 2013) appeared as paraphyletic in the GBDP tree (Figure 7) because particularly Defluviimonas alba (Pan et al., 2015) but also D. indica (Jiang et al., 2014) were placed apart from the type species of Defluviimonas, D. denitrificans, which appeared more closely related to the type species of Albidovulum, A. inexpectatum (Albuquerque et al., 2002) whereas Albidovulum xiamenense (Yin et al., 2012) was in turn placed apart from its type species. In the original proposal of A. xiamenense, only weak support was obtained for the monophyly of Albidovulum. When D. indica was proposed, 90% support was obtained for the monophyly of Defluviimonas in a 16S rRNA gene analysis based on the neighbor-joining algorithm and a simplistic evolutionary model. We could not reproduce this finding here based on our ML analysis with a much larger taxon sampling (Supplementary File S2) which, even when unconstrained, placed D. indica as sister group of the clade comprising A. inexpectatum and core Defluviimonas. In the 16S rRNA gene trees (Supplementary File S2) D. alba was placed as sister group of Frigidibacter albus (Li and Zhou, 2015). When D. alba was proposed, no support for the monophyly of Defluviimonas was obtained, as the phylogenetic position of D. alba remained uncertain; Albidovulum was not considered. D. alba and F. albus show similar phenotypic characteristics (Supplementary Table S1). Therefore, one could propose to include D. alba in Frigidibacter. However, the 16S rRNA gene similarity between the two type strains is 99.79% when calculated using the recommended settings, which indicates that DNA:DNA hybridization should be carried out to rule out that the strains are conspecific (Meier-Kolthoff et al., 2013b). In contrast, we conclude that a new genus should be proposed to accommodate D. indica, which is also in accordance with the gene-content analysis (Supplementary File S2) as it yielded strong support for the remaining clade after removal of D. indica. Including, D. aestuarii, D. aquaemixtae, D. denitrificans, and D. nitratireducens in Albidovulum is a phylogenetically obvious solution but would render the remaining species of the genus, D. alba and D. pyrenivorans, illegitimate, whose phylogenetic position is still uncertain. In contrast, its isolated phylogenetic position indicates that Albidovulum xiamenense can safely be placed into a genus of its own, which is not precluded by its phenotype. The removal of A. xiamenense does not cause a need for the emendation of Albidovulum (Supplementary Table S1).

Gemmobacter (Rothe et al., 1987; Chen W.-M. et al., 2013) was shown as polyphyletic in the GBDP tree (Figure 7) because G. nectariphilus (Tanaka et al., 2004; Chen W.-M. et al., 2013) and G. megaterium (Liu J.-J. et al., 2014) were placed in a distant position relative to the type species, G. aquatilis. In addition, in the CCT G. intermedius (Kämpfer et al., 2015b) and G. straminiformis (Kang et al., 2017) formed an unsupported group together with G. nectariphilus and G. megaterium (Supplementary File S2). When these Gemmobacter species were proposed the monophyly of the genus was already unsupported in 16S rRNA gene trees. We did not detect an obvious synapomorphy for the genus as currently circumscribed (Supplementary Table S1). Therefore, a new genus is proposed to accommodate G. megaterium and G. nectariphilus; G. intermedius, and G. straminiformis should also be assigned to this genus once their genome sequences confirm their position, which is currently uncertain in 16S rRNA gene trees. The removal of G. megaterium and G. nectariphilus does not cause a need for the emendation of Gemmobacter (Supplementary Table S1). The two resulting genera received high support in the gene-content analysis (Supplementary File S2).

Pseudorhodobacter (Uchino et al., 2002; Jung et al., 2012a; Chen C.-X. et al., 2013; Lee et al., 2013, 2016) appeared as paraphyletic in the GBDP tree (Figure 7) because P. psychrotolerans (Lee et al., 2016) was placed as sister group of Rhodobacter blasticus and Tabrizicola. In the CCT and UCT, P. aquaticus (Li A.-H. et al., 2016), P. collinsensis (Zhang et al., 2016), P. psychrotolerans (Lee et al., 2016) and P. sinensis (Li A.-H. et al., 2016) formed a reasonably well supported clade separate from core Pseudorhodobacter (Supplementary File S2) and showed similar phenotypic characteristics (Supplementary Table S1). The descriptions of these species were accompanied by poorly resolved 16S rRNA gene trees which did not clearly indicate the monophyly of Pseudorhodobacter. Given the results presented here, it is proposed to place these deviating Pseudorhodobacter species in a separate genus.

Rhodobacter (Imhoff et al., 1984; Srinivas et al., 2007b; Wang et al., 2014) was shown to be non-monophyletic in the GBDP tree with strong support (Figure 7) because a clade containing R. veldkampii (Hansen and Imhoff, 1985) and R. vinaykumarii (Srinivas et al., 2007b) was placed apart from the clade containing the type species of the genus. In addition, R. azotoformans (Hiraishi et al., 1996), R. johrii (Girija et al., 2010), R. megalophilus (Arunasri et al., 2008), R. ovatus (Srinivas et al., 2008) and R. sphaeroides (van Niel, 1944; Imhoff et al., 1984) were placed apart from each of these two clades and formed the sister group of Cereibacter (Suresh et al., 2015) instead. Finally, R. blasticus (Kawasaki et al., 1993) appeared as sister group of Tabrizicola aquatica (Tarhriz et al., 2013). Rhodobacter already appeared non-monophyletic in the 16S rRNA gene trees presented in recent taxonomic studies such as the one that proposed Tabrizicola although phylogenetic resolution remained limited. The five Rhodobacter species placed adjacent to Cereibacter are phenotypically in agreement with this genus (Supplementary Table S1), which was also supported by the gene-content analysis (Supplementary File S2). Assigning these deviating Rhodobacter species to Cereibacter is more conservative than establishing a separate genus. Based on dDDH values, R. megalophilus is a later heterotypic synonym of R. sphaeroides (Table 1). Consequently, we propose to place R. azotoformans, R. johrii, R. ovatus, and R. sphaeroides in Cereibacter. Given their relatively isolated phylogenetic position a separate genus is proposed to accommodate R. vinaykumarii and R. veldkampii, which is not in conflict with their phenotype (Supplementary Table S1) even though the group is not supported by the gene-content analysis. Similarly, R. blasticus can be placed in Tabrizicola; Xinfangfangia (Hu et al., 2018) was also shown to be closely related to this group in the 16S rRNA gene trees (Supplementary File S2) but Tabrizicola has priority. The removal of R. azotoformans, R. johrii, R. ovatus, R. sphaeroides, R. vinaykumarii, R. veldkampii and R. blasticus does not cause a need for the emendation of Rhodobacter (Supplementary Table S1).

Paracoccus (Davis et al., 1969; Ludwig et al., 1993; Liu et al., 2008) appeared as paraphyletic in the GBDP tree (Figure 7) because Methylarcula marina (Doronina et al., 2000) was placed as sister group of Paracoccus saliphilus (Wang Y.-X. et al., 2009). In the CCT, Methylarcula terricola (Doronina et al., 2000) was also nested within the genus Paracoccus. When M. marina and M. terricola were proposed, a 16S rRNA gene phylogenetic analysis showed strong support for a clade comprising Methylarcula and Paracoccus but only weak support for the monophyly of Paracoccus to the exclusion of Methylarcula. The known phenotypic features of the two genera also fit excellently to each other (Supplementary Table S1). Thus, we conclude that Methylarcula should be included in Paracoccus, which has priority. However, since the types strains of both Methylarcula species appear to have been deposited in a single culture collection only, alternative species names cannot currently be proposed (Parker et al., 2019).

Actibacterium (Lucena et al., 2012; Guo et al., 2017) appeared as paraphyletic in the GBDP tree (Figure 8) because Confluentimicrobium lipolyticum (Park et al., 2014d) was placed as sister group of Actibacterium ureilyticum (Lin et al., 2016). In the CCT, in addition to C. lipolyticum, C. naphthalenivorans (Jeong et al., 2015) also was placed within Actibacterium (Supplementary File S2). When C. lipolyticum was proposed, its phylogenetic position could not fully be resolved by 16S rRNA gene analysis, as it appeared external to Actibacterium but with low support. C. lipolyticum and C. naphthalenivorans show phenotypic characteristics similar to those of Actibacterium (Supplementary Table S1). Therefore, it is proposed to assign C. lipolyticum and C. naphthalenivorans to Actibacterium, which has priority.

Celeribacter (Ivanova et al., 2010; Lee et al., 2012) was shown as paraphyletic in the GBDP tree (Figure 8) because C. manganoxidans (Wang L. et al., 2015) was placed in a position relatively distant to a clade harboring Pacificibacter marinus (Park et al., 2014c) together with core Celeribacter including the type species, C. marinus (Baek et al., 2014). In the CCT and UCT, C. manganoxidans was also placed apart from the other Celeribacter species without any obvious phylogenetic affiliation to another genus (Supplementary File S2). When C. manganoxidans was proposed, its phylogenetic placement was already only poorly supported. Although C. manganoxidans and Celeribacter do not display consistent phenotypic differences from each other (Supplementary Table S1), the phylogenomic results coupled with the lack of phylogenetic evidence to the contrary indicate that a new genus should be proposed to accommodate C. manganoxidans.

Celeribacter also appeared as paraphyletic in the CCT because Vadicella arenosi (Romanenko et al., 2011c) was placed within this genus with high support. When Vadicella was proposed the only phylogenetic evidence was a 16S rRNA gene tree with low backbone support. In addition at that time only Celeribacter neptunius (Ivanova et al., 2010) was considered for comparison in the 16S rRNA gene tree. Taking all other Celeribacter species and the phenotype of these species (Supplementary File S1) into account, the preferred solution is to include Vadicella in Celeribacter.

Tropicimonas (Harwati et al., 2009a; Oh et al., 2012) was shown as paraphyletic in GBDP tree (Figure 8) because Pseudoruegeria marinistellae (Zhang Y. et al., 2017) was placed as sister group of type species of Tropicimonas, T. isoalkanivorans (Harwati et al., 2009a). In addition, in CCT Pseudoruegeria aestuarii was also nested within the Tropicimonas clade (Supplementary File S2). When P. marinistellae was proposed, T. isoalkanivorans was not included in the phylogenetic analysis. Furthermore, species such as P. aestuarii (Cha et al., 2016), P. haliotis (Hyun et al., 2013b), P. lutimaris (Jung et al., 2010), and P. sabulilitoris (Park et al., 2014a) were also apart from the clade harboring the type species of the genus. However, the lack of a genome sequence for the type species of Pseudoruegeria, P. aquimaris, currently precludes re-classifications of Pseudoruegeria species, particularly because Pseudoruegeria has priority over Tropicimonas. In the CCT, Tropicimonas arenosa (Oh et al., 2016) was placed as sister group of Oceanicola litoreus (Park et al., 2013c) but supported with a low bootstrap value (Supplementary File S2). Due to the lack the genome sequences of T. arenosa, we refrain from taxonomic consequences for this species.

Oceanicola (Cho and Giovannoni, 2004) appeared as polyphyletic in the GBDP tree and CCT (Figure 8 and Supplementary File S2) because Oceanicola granulosus (Cho and Giovannoni, 2004) was placed as sister group of Roseisalinus antarcticus (Labrenz et al., 2005) with strong support (even in the gene-content analysis) whereas Oceanicola litoreus (Park et al., 2013c) was placed in a distinct clade as sister group of Celeribacter manganoxidans. While this clade had only moderate support in the GBDP tree, the tree clearly indicated that O. litoreus does not form a clade together with O. granulosus, which is the type species of the genus. Similarly, O. litoreus was placed distantly to O. granulosus in the CCT, in that case close to Tropicimonas arenosa with low support (Supplementary File S2). When O. litoreus was proposed, it was placed as sister group of the other Oceanicola species with low support in a 16S rRNA gene tree. When Roseisalinus antarcticus was originally proposed, the presented tree displayed low support at the backbone, and the sampling did not include Oceanicola. The DSMZ nomenclature database indicates that a total of nine Oceanicola species had been proposed in the literature all except two were assigned to other genera in later studies. As revealed here it makes sense to taxonomically assign O. litoreus to a genus of its own, which is not precluded by its phenotype (Supplementary Table S1).

Primorskyibacter (Romanenko et al., 2011b) was shown as non-monophyletic in the GBDP tree (Figure 9) because P. insulae (Park et al., 2015a) occupied a relatively isolated position, distant from core Primorskyibacter including the type species, P. sedentarius (Romanenko et al., 2011b); this arrangement was also shown in the CCT (Supplementary File S1). When P. insulae was originally proposed (Park et al., 2015a), the analysis of the 16S rRNA gene yielded only low branch support for the monophyly of Primorskyibacter. As the sister-group relationship between P. insulae and Marivita (Hwang et al., 2009a) is not supported in the phylogenomic tree, P. insulae should rather be placed in a genus of its own, which is not contradicted by the phenotype (Supplementary Table 1).

Salipiger (Martínez-Cánovas et al., 2004) appeared as paraphyletic in GBDP tree (Figure 9) because Yangia pacifica (Dai et al., 2006) was placed as sister group of S. marinus (Wirth and Whitman, 2018) with strong support. The taxonomic history of the genera is characterized by a decrease of support in 16S rRNA gene trees with an increasing number of species. When Yangia was proposed, 96% support for a clade comprising Salipiger and Roseivivax to the exclusion of Yangia was obtained. However, when Citreicella marina was proposed (Lai et al., 2011a), which was later on included in Salipiger, 98% support for a sister-group relationship with C. thiooxidans was obtained while Yangia was shown as sister group of this clade with low support. Because the type species of Salipiger, S. mucosus, was branching first within the Salipiger–Yangia clade in the GBDP tree, generating separate genera would imply reclassifying all Salipiger species except for the type species. For this reason, it is taxonomically more conservative to include Yangia in Salipiger, which is supported by the lack of significant phenotypic differences between the two genera (Supplementary Table S1), while the gene-content analysis is inconclusive (Supplementary File S2). Paraphaeobacter pallidus (Cai et al., 2017) may also have to be placed in Salipiger according to the 16S rRNA gene analyses. But due to low support and since the genome sequence of P. pallidus is not yet available, we refrain from taxonomic consequences for this species.

Tropicibacter (Harwati et al., 2009b) formed a paraphyletic group in the GBDP tree and the CCT (Figure 9 and Supplementary File S2) because T. phthalicicus (Iwaki et al., 2012) formed the sister group of Pelagimonas with strong support. When T. phthalicicus was proposed, the 16S rRNA gene tree showed only poor support for a sister-group relationship between T. phthalicicus and the type species, T. naphthalenivorans. While the gene-content analysis is inconclusive in this respect, assigning T. phthalicicus to Pelagimonas is certainly the taxonomically most conservative solution to solve the discrepancy. This is not precluded by the known major phenotypic features of these taxa, as the only known difference is motility in T. phthalicicus (Supplementary Table 1). The removal of T. phthalicicus does not cause a need for the emendation of Tropicibacter (Supplementary Table S1).

Roseovarius (Labrenz et al., 1999) was shown as paraphyletic in the GBDP tree and in the CCT (Figure 9 and Supplementary File S2) because Pelagicola litorisediminis (Park et al., 2013a) was nested within Roseovarius with a high bootstrap value. When P. litorisediminis was proposed, a phylogenetic analysis was presented that was unresolved at the backbone; in particular, the monophyly of Roseovarius was not supported at all. P. litorisediminis and Roseovarius show similar phenotypic characteristics (Supplementary Table S1). Consequently, we propose to include P. litorisediminis in Roseovarius, which is the taxonomically most conservative solution. The clade obtains some support even in the gene-content analysis (Supplementary File S2), and the inclusion of P. litorisediminis does not cause a need for the emendation of Roseovarius (Supplementary Table S1).

Lutimaribacter (Yoon et al., 2009) appeared as non-monophyletic in the GBDP tree (Figure 9) because Lutimaribacter litoralis (Iwaki et al., 2013) was placed within a strongly supported clade containing Litorimicrobium taeanense (Jin et al., 2011) and core Thalassobius (Arahal et al., 2005) to the exclusion of the type species of Lutimaribacter. The original proposal of Lutimaribacter litoralis was accompanied by a 16S rRNA gene analysis with 90% support for a clade comprising Lutimaribacter together with Oceanicola pacificus. When Litorimicrobium was proposed (Jin et al., 2011), the presented tree displayed low support at backbone and the relationships between the genera remained ambiguous. This problem was already evident in the study that proposed Thalassobius. Given the phylogenomic results and the lack of evidence to the contrary in the 16S rRNA gene data, it is proposed to include Lutimaribacter litoralis and Litorimicrobium taeanense in Thalassobius, which has priority. Even though the clade obtains no support in the gene-content analysis, this solution is not precluded by the known phenotypic features (Supplementary Table S1). The removal of Lutimaribacter litoralis does not cause a need for the emendation of Lutimaribacter (Supplementary Table S1).

Thalassobius (Arahal et al., 2005) was shown as non-monophyletic in the GBDP and 16S rRNA gene trees (Figure 9 and Supplementary File S2) because T. activus appeared more closely related to Cognatishimia (Wirth and Whitman, 2018) than to the type species, T. mediterraneus. When T. activus was proposed (Pujalte et al., 2018), it already did not form a clade together with the type species of the genus in a phylogenetic analysis of the 16S rRNA gene. Given their close phylogenetic relationship, which is strongly supported in the GBDP tree albeit unresolved in the gene-content analysis, it is proposed to include T. activus in Cognatishimia, which is not precluded by the known phenotypic features (Supplementary Table S1); among the major phenotypic features, the sole difference appears to be the lack of flagella in T. activus. Thalassobius activus was placed in Cognatishimia (Arahal et al., 2019) while the current study was under revision, hence no further taxonomic proposal needs to be made although it appears advisable to provide an emended description of Cognatishimia.

Sulfitobacter (Sorokin, 1995) was shown as a paraphyletic group in the GBDP tree and the CCT (Figure 9 and Supplementary File S2) because S. pseudonitzschiae (Hong et al., 2015) was shown to be more closely related to Ascidiaceihabitans (Kim et al., 2016b) than to the type species of Sulfitobacter, albeit with low support. The original description of S. pseudonitzschiae showed a 16S rRNA gene in which Sulfitobacter did not appear monophyletic. Two Roseobacter species were even shown as more closely related as S. pseudonitzschiae to the type species of Sulfitobacter, S. pontiacus, with strong support, whereas Ascidiaceihabitans could not be considered. The original description of Ascidiaceihabitans in turn did not consider S. pseudonitzschiae. The CCT indicated Pseudoseohaeicola (Park et al., 2015b) as sister group of S. pseudonitzschiae to the exclusion of Ascidiaceihabitans but since the support was only moderate we here refrain from taxonomic proposals for Pseudoseohaeicola. The issue should be revisited once the genome of the type strain of Pseudoseohaeicola caenipelagi becomes available. Similarly, as Ascidiaceihabitans has priority over Pseudoseohaeicola, it would be safe to place S. pseudonitzschiae in Ascidiaceihabitans, but resolution is low even in the GBDP tree.

Pseudooceanicola (Lai et al., 2015) formed a polyphyletic group in the GBDP tree (Figure 9) as P. lipolyticus (Huang et al., 2018) appeared as the sister taxon of Ruegeria kandeliae (Zhang L. et al., 2018) with strong support. R. kandeliae was in turn phylogenetically located apart from core Ruegeria. The CCT showed the same relationships while the UCT was unresolved (Supplementary File S2). Given their relatively isolated position in the genome-based phylogeny, we would propose to transfer P. lipolyticus and R. kandeliae to a single new genus, which is not contradicted by their phenotype (Supplementary Table S1). While not supported by the gene-content analysis (Supplementary File S2), this solution is certainly more conservative than assigning them to two distinct genera. The removal of R. kandeliae does not cause a need for the emendation of Ruegeria (Uchino et al., 1998; Martens et al., 2006; Yi et al., 2007). In the case of P. lipolyticus, however, we observed a significant discrepancy between the 16S rRNA gene tree and the GBDP tree (Supplementary File S2). While the remainder of the used GenBank genome sequence did not show signs of contamination, the rRNA genes, which covered almost the complete contig NZ_PGTB01000197, yielded a distinct phylogenetic location. For this reason, we cannot rule out that the protein-coding genes of the genome sequence do not, in contrast to the 16S rRNA gene, originate from P. lipolyticus. Hence, we only propose a new genus for Ruegeria kandeliae.

Epibacterium (Penesyan et al., 2013) was shown as a paraphyletic group in the GBDP tree and the CCT (Figure 9 and Supplementary File S2) because all Epibacterium species except for the type species, E. ulvae, formed a strongly supported clade together with Tritonibacter (Klotz et al., 2018) and Ruegeria pelagia (Lee et al., 2007c). R. pelagia in turn appeared only distantly related to the type species of Ruegeria, R. atlantica. The taxonomically most conservative solution to this discrepancy between phylogeny and classification is to assign all deviating Epibacterium species – including E. scottomollicae (Vandecandelaere et al., 2008a; Wirth and Whitman, 2018) – to Tritonibacter, which is not precluded by their major phenotypic features (Supplementary Table S1) even though the gene-content analysis is inconclusive (Supplementary File S2). The inclusion of E. scottomollicae does not cause a need for the emendation of Tritonibacter (Supplementary Table S1).

Agrobacterium (Conn, 1942; Sawada et al., 1993) appeared as polyphyletic in the GBDP and 16S rRNA gene trees (Figure 9 and Supplementary File S2) because Agrobacterium meteori (Rüger and Höfle, 1992) was placed as a sister group of Ruegeria atlantica (Rüger and Höfle, 1992; Uchino et al., 1998; Muramatsu et al., 2007; Yi et al., 2007; Vandecandelaere et al., 2008a) with high support. A. meteori was originally proposed on basis of phenotypic characteristics without taking into account the 16S rRNA gene sequence as phylogenetic marker. The phenotype of A. meteori is quite similar to the one of Ruegeria atlantica (Supplementary Table S1), and according to an earlier study (Uchino et al., 1998) A. meteori is a later heterotypic synonym of R. atlantica (=Agrobacterium atlanticum). However, the dDDH value between their genome sequences was lower than the species boundary of 70% (Table 1). Accordingly, it is proposed to include A. meteori in Ruegeria as R. meteori. The removal of A. meteori does not cause a need for the emendation of Agrobacterium (Supplementary Table S1) and the inclusion of A. meteori does not cause a need for the emendation of Ruegeria (Uchino et al., 1998; Martens et al., 2006; Yi et al., 2007).

Aminobacter (Urakami et al., 1992; Kämpfer et al., 2002a) appeared as paraphyletic in the CCT (Supplementary File S2) because the monotypic genus Carbophilus (Meyer et al., 1993) was nested within Aminobacter with strong support. When Carbophilus was originally proposed Aminobacter was not taken into account. As the two share most phenotypic characteristics and mostly differ regarding their minor fatty acids (Supplementary Table S1) we propose to transfer Carbophilus to Aminobacter, which has priority; this is also the taxonomically most conservative solution.

Sphingomonadales Families

Sphingomonadales appeared to be a monophyletic order of Alphaproteobacteria in our analyses (Figure 6), which is in agreement with the presence of sphingolipids (Supplementary Table S1 and Supplementary File S2), a likely apomorphy of the group. The arrangement of the order into families seemed to be in need of a revision, however.

Within Sphingomonadales, Sphingomonadaceae (Kosako et al., 2000) appeared paraphyletic in the GBDP tree (Figure 6) because the clade comprising Sandarakinorhabdus cyanobacteriorum (Cai et al., 2018a), S. limnophila (Gich and Overmann, 2006; Kim M.C. et al., 2016), Sphingosinicella microcystinivorans (Maruyama et al., 2006; Geueke et al., 2007) and Pacificimonas flava (Liu K. et al., 2014), which appeared as sister group of the remaining taxa. These genera should better be placed in a separate family, which is supported by high overall genomic divergence within Sphingomonadales (Figure 6 and Supplementary File S2). The gene-content analysis did not resolve the backbone of the Sphingomonadales tree but did not yield significant conflict either. In published 16S rRNA gene trees Sphingosinicella was placed with only low support in Sphingomonadaceae (Maruyama et al., 2006). The original descriptions of Sandarakinorhabdus (Gich and Overmann, 2006) and Pacificimonas (Liu K. et al., 2014) were also accompanied by 16S rRNA gene trees with low support. In the CCT the additional genera Polymorphobacter (Fukuda et al., 2014), Sandaracinobacter (Yurkov et al., 2017) and Sphingoaurantiacus (Kim M.C. et al., 2016) appeared intermixed with the genera represented by genome sequences in the same clade and thus should tentatively also be assigned to the new family. The establishment of a new family for these genera is not in conflict with their phenotype (Supplementary Table S1).

The second conflict within Sphingomonadales was due to the fact that Erythrobacteraceae (Lee et al., 2005; Xu X.-W. et al., 2009) was nested within Sphingomonadaceae with high support in the GBDP tree (Figure 6 and Supplementary File S2). In particular, Novosphingobium (Takeuchi et al., 2001) appeared intermixed with Erythrobacteraceae. Sphingomonadaceae originally encompassed the genera currently placed in Erythrobacteraceae such as Erythrobacter and Porphyrobacter (Kosako et al., 2000) before Erythrobacteraceae were proposed (Lee et al., 2005) on the basis of the analysis of 16S rRNA gene and chemotaxonomic data. The high branch support obtained in that study for the monophyly of both Erythrobacteraceae and Sphingomonadaceae sensu stricto could not be confirmed by our analyses, which are based on a much larger taxon sampling (Supplementary File S2). The presence of pigments (including bacteriochlorophyll α) in Erythrobacteraceae which are absent in genera such as Sphingomonas was also used as argument for the separation of the two families (Lee et al., 2005). However, a single character with two character states cannot properly be used to separate two taxa because this character would yield an apomorphy for at most one of the two taxa (Nouioui et al., 2018). The sizeable overall genomic divergence within Sphingomonadales (Figure 6 and Supplementary File S2) argues against placing all of its genera into a single family. For this reason, we propose to include Novosphingobium in Erythrobacteraceae. Considering that only apomorphies can be used to justify a taxon (Hennig, 1965; Wiley and Lieberman, 2011), this proposal did not appear to be in conflict with the phenotype of the involved taxa (Supplementary Table S1).

Additionally the positioning of a clade comprised of Zymomonas mobilis (Kluyver and van Niel, 1936) including Zymomonas mobilis subsp. pomaceae (De Ley and Swings, 1976; Coton et al., 2006) caused conflict regarding Sphingomonadaceae in the GBDP tree (Figure 6). Zymomonas (Kluyver and van Niel, 1936) also formed a branch isolated from the remaining Sphingomonadaceae in the CCT (Supplementary File S2). While the resolution of the backbone within Sphingomonadales was partially low in the GBDP tree, an additional supermatrix analysis (Supplementary File S2) confirmed the placement of Zymomonas apart from the type species of the family, Sphingomonas. In contrast to most Sphingomonadaceae, Zymomonas was described as facultatively anaerobic (Supplementary File S2). While this deviation, much like the lower G+C content and genome size, is probably an autopomorphy of Zymomonas, there is no obvious phenotypic argument for placing the genus within Sphingomonadaceae as previously suggested (Kosako et al., 2000). Therefore we propose to transfer Zymomonas to a family of its own.

Sphingomonadales Genera

Within Sphingomonadales, Novosphingobium (Takeuchi et al., 2001) appeared as non-monophyletic in the GBDP tree (Figure 6) as well as in the CCT (Supplementary File S2) because N. tardaugens (Fujii et al., 2003) was placed in a relatively isolated position with respect to core Novosphingobium, more closely related to Altererythrobacter, Erythrobacter, and Porphyrobacter. When N. tardaugens was proposed the reduced taxon sampling in the presented 16S rRNA gene tree did not allow for an assessment of the monophyly of the genus. Given the overall genomic divergence of Erythrobacteraceae (Lee et al., 2005; Xu X.-W. et al., 2009), into which we propose to assign Novosphingobium as explained above, solving the non-monophyly of the major genera of the family by merging these genera does not seem taxonomically advisable. For this reason, we suggest a new genus to accommodate N. tardaugens, which is not contradicted by its phenotype (Supplementary Table S1). N. tardaugens also displayed a genome size more in accordance with the one found in other genera of Erythrobacteraceae than with the larger genomes of core Novosphingobium.

Erythrobacter (Shiba and Smidu, 1982; Subhash et al., 2013) was shown as non-monophyletic in the GBDP tree (Figure 6) because species such as E. gangjinensis (Lee et al., 2010), E. luteus (Lei et al., 2015), E. atlanticus (Zhuang et al., 2015), E. marinus (Jung et al., 2012b), E. seohaensis (Yoon et al., 2005b) and E. nanhaisediminis (Xu et al., 2010) were placed apart from the type species of Erythrobacter, E. longus (Shiba and Smidu, 1982) causing Altererythrobacter (Kwon et al., 2007; Xue et al., 2012, 2016) to appear intermixed with Erythrobacter. Moreover, E. longus was placed more closely to Porphyrobacter (Fuerst et al., 1993; Coil et al., 2015) than to the majority of the Erythrobacter species. Porphyrobacter appeared as polyphyletic in the GBDP tree (Figure 6) because P. mercurialis (Coil et al., 2015) was placed apart from the remaining Porphyrobacter species. Additionally, the single representative of Qipengyuania, Q. sediminis (Feng et al., 2015), was nested in the CCT within Altererythrobacter but with low support as in the original publication. Blastomonas marina (Meng et al., 2017) also was nested in the CCT within Altererythrobacter. These taxonomic problems were already observed by other authors as Erythrobacter, Altererythrobacter, and Porphyrobacter appeared intermixed in 16S rRNA gene phylogenies (Coil et al., 2015). Thus, we are well aware of the fact that after this modification Altererythrobacter, Erythrobacter and Porphyrobacter would still be intermixed (Figure 6) but given the low support in the 16S rRNA gene analyses (Supplementary File S2) the taxonomy of the family should be revisited once more type-strain genome sequences are available. For instance, the genome sequence of the type strain of the type species of Altererythrobacter, A. epoxidivorans (Kwon et al., 2007), was not yet available at the time of writing. We thus refrain from proposing taxonomic changes for these three genera.

Sphingosinicella (Maruyama et al., 2006; Geueke et al., 2007; Yasir et al., 2010) was shown as polyphyletic in the GBDP tree and in the CCT (Figure 6 and Supplementary File S2) because S. vermicomposti (Yasir et al., 2010) was placed as sister group of Sphingomonas indica (Niharika et al., 2012) supported by a high bootstrap value. When S. vermicomposti was proposed, 98% support was obtained for the monophyly of Sphingosinicella in a 16S rRNA gene analysis based on the neighbor-joining algorithm and a simplistic evolutionary model. We could not reproduce this finding here based on our ML and MP analyses with a much larger taxon sampling (Supplementary File S2), as even the unconstrained analyses placed S. vermicomposti in a position quite distinct from core Sphingosinicella. The proposal of Sphingomonas indica was accompanied by a 16S rRNA gene tree that did not resolved the monophyly of Sphingomonas (Yabuuchi et al., 1990, 1999, 2001; Takeuchi et al., 1993, 2001; Busse et al., 2003; Chen et al., 2012). Given the overall genomic divergence of Sphingomonadaceae, solving the non-monophyly of the major genera of the family by merging these genera does not seem taxonomically advisable. Our phylogenetic analyses thus suggested that S. vermicomposti and Sphingomonas indica are best assigned to an independent genus. The two species also display similar phenotypic characteristics (Supplementary Table S1) even though the gene-content analysis did not provide support (Supplementary File S2). The removal of Sphingomonas indica does not cause a need for the emendation of Sphingomonas (Supplementary Table S1).

Sphingopyxis (Takeuchi et al., 2001; Baik et al., 2013) appeared as paraphyletic in the GBDP tree and in and in the CCT (Figure 6 and Supplementary File S2) because Sphingopyxis baekryungensis (Yoon et al., 2005a) was placed not within core Sphingopyxis but in a clade together with Blastomonas (Sly and Cahill, 1997) and two species of Sphingorhabdus (Jogler et al., 2013; Yang et al., 2017) with high support. Since these two Sphingorhabdus species were placed apart from the type species of the genus in the 16S rRNA gene analyses and because of the unclear assignment of S. baekryungensis to either Blastomonas or Sphingorhabdus we propose to place S. baekryungensis into a genus of its own, which is not precluded by its phenotype (Supplementary Table S1). When S. baekryungensis was originally proposed (Yoon et al., 2005a) a phylogenetic analysis with low taxon sampling was presented which lacked any support for the monophyly of Sphingopyxis; other evidence for the monophyly of the genus was not detected either. Taxonomic consequences for Sphingorhabdus cannot currently be drawn because of the lack of a genome sequence for the type strain of its type species, S. planktonica, and the lack of resolution in even the constrained 16S rRNA gene analyses.

We are aware of the fact that after this modification Sphingomonas still remained non-monophyletic because genera such as Hephaestia (Felföldi et al., 2014), Rhizorhabdus (Francis et al., 2014), and Stakelama (Chen C. et al., 2010) were nested within its range (Figure 6). Given the overall genomic divergence of the group, solving the non-monophyly by including these genera in Sphingomonas does not seem taxonomically advisable. When Rhizorhabdus was proposed, a couple of Sphingomonas species already appeared more closely related to it than to the type species of Sphingomonas, S. paucimobilis (Yabuuchi et al., 1990), but they were not included in the new genus. The type species of Rhizorhabdus, R. argentea, formed in the CCT and UCT a well-supported clade together with R. dicambivorans (Yao et al., 2016), S. histidinilytica (Nigam et al., 2010), S. starnbergensis (Chen H. et al., 2013) and S. wittichii (Yabuuchi et al., 2001; Kim M.C. et al., 2016). It thus makes sense to include these Sphingomonas species in Rhizorhabdus, which is not precluded by their phenotype (Supplementary Table S1). Given the otherwise low support in the 16S rRNA gene analyses (Supplementary File S2) we cannot propose analogous taxonomic consequences for Hephaestia and Stakelama. The taxonomy of Sphingomonas should be revisited once more type-strain genome sequences are available. The removal of S. histidinilytica, S. starnbergensis and S. wittichii does not cause a need for the emendation of Sphingomonas (Supplementary Table S1).

Kordiimonadales and Caulobacterales Families

Only few discrepancies between taxonomic classification and phylogeny were observed in these relative small orders.

When Kordiimonas and with it the order Kordiimonadales was proposed (Kwon et al., 2005) no family to accommodate Kordiimonas was given. Kordiimonadaceae was later proposed (Xu et al., 2014) but has not been validly published yet. Kordiimonas formed a strongly supported clade together with Eilatimonas (Paramasivam et al., 2013) in the GBDP tree (Figure 8) and in the CCT (Supplementary File S2). When Eilatimonas was proposed, the genus already was shown in a 16S rRNA gene analysis as the sister group of Kordiimonas with strong support. Furthermore, Temperatibacter (Teramoto and Nishijima, 2014) of Temperatibacteraceae appeared as nested within Kordiimonadaceae in the CCT (Supplementary File S2), wherein the three genera formed a clade with strong support. The taxonomically most conservative solution is to include all these genera in an emended family Temperatibacteraceae, which is not precluded by their respective phenotypes (Supplementary Table S1).

Within Caulobacterales, Hyphomonadaceae (Lee et al., 2007d) appeared as paraphyletic in the GBDP tree because a clade comprising Robiginitomaculum to Maricaulis branched first, rendering core Hyphomonadaceae the sister group of Caulobacteraceae (Figure 8 and Supplementary File S2). However, support against the monophyly of Hyphomonadaceae was low. In the CCT an according clade was apparent that was composed of the genera Algimonas (Fukui et al., 2013), Fretibacter (Cho et al., 2013), Glycocaulis (Abraham et al., 2013; Lv et al., 2014), Hellea (Alain et al., 2008), Hyphobacterium (Sun et al., 2017), Litorimonas (Jung et al., 2011), Maricaulis (Abraham et al., 1999), Marinicauda (Zhang et al., 2013), Oceanicaulis (Strömpl et al., 2003), Robiginitomaculum (Lee et al., 2007d), and Woodsholea (Abraham et al., 2004), but this clade formed the sister group of core Hyphomonadaceae with moderate support. For this reason, we refrain from re-classifying these eleven genera into a separate family, which is not precluded by their phenotype (Supplementary Table S1).

Species and Subspecies

Values of dDDH similarity (Meier-Kolthoff et al., 2013a; Meier-Kolthoff and Göker, 2019) found to be higher or lower than expected given the current species and subspecies thresholds of 70% (Wayne et al., 1987) and 79%, respectively (Meier-Kolthoff et al., 2014b), as well as known and confirmed heterotypic synonyms, are shown in Table 1 for pairs of closely related strains. Multiple species and subspecies displayed a value above the 79% threshold, hence it is proposed that the according taxa be recognised as heterotypic synonyms at the subspecies level (see Table 1 for proposed synonyms). Conversely, some subspecies were shown to merit species status (Tindall, 2019b), specifically Acetobacter pasteurianus subsp. ascendens (De Ley and Frateur, 1974) and Acetobacter pasteurianus subsp. paradoxus (De Ley and Frateur, 1974). The name Acetobacter ascendens was already proposed (Kim et al., 2018) but the name does not appear to be validly published.

Brevirhabdus pacifica (Wu et al., 2015) was shown as paraphyletic in the phylogenomic analysis (Figure 7) because Xuhuaishuia manganoxidans (Wang L. et al., 2016) was nested within the two genome-sequenced type-strain deposits of the species. The dDDH analysis confirmed this result, as X. manganoxidans appeared as a later heterotypic synonym of B. pacifica. As the genus Brevirhabdus was not considered in the study where X. manganoxidans was proposed, the phylogenetic relationship of those taxa could not be elucidated back then.

Mameliella (Zheng et al., 2010; Chen Z. et al., 2015) appeared as paraphyletic in the GBDP and in the 16S trees (Figure 9 and Supplementary File S2) because Alkalimicrobium pacificum (Zhang D.-C. et al., 2015) and Ponticoccus lacteus (Yang Y. et al., 2015) were nested within Mameliella. When A. pacificum was proposed, it appeared as sister group of Mameliella alba (Zheng et al., 2010). When P. lacteus was proposed, the genus Mameliella was not considered in the study, and hence their phylogenetic relationship could not be elucidated. According to a recent study (Liu et al., 2018), A. pacificum and P. lacteus are later heterotypic synonyms of Mameliella alba, as is Mameliella atlantica (Xu et al., 2015). This was confirmed by the here calculated dDDH values (Table 1).

Some dDDH values between pairs of species were found to be higher than 70%, the currently accepted threshold to differentiate among species (Wayne et al., 1987) and lower than 79%, the threshold defined to differentiate among subspecies (Meier-Kolthoff et al., 2014b). Based on the dDDH values we here concluded that Rickettsia japonica (Uchida et al., 1992) is best assigned to a subspecies of Rickettsia conorii (Brumpt, 1932). However, the type strain of Rickettsia japonica is only deposited in a single culture collection, which prevents us from proposing an according new combination. In addition, we propose that Borrelia bavariensis (Margos et al., 2013), Gluconobacter nephelii (Kommanee et al., 2011), Methylobacterium chloromethanicum (McDonald et al., 2001), Methylorubrum extorquens (Urakami and Komagata, 1984; Green and Ardley, 2018), Pseudorhizobium pelagicum (Kimes et al., 2015), Rhizobium loessense (Wei et al., 2003), Rickettsia gravesii, Rickettsia heilongjiangensis, Rickettsia raoultii (Mediannikov et al., 2008) Rickettsia buchneri (Kurtti et al., 2015), and Ruegeria pelagia (Lee et al., 2007c) be classified as Borrelia garinii subsp. bavariensis, subsp. nov., Gluconobacter japonicus subsp. nephelii, subsp. nov., Methylobacterium dichloromethanicum subsp. chloromethanicum, subsp. nov., Methylobacterium dichloromethanicum subsp. extorquens, subsp. nov., Rhizobium marinum subsp. pelagicum, subsp. nov., Rhizobium mongolense subsp. loessense, subsp. nov., Rickettsia conorii subsp. gravesii, subsp. nov., Rickettsia conorii subsp. heilongjiangensis, subsp. nov., Rickettsia conorii subsp. raoultii, subsp. nov., Rickettsia tamurae subsp. buchneri, subsp. nov., and Tritonibacter mobilis subsp. pelagius, subsp. nov., respectively (Table 1).

Finally, in the present study, all of the pairs of strains considered to represent distinct deposits of the same type strain were found to have dDDH similarities of 99.0% or above with the exception of Celeribacter indicus (Lai et al., 2014) strain MCCC 1A01112 and DSM 27257 (87.8%), Celeribacter marinus (Baek et al., 2014) strain IMCC12053 and DSM 100036 (94.8%), Gluconacetobacter diazotrophicus (Yamada et al., 1997) strain PAl 5 and DSM 5601 (92.1%), Thalassobius mediterraneus (Arahal et al., 2005) strain CECT 5383 and DSM 16398 (98.3%), Thalassospira xiamenensis (Liu et al., 2007) strain M-5 and DSM 17429 (94.7%), results which may account for the separation of each pair of these strains.

Phylogenetic Conservation of Genomic and Phenotypic Markers of Interest

Table 2 shows the p-values obtained by the tip-permutation test and the retention values of selected phenotypic and genomic features. All investigated characters showed a significant phylogenetic conservation (α = 0.001) but the fit of each character to the tree varied considerably, as indicated by the retention index. A relatively high correspondence between G+C content and phylogeny and genome size and phylogeny was observed. Genome size varied between 0.85 Mbp in Neorickettsia sennetsu and 9.79 Mbp in Bradyrhizobium arachidis while G+C content varied between 26.86% in Brachyspira alvinipulli and 73.38% in Rubritepida flocculans. Morphology showed a significant but relatively low conservation but motility by flagella appeared certainly more conserved than cell length and particularly cell width.

Presence or absence of flagella also showed a better fit to the tree than the relationship to oxygen and the investigated chemotaxonomic features. Among these, presence or absence of carotenoids showed the highest conservation, followed by average number of isoprene residues in major ubiquinones and, finally, presence or absence of bacteriochlorophyll α. Particularly the chemotaxonomic features showed a low coverage in the taxon descriptions from the literature. The screening of the literature indicated that this may at least partially be due to the reluctance of researchers to report negative results. It is reasonable to assume that a perceived absence of bacteriochlorophyll α, carotenoids, or sphingolipids is not necessarily reported as such but simply omitted. In the case of sphingolipids the lack of reports for them outside Sphingomonadales even prevented the calculation of the metrics presented in Table 2. The less cautious coding that treated all missing values as indicating absence (Supplementary Table S1) yielded a high retention index for sphingolipids (0.739) whereas the retention indices of the other binary characters dropped compared to the values depicted in Table 2.

The gene-content analysis, specific aspects of which have already been discussed above, was largely in agreement with the GBDP tree even though certain parts of the trees were, predictably, in conflict (Supplementary File S2). For instance, using a branch-support threshold of 95% to indicate strong support or conflict, among the branches strongly supported by the GBDP analysis 40.6% were also strongly supported by the gene-content analysis, 6.7% were strongly contradicted and 52.7% of the cases the gene-content analysis remained neutral (Supplementary Table S1). Compared to the average branch support of the GBDP analysis of 92.1%, the gene-content analysis yielded 68.0% average support by itself and on average 45.2% support for the branches in the GBDP tree (Supplementary File S2).

Discussion

Causes of Conflict Between Phylogenomic Analyses and Taxonomic Classification

Most of the Alphaproteobacteria taxa already appeared monophyletic in this study eliminating need for a taxonomic revision. This partially reflects the work previously done by other authors where they proposed taxonomic changes within Alphaproteobacteria based on at least multi-gene if not phylogenomic datasets as, e.g., in the case of revisions of genera of Rhodobacteraceae (Breider et al., 2014; Wirth and Whitman, 2018). Much of the remaining discrepancies can be traced back to the way the 16S rRNA gene was interpreted in the taxonomic literature, which is discussed here while the interpretation of phenotypic features is treated in the next section.

In the case of some Alphaproteobacteria taxa such as Mycoplana (Urakami et al., 1990) and Agrobacterium meteori (Rüger and Höfle, 1992) the discrepancies between their classification and the modern methods was caused by the fact that they were described prior to the availability of 16S rRNA gene sequencing. In these low number of cases, the taxonomic conclusions that could be drawn with the 16S rRNA gene and those indicated by phylogenomic trees were essentially the same.

While in few cases, such as the original descriptions of Angulomicrobium and Meganema, branch support was not even calculated, most of the taxonomic discrepancies observed within the Alphaproteobacteria appeared to be caused by low resolution of the 16S rRNA genes used to propose the respective taxa. This held for genera such as Alkalispirochaeta, Allorhizobium, Altererythrobacter, Celeribacter, Citreicella, Erythrobacter, Lutimaribacter, Mesorhizobium, Neorhizobium, Novosphingobium, Ochrobactrum, Porphyrobacter, Rhizobium, Roseomonas, Sphingomonas, Sphingopyxis, Thalassobius, and Tropicimonas, as well as the families Hyphomicrobiaceae, Phyllobacteriaceae, Rhodobiaceae, Rhodospirillaceae, and Sphingomonadaceae. Appropriately calculating branch support is a necessary but not a sufficient prerequisite for safely generating monophyletic taxa. When drawing taxonomic conclusions from some phylogenetic tree, taxa must also be chosen so as to correspond to highly supported clades (Vences et al., 2013). Such clades are not always present in 16S rRNA genes although all species must be assigned to a genus in the Linnaean system.

The second most important cause of non-monophyly in the class Alphaproteobacteria detected in the current study was incomplete taxon sampling. Actually non-monophyletic taxa may easily appear monophyletic when species or strains of relevance, in particular type strains of type species of genera or even type genera of families or orders, are omitted from phylogenetic analysis. Incomplete taxon sampling affected the taxonomic assignment to families in the case of Acuticoccus, Aquamicrobium, Breoghania, Cucumibacter, Devosia, Magnetovibrio, Maritalea, Marivibrio, Pelagibacterium, Reyranella, Tagaea, Terasakiella, Thalassospira, and Varunaivibrio. Insufficient sampling also affected the assignment of species to genera at least in the case of Novosphingobium tardaugens and Sphingopyxis baekryungensis.

We found no evidence for a real conflict between the 16S rRNA gene and entire genomes in this study. In previous analyses (Hahnke et al., 2016; Nouioui et al., 2018) such conflicts already appeared to be rare compared to the overall number of taxa investigated and compared to the more common causes of taxonomic conflicts, which were the same as the ones listed above. Analyzing comprehensive sets of 16S rRNA gene sequences appears to be necessary unless a genome sequence is available for all type strains, particularly because taxonomic problems caused by insufficient taxon sampling seem to be quite common. The use of a backbone constraint can integrate information from analyses of more genes from few organisms into comprehensible sampled single-gene data (Liu X.-Z. et al., 2015; Hahnke et al., 2016; Nouioui et al., 2018). For example, the CCT was necessary to safely place type species that still lack a genome sequence and to detect instances where taxonomic conclusions would have been premature.

In previous studies (Hahnke et al., 2016; Nouioui et al., 2018) the taxonomic conclusions drawn from the GBDP tree were confirmed by corresponding supermatrix analyses in all of the cases investigated. In the present study it was also assessed whether conflict was evident between trees inferred from distinct data sets and, if so, to conduct analyses with alternative methods. We could not detect such conflict in Alphaproteobacteria, however, and accordingly restricted supermatrix analyses to cases in which the GBDP tree was insufficiently resolved for clarifying specific taxonomic questions. In these situations, conflict between the GBDP tree and the supermatrix analyses was not detected either. Our approach thus appeared to be as robust for Alphaproteobacteria as in previous studies on other phyla (Hahnke et al., 2016; Nouioui et al., 2018). Whole-genome methods, such as GBDP, yield truly genome-based phylogenies instead of approaches that use only a limited number of genes, which also rely on assumptions about the relative suitability of the selected genes compared to other genes (Lienau and DeSalle, 2009; Klenk and Göker, 2010). Distance methods for tree reconstruction still represent the most promising approach for accurately building phylogenies with a huge number of tips (Desper and Gascuel, 2004, 2006; Lefort et al., 2015). In an approach using GBDP in conjunction with FastME the time-consuming step is the calculation of the intergenomic distances, which can be done incrementally since the pairwise distances can be calculated independently of each other.

While the use of genome-scale data often yields more strongly resolved trees, it may also increase incongruities between distinct analyses (Jeffroy et al., 2006; Klenk and Göker, 2010). In this context overestimating phylogenetic confidence from genome-scale data must be avoided (Taylor and Piel, 2004). A reduction of the supposed incongruities between phylogenies and thereby more realistic support values for phylogenomic analyses can be obtained by bootstrapping entire genes instead of single alignment positions (Falush et al., 2006; Siddall, 2010; Hahnke et al., 2016; Simon et al., 2017). GBDP pseudo-bootstrapping in conjunction with the greedy-with-trimming algorithm (Meier-Kolthoff et al., 2014a; Meier-Kolthoff and Göker, 2019) is akin to such a “partition bootstrap.” The concept of hierarchical classification itself has been called into question based on topological incongruities between analyses of single genes that were attributed to horizontal gene transfer (Bapteste and Boucher, 2009; Klenk and Göker, 2010). However, the addition of more genes (Breider et al., 2014), up to virtually all available ones, as in the present study, yields strong support even when based on a careful approach to statistical resampling. This observation indicates a strong hierarchical signal and thus no reason to abandon the hierarchical system of taxonomic classification.

Agreement and Conflict of Taxonomic Markers With Genome-Scale Phylogeny

Phylogenetic conservation of investigated phenotypic markers was detectable but varied and was on average not particularly pronounced (Table 2). This outcome is in overall agreement with an earlier study (Barberán et al., 2017) which compared a 16S rRNA gene tree to selected phenotypic features collected from recent taxonomic descriptions published in the International Journal of Systematic and Evolutionary Microbiology across phyla. While it comes as no surprise that cell shape is less conserved than presence or absence of flagella (Barberán et al., 2017), the relative performance of certain character deserves further discussion. Two distinct kinds of causes for the discrepancies between the phylogenomic trees and the traditional classification, which was at least partially based on phenotypic characters, are possible. The taxonomic interpretation of these characters could be insufficient (Montero-Calasanz et al., 2017; Nouioui et al., 2018), and the characters themselves could conflict with the phylogenomic trees. For distinguishing between these two options, the characters need to be examined as they were used in the taxonomic literature. For a historical assessment such as ours it makes no sense to analyze new, modified characters derived from the earlier ones. While it may be of interest in other respects to take the genetic background into account to form new characters, these new characters were not used for establishing the traditional classification in the first place.

Characters such as flagellum production or capability for photosynthesis are based on multiple genes usually arranged in gene clusters (Frank et al., 2015; Brinkmann et al., 2018) and thus can be assumed to be rather complex characters. According to Dollo’s law, complex features arise only once but can be lost several times in evolution (Le Quesne, 1974; Farris, 1977). Accordingly, a group of organisms which display a complex feature are expected to be monophyletic or paraphyletic in a tree, but not polyphyletic (Nouioui et al., 2018). The failure to properly distinguish between plesiomorphic (ancestral) character states, which indicate paraphyletic groups, and apomorphic (derived) character states, which indicate monophyletic groups (Hennig, 1965; Wiley and Lieberman, 2011), may account for some of the discrepancies between the current taxonomic classification and genome-scale trees (Montero-Calasanz et al., 2017; Nouioui et al., 2018). Homoplasy in the investigated characters most likely plays a role in other cases, and evolution according to Dollo’s law alone could cause homoplasy in a complex character. In the case of the photosynthesis in Alphaproteobacteria, however, it was shown that horizontal transfer of single genes as well as of entire photosynthesis operons occurred in addition to frequent losses of the capability for photosynthesis (Brinkmann et al., 2018). Such vertical transfer of complex characters may not only be facilitated by the arrangement of the according genes in gene clusters but also by the location of these gene clusters on extrachromosomal elements. This holds not only for the photosynthesis operon (Brinkmann et al., 2018) but also for the flagellum gene cluster (Frank et al., 2015). Losses and horizontal transfer can both contribute to the low phylogenetic conservation of phenotypic features, which in turn can contribute to discrepancies between the taxonomic classification and the phylogeny. It should not be overlooked, however, that certain phenotypic features can have an excellent fit to the phylogeny, as most likely in the case of the production of sphingolipids in Alphaproteobacteria (Supplementary File S2).

As for genomic features, bacterial G+C content and bacterial genome size appeared to be strongly phylogenetically conserved in Alphaproteobacteria (Supplementary File S2 and Table 2). For this reason, genome sizes have been added to the description of the reclassified and emended species as shown below. A significant correlation between genome size and G+C content was found in previous studies (Almpanis et al., 2018; Nouioui et al., 2018), which is not unexpected because symbiotic bacteria tend to have smaller genomes and to be richer in A+T content (Rocha and Danchin, 2002; Mann and Chen, 2010), an effect that may even impact the amino-acid content (Cole et al., 1998). Conversely, positive selection (Hildebrand et al., 2010) and G+C-biased gene conversion (Lassalle et al., 2015) can increase the G+C content. Although exceptions from the rule that reduced genomes have a low G+C content are known (McCutcheon et al., 2009), the rule was confirmed in the present study, particularly regarding genome size and G+C content reduction in pathogens such as Bartonella, Borrelia and Rickettsiales (Supplementary File S2). Because of this correlation caused by adaptive processes, genome size could thus be regarded as non-independent of G+C content, which would cast some doubt on its use as a taxonomic marker. Yet as shown previously (Nouioui et al., 2018) the overall correlation between G+C content and genome size is considerably reduced in strength after accounting for the impact of the phylogeny. This correlation may even only be due to G+C reduction effects in symbiotic bacteria with dramatically reduced genome sizes. The apparently non-random distribution with respect to the phylogeny of such reductions in Alphaproteobacteria (Supplementary File S2) also underlines the value of both G+C content and genome size as taxonomic markers in the class.

Now that it has been shown that within-species deviation in G+C content is at most 1% (Meier-Kolthoff et al., 2014c), many Alphaproteobacteria species descriptions were found to be inaccurate or too imprecise. The same observation was made in our earlier studies (Hahnke et al., 2016; Nouioui et al., 2018) on other groups of bacteria. As such values not only assist in detecting strains that do not belong to the same species but also show significant correlation to phylogenetic trees it is good practise to strengthen species descriptions in this way. In contrast, in most cases it is premature to redefine genera and higher taxa of Alphaproteobacteria in this way since additional type-strain genome sequences would be needed before this issue could be addressed. Only few genera were comprehensively sampled genomically in this study, which allows for providing respective emendations below.

Single characters may or may not be optimal as taxonomic markers, as they are often chosen for historical reasons. The application of gene-content phylogenies is more attractive for taxonomic purposes as they can be based on a huge number of characters now that genome sequences are available for many species. Using distance methods for inferring gene-content phylogenies is not a new idea; in particular, distance formulas were favoured which disregard double absence of genes to account for independent genome reduction (Wolf et al., 2002; Huson and Steel, 2004; Yang et al., 2005). GBDP formula d3 follows the same approach (Henz et al., 2005). While more sophisticated approaches have been proposed (Gu and Zhang, 2004; Huson and Steel, 2004), we here used GBDP to infer gene-content phylogenies because the method is established. GBDP also allows for a direct comparison with methods based on sequence identity as only a single factor, the distance formula (d₃ vs. d₅), needs to be considered.

It must be taken into account that gene-content phylogenies may fail to recover the “true” tree for a variety of reasons, including horizontal gene transfer and gene loss; depending on the perspective, this may lead to big-genome attraction or small-genome attraction (Wolf et al., 2001; Lake and Rivera, 2004). Last but not least, many published genome sequences are not closed, which does not necessarily add a bias but most likely some noise to gene-content data. For these reasons we do not recommend gene-content approaches for inferring phylogenies although for some data sets they can certainly yield the same topology as approaches based on sequence alignment (Breider et al., 2014). Rather, we regard gene content as of interest for selecting branches from a phylogenetic tree, which was inferred by using a standard genome-scale approach, to obtain a taxonomic classification. Using the gene content directly for the purpose of classification irrespective of whether or not monophyletic taxa are generated (Zhu et al., 2015) is not in agreement with the principles of phylogenetic classification (Wiley and Lieberman, 2011). However, taking gene-content data into account is of interest because gene content conveys metabolic and other phenotypic capabilities (Zhu et al., 2015).

Our results indicate that gene content can be used in this manner to improve the taxonomic classification. In this respect is may even be advantageous if certain parts of the gene-content tree conflicted with the standard genome-scale analysis because this would narrow down the number of branches of the tree to be selected to form taxa. The results of the present study also indicated that gene-content analysis can provide useful insights into the evolution of Alphaproteobacteria and most likely also other groups of bacteria. For instance, the basal branches within Rhodobacterales are well supported in the gene-content analysis, reaching 100% for Rhodobacteraceae. This may be indicative of specific changes in gene content that occurred early on within the radiation of the order. As a variety of representatives of the family are dominant in marine habitats (Brinkhoff et al., 2008), these changes may be of considerably ecological interest. The genomic basis of the switches of Rhodobacteraceae between marine and non-marine habitats has recently been elucidated (Simon et al., 2017). A logical next step in future studies is to reconstruct the changes in gene content which took place at the basis of the Rhodobacterales part of the tree and led to the organisms that later on split into the numerous species of Rhodobacteraceae. For reasons of taxonomic conservatism, we here refrained from reclassifying monophyletic taxa solely based on their lack of support by the gene-content analysis. However, in this study the gene content appeared valuable for delineating new taxa in a variety of cases in which the existing taxa needed to be revised because they were evidently non-monophyletic. Considering gene content can apparently reduce arbitrariness in taxonomic decisions and increase the information content of the taxonomic system. It thus deserves attention in future studies on genome-based taxonomic classification of Alphaproteobacteria and other groups of organisms.

Conclusion and Outlook

The results of this study provide a further improved framework for the classification of the class Alphaproteobacteria. The newly proposed taxonomic classification provides a sound basis for future studies on these bacteria, not least on those of ecological interest such as Rhodobacteraceae, which are major players in many marine habitats. Discrepancies of the results from the comparative phylogenomic approach with aspects of the previous taxonomic classification based on the 16S rRNA gene were mainly caused by insufficient taxon sampling and disregarded or overestimated branch support. Exceptions in which the 16S rRNA gene is in real conflict with genome-scale phylogenies were not observed. While they are, expectedly, not in full agreement, the relatively high correspondence between gene-content phylogeny and standard genome-scale analyses yields “grist to the taxonomic mill.” Also encouraging is the strong agreement between G+C content, genome size and selected phenotypic features on the one hand and truly genome-scale phylogenies of Alphaproteobacteria on the other hand. Future phylogenomic studies should try to link such key features, their genomic basis and their evolutionary relationships, and make an attempt to clarify evolutionary relationships that could as yet not be resolved. This affects particularly those groups that were thoroughly sampled but are not yet well covered by genome sequencing. Success in Alphaproteobacteria may help to revitalize prokaryotic systematics as a fundamental scientific discipline in other parts of the bacterial tree of life, particularly if it manages to integrate the distinct types of rich data that are nowadays available.

Taxonomic Consequences: New Orders

Description of Mariprofundales ord. nov.

Ma.ri.pro.fun.da’les (N.L. masc. n. Mariprofundus, the type genus of the order; -ales, ending to denote an order; N.L. fem. pl. n. Mariprofundales, the order of Mariprofundus).

The description is the same as for the family Mariprofundaceae fam. nov. (Emerson et al., 2007), the sole family in the order, which was proposed earlier on but has not been validly published yet. The type genus of the order is Mariprofundus. Phylogenetic analyses of genome and 16S rRNA gene sequences indicate that the genus is best placed into an order of its own.

Taxonomic Consequences: New Families

Description of Afifellaceae fam. nov.

A.fi.fel.la’ce.ae (N.L. fem. dim. n. Afifella, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Afifellaceae, the Afifella family).

The description is as given for Afifella (Urdiain et al., 2008), which is the type and currently the sole genus of the family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Ahrensiaceae fam. nov.

Ah.rens.i.a’ce.ae (N.L. fem. n. Ahrensia, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Ahrensiaceae, the Ahrensia family).

Cells are Gram-negative, ovoid to rod-shaped, non-motile or motile by means of polar or peritrichous flagella. Aerobic, oxidase and catalase positive. The major ubiquinone is Q-10. The major named polar lipids are phosphatidylcholin, phosphatidylethanolamine, phosphatidylglycerol and diphosphatidylglycerol. The major fatty acid is C_18:1 ω7c. The G+C content as calculated from genome sequences is around 48.1–50.1% while the range provided in the literature is 48.1–60.1 mol%. The family currently comprises the genera Ahrensia (the type genus) and Pseudahrensia.

Description of Amorphaceae fam. nov.

A.mor.pha’ce.ae (N.L. masc. n. Amorphus, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Amorphaceae, the Amorphus family).

Cells are Gram-negative, aerobic, with variable morphology and non-flagellated. The major ubiquinone is Q-10. The major fatty acids are C_18:1 ω7c and C_19:0 cyclo ω8c. The family currently comprises the genera Acuticoccus and Amorphus (the type genus). This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Aurantimonadaceae fam. nov.

Au.ran.ti.mo.na.da’ce.ae (N.L. fem. n. Aurantimonas, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Aurantimonadaceae, the Aurantimonas family).

Cells are Gram-negative, rod-shaped, aerobic to facultatively aerobic and either non-motile or motile by means of flagella. The major ubiquinone is Q-10. Usually catalase- and oxidase positive. Growth occurs under mesophilic conditions. Carotenoids can be present. NaCl requirement is variable. The major named polar lipids are diphosphatidylglycerol, phosphatidylglycerol, phosphatidylcholin, phosphatidylethanolamine, phosphati- dylmonomethylethanolamine and aminolipids. The major named fatty acids are usually C_18:1 ω7c and C_19:0 cyclo ω8c. The family currently comprises the genera Mangrovicella, Jiella, Aurantimonas (the type genus), Aureimonas, Fulvimarina and Consotaella.

Description of Azospirillaceae fam. nov.

A.zo.spi.ril.la’ce.ae (N.L. neut. dim. n. Azospirillum, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Azospirillaceae, the Azospirillum family).

Cells are Gram-negative, mostly rod-shaped, in general aerobic, usually motile by flagella, usually heterotrophic. The major ubiquinone is predominantly Q-10. The major named polar lipids are phosphatidylcholin, phosphatidylethanolamine, phosphatidylglycerol and diphosphatidylglycerol. The major fatty acids are usually C_18:1 ω7c, C_18:1 ω6c, C_16:0 and C_16:1 ω6c. The family currently comprises the genera Niveispirillum, Azospirillum (including Conglomeromonas), Skermanella, Nitrospirillum, Rhodocista, and Desertibacter. The genera Elstera, Inquilinus and Lacibacterium are tentatively assigned to this family. The type genus is Azospirillum.

Description of Blastochloridaceae fam. nov.

Blas.to.chlo.ri.da’ce.ae (N.L. fem. n. Blastochloris, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Blastochloridaceae, the Blastochloris family).

The description is as given for Blastochloris (Hiraishi, 1997), which is the type and currently the sole genus of the family. This family has been separated from Hyphomicrobiaceae based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Boseaceae fam. nov.

Bo.se.a’ce.ae (N.L. fem. n. Bosea, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Boseaceae, the Bosea family).

The description is as given for Bosea (Das et al., 1996; La Scola et al., 2003), which is the type genus and currently the sole genus of the family. This family has been separated from Nitrobacteraceae (whose illegitimate synonym is Bradyrhizobiaceae) based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Breoghaniaceae fam. nov.

Bre.o.gha.ni.a’ce.ae (N.L. fem. n. Breoghania, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Breoghaniaceae, the Breoghania family).

The description is as given for Breoghania (Gallego et al., 2010),which is the type and currently the sole genus of the family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Devosiaceae fam. nov.

De.vo.si.a’ce.ae (N.L. fem. n. Devosia, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Devosiaceae, the Devosia family).

Cells are Gram-negative, predominantly rod-shaped, aerobic, usually motile by flagella, and heterotrophic. The major ubiquinone is predominantly Q-10 whereas Q-9 and Q-11 present as minor ubiquinone in some species. The major polar lipids are phosphatidylglycerol, diphosphatidylglycerol and glycolipid. The major fatty acids are usually C_18:1 ω7c, 11-methyl C_18:1 ω7c, C_18:1 ω6c, C_18:0, C_16:0 ω6c and C_19:0 cyclo ω8c. The family currently comprises the genera Arsenicitalea, Cucumibacter, Devosia (the type genus), Maritalea (including Zhangella), Methyloterrigena, Paradevosia, Pelagibacterium, and Youhaiella. This family has been separated from Hyphomicrobiaceae based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Kaistiaceae fam. nov.

Kais.ti.a’ce.ae (N.L. fem. n. Kaistia, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Kaistiaceae, the Kaistia family).

Cells are Gram-negative, cocci or rod-shaped, non-motile, aerobic and chemoorganotrophic. The G+C content as calculated from genome sequences is around 64.5–67.1% while the range provided in the literature is 61.6–69.0 mol%. The family currently comprises the genera Kaistia (the type genus) and Bauldia.

Description of Mariprofundaceae fam. nov.

Ma.ri.pro.fun.da’ce.ae (N.L. masc. n. Mariprofundus, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Mariprofundaceae, the Mariprofundus family).

The description is that for Mariprofundus (Emerson et al., 2007), which is the type and currently sole genus of the family. The family was proposed earlier on but has not been validly published yet. It has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Neomegalonemataceae fam. nov.

Ne.o.me.ga.lo.ne.ma.ta’ce.ae (N.L. neut. n. Neomegalonema, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Neomegalonemataceae, the Neomegalonema family).

The description is as given for Neomegalonema (Oren, 2017b), which is the type and currently the sole genus of the family. This family has been separated from Hyphomicrobiales (known under the illegitimate synonym Rhizobiales) and from Rhodobacteraceae based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Parvibaculaceae fam. nov.

Par.vi.ba.cu.la’ce.ae (N.L. neut. n. Parvibaculum, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Parvibaculaceae, the Parvibaculum family).

Cells are Gram-negative, usually rod-shaped, generally aerobic, motile by flagella or non-motile and heterotrophic. The major ubiquinone is Q-10 or Q-11. The major polar lipids are diphosphatidylglycerol, phosphatidylethanolamine and phosphatidylglycerol. The major fatty acid is usually C_18:1 ω7c and in some cases C_16:0 and C_19:0 cyclo. The family currently comprises the genera Parvibaculum (the type genus) and Tepidicaulis. Anderseniella, Rhodoligotrophos and Pyruvatibacter are tentatively assigned to this family as well.

Description of Phreatobacteraceae fam. nov.

Phre.a.to.bac.te.ra’ce.ae (N.L. masc. n. Phreatobacter, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Phreatobacteraceae, the Phreatobacter family).

The description is as given for Phreatobacter (Tóth et al., 2014; Lee S.D. et al., 2017), which is the type and currently the sole genus of the family. This family is proposed based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Pleomorphomonadaceae fam. nov.

Ple.o.mor.pho.mo.na.da’ce.ae (N.L. fem. n. Pleomorphomonas, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Pleomorphomonadaceae, the Pleomorphomonas family).

Cells are Gram-negative, usually rod-shaped or spherical, generally aerobic and mostly diazotrophic. The major ubiquinone is predominantly Q-10. The major polar lipids are phosphatidylglycerol, phosphatidylcholin, phosphatidylethanolamine and sometimes phospholipid. The major fatty are usually C_18:1 ω7c and C_18:0. The family currently comprises the genera Chthonobacter, Hartmannibacter, Methylobrevis, Mongoliimonas, Oharaeibacter, and Pleomorphomonas (the type genus). This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Pseudoxanthobacteraceae fam. nov.

Pseu.do.xan.tho.bac.te.ra’ce.ae (N.L. masc. n. Pseudoxanthobacter, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Pseudoxanthobacteraceae, the Pseudoxanthobacter family).

The description is as given for Pseudoxanthobacter (Arun et al., 2008), which is the type and currently sole genus of the family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Reyranellaceae fam. nov.

Rey.ra.nel.la’ce.ae (N.L. fem. dim. n. Reyranella, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Reyranellaceae, the Reyranella family).

The description is as given for Reyranella (Pagnier et al., 2011) which is the type and currently sole genus of the family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Rhodovibrionaceae fam. nov.

Rho.do.vi.bri.o.na’ce.ae (N.L. masc. n. Rhodovibrio, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Rhodovibrionaceae, the Rhodovibrio family).

Gram-negative, vibrioid, spiral or rod-shaped, non-motile or motile by means of polar flagella. Aerobic, facultatively anaerobic or anaerobic, with chemoorganotrophic or photoorganoheterotrophic metabolism. The major ubiquinone is Q-10; MK-10 was reported for Rhodovibrio. The major fatty acids are C_19:0 cyclo ω8c, C_18:1 ω7c and C_18:0. The G+C content provided in the literature is 61.5–69.0 mol%. The family currently comprises the genera Fodinicurvata, Limimonas, Rhodovibrio (the type genus) and Tistlia. Limibacillus and Pelagibius are tentatively assigned to this family.

Description of Sphaerochaetaceae fam. nov.

Sphae.ro.chae.ta’ce.ae (N.L. fem. n. Sphaerochaeta, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Sphaerochaetaceae, the Sphaerochaeta family).

Cells are Gram-negative, mostly non-motile, pleomorphic, anaerobic and heterotrophic. The family currently comprises the genera Pleomorphochaeta and Sphaerochaeta (the type genus). This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Sphingosinicellaceae fam. nov.

Sphin.go.si.ni.cel.la’ce.ae (N.L. fem. n. Sphingosinicella, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Sphingosinicellaceae, the Sphingosinicella family).

Cells are Gram-negative, aerobic, rod-shaped, mostly motile heterotrophs. The major ubiquinone is Q-10. The major named polar lipids are phosphatidylethanolamine, phosphatidylglycerol and in some species diphosphatidylglycerol. The major fatty acids are usually C_18:1 ω7c, C_16:0 and C_16:1 ω7c. The family currently comprises the genera Pacificimonas, Sphingosinicella (the type genus), Sandaracinobacter and Sandarakinorhabdus. Polymorphobacter and Sphingoaurantiacus are tentatively assigned to this family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Stappiaceae fam. nov.

Stap.pi.a’ce.ae (N.L. fem. n. Stappia, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Stappiaceae, the Stappia family).

Cells are Gram-negative, usually rod-shaped, motile, mostly aerobic or facultatively anaerobic and heterotroph. The major ubiquinone is usually Q-10. The major fatty acids are predominantly C_18:1 ω7c and C_16:1 ω7c. The family currently comprises the genera Pannonibacter, Pseudovibrio, Roseibium (including Labrenzia) and Stappia (the type genus). This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Stellaceae fam. nov.

Stel.la’ce.ae (L. fem. n. Stella, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Stellaceae, the Stella family).

The description is as given for Stella (Vasilyeva, 1985), which is the type genus of the family, with the following modification. Cells are six-pronged stars, rods or ovoids. The family houses Stella and tentatively also Constrictibacter. This family has been separated from Acetobacteraceae based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Tepidamorphaceae fam. nov.

Te.pid.a.mor.pha’ce.ae (N.L. masc. n. Tepidamorphus, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Tepidamorphaceae, the Tepidamorphus family).

Cells are Gram-negative, either ovoid or rod-shaped, predominantly aerobic, mostly motile via flagella and generally chemoorganothrophs. The major ubiquinone is Q-10. The major named polar lipids are diphosphatidylglycerol, phosphatidylethanolamine, and phosphatidylglycerol. The major fatty acids are usually C_19:0 cyclo ω8c, C_18:1 ω7c, C_18:0 and C_16:0. The family currently comprises the genera Butyratibacter, Lutibaculum, Microbaculum, and Tepidamorphus (the type genus).

Description of Terasakiellaceae fam. nov.

Te.ra.sa.ki.el.la’ce.ae (N.L. fem. dim. n. Terasakiella, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Terasakiellaceae, the Terasakiella family).

The description is as given for Terasakiella (Satomi et al., 2002; Han et al., 2016), which is the type and currently sole genus of the family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Thalassobaculaceae fam. nov.

Tha.las.so.ba.cu.la’ce.ae (N.L. neut. n. Thalassobaculum, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Thalassobaculaceae, the Thalassobaculum family).

Gram-negative, rod-shaped, motile by means of one polar flagellum. Aerobic or facultatively anaerobic respiration and chemoorganotrophic metabolism. The major ubiquinone is Q-10. The major fatty acids are C_18:1 ω7c, C_16:1 ω7c and C_16:0. The G+C content as calculated from genome sequences is around 60.5–67.4% while the range provided in the literature is 60–68 mol%. The family currently comprises the genera Nisaea, Oceanibaculum and Thalassobaculum (the type genus).

Description of Thalassospiraceae fam. nov.

Tha.las.so.spi.ra’ce.ae (N.L. fem. n. Thalassospira, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Thalassospiraceae, the Thalassospira family).

Cells are Gram-negative, usually motile by flagella, mostly aerobic or microaerophilic with a heterotrophic or autolithothrophic metabolism. The major ubiquinone is predominantly Q-10. The major polar lipids are phosphatidylethanolamine and phosphatidylglycerol. The major fatty acids are usually C_18:1 ω7c, C_16:1 ω7c and C_16:0. The family currently comprises the genera Magnetovibrio, Terasakiella, Thalassospira (the type genus), Varunaivibrio. Magnetospira is tentatively assigned to this family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Zavarziniaceae fam. nov.

Za.var.zi.ni.a’ce.ae (N.L. fem. n. Zavarzinia, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Zavarziniaceae, the Zavarzinia family).

The description is as given for Zavarzinia (Meyer et al., 1993), which is the type and currently sole genus of the family. This family has been separated from Acetobacteraceae based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Zymomonadaceae fam. nov.

Zy.mo.mo.na.da’ce.ae (N.L. fem. n. Zymomonas, type genus of the family; -aceae, ending to denote a family; N.L. fem. pl. n. Zymomonadaceae, the Zymomonas family).

The description is as given for Zymomonas (Kluyver and van Niel, 1936), which is the type and currently sole genus of the family. This family has been separated from other families based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Taxonomic Consequences: New Genera

Description of Albibacillus gen. nov.

Al.bi.ba.cil’lus (L. masc. adj. albus, white; L. masc. n. bacillus, rod; N.L. masc. n. Albibacillus, white rod).

Gram-negative, rod-shaped, aerobic, oxidase and catalase positive. The major ubiquinone is Q-10. The major polar lipids are phosphatidylglycerol, phosphatidylethanolamine, phosphatidylmonomethylethanolamine and aminolipids. The major fatty acids are C_18:1 ω7c and C_19:0 cyclo ω8c. The G+C content is around 64-66%. The type species is Albibacillus kandeliae, comb. nov.

Description of Allgaiera gen. nov.

All.gai’er.a (N.L. fem. n. Allgaiera, named after the German microbiologist Martin Allgaier for his work on marine Rhodobacteraceae).

The description is as given for Allgaiera indica, comb. nov., which is the type species. The genus has been separated from Defluviimonas based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Allosediminivita gen. nov.

Al.lo.se.di.mi.ni.vi’ta (Gr. masc. adj. allos, another, other, different; N.L. fem. n. Sediminivita, a bacterial genus; N.L. fem. n. Allosediminivita, a genus different from Sediminivita).

The description is as given for Allosediminivita pacifica, comb. nov., which is the type species. The genus has been separated from Roseivivax based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Allosphingosinicella gen. nov.

Al.lo.sphin.go.si.ni.cel’la (Gr. masc. adj. allos, another, other, different; N.L. fem. n. Sphingosinicella, a bacterial genus; N.L. fem. n. Allosphingosinicella, a genus different from Sphingosinicella).

Cells are Gram-negative, non-motile, non-spore-forming and rod-shaped. Strictly aerobic. The major ubiquinone is Q-10. The major polar lipids are diphosphatidylglycerol, phosphatidylglycerol and phosphatidylethanolamine. The major fatty acids are C_18:1 ω7c, C_16:1 ω7c, C_14:0 2-OH and C_16:0. The G+C content as calculated from genome sequences is around 62.4–67.0% while the range provided in the literature is 59.4–65.8 mol%. The type species is Allosphingosinicella vermicomposti comb. nov.

Description of Bieblia gen. nov.

Biebl’i.a (N.L. fem. n. Bieblia, named after the German microbiologist Hanno Biebl for his work on marine Rhodobacteraceae).

Facultatively aerobic or anaerobic, Gram-negative, non-motile, rods or ovoids, mostly autotrophs. bacteriochlorophyll αnd carotenoids present. The G+C content is 64–69%. The type species is Bieblia veldkampii, comb. nov. The genus has been separated from Rhodobacter based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Caenibius gen. nov.

Cae.ni’bi.us (L. neut. n. caenum, mud, referring to the isolation of the type strain from activated sludge; N.L. masc. n. bius, life (from Gr. n. bios); N.L. masc. n. Caenibius, sludge life).

The description is as given for Caenibius tardaugens, comb. nov., which is the type species. The genus has been separated from Novosphingobium based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Cypionkella gen. nov.

Cy.pi.on.kel’la (N.L. fem. dim. n. Cypionkella, named after the German microbiologist Heribert Cypionka for his work on marine Rhodobacteraceae).

Cells are Gram-negative, rod-shaped or oval, non-motile, mostly aerobic, heterotrophic. Catalase and oxidase positive. The predominant ubiquinone is Q-10. The major polar lipids are phosphatidylglycerol and phosphatidylcholin. The major fatty acid is C_18:1 ω7c. The G+C content is 60–62%. The type species is Cypionkella psychrotolerans, comb. nov. The genus has been separated from Pseudorhodobacter based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Elkelangia gen. nov.

El.ke.lang’i.a (N.L. fem. n. Elkelangia, named after Elke Lang, a German microbiologist known for her work as long term curator of Gram negative bacteria at DSMZ).

The description is as given for Elkelangia baekryungensis, comb. nov., which is the type species. The genus has been separated from Sphingopyxis based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Meinhardsimonia gen. nov.

Mein.hard.si.mon’i.a (N.L. fem. n. Meinhardsimonia, named after the German microbiologist Meinhard Simon for his work on marine Rhodobacteraceae).

The description is as given for Meinhardsimonia xiamenensis, comb. nov., which is the type species. The genus has been separated from Albidovulum based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Neopararhizobium gen. nov.

Ne.o.pa.ra.rhi.zo’bi.um (Gr. pref. neo, new; N.L. neut. n. Pararhizobium, the genus Pararhizobium; N.L. neut. n. Neopararhizobium, new Pararhizobium).

The type species is Neopararhizobium haloflavum, comb. nov. The genus has been separated from Pararhizobium based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Pacificitalea gen. nov.

Pa.ci.fi.ci.ta’le.a (L. masc. adj. pacificus, peaceful, referring to the Pacific Ocean; L. fem. n. talea, a rod; N.L. fem. n. Pacificitalea, a rod isolated from the Pacific Ocean).

The description is as given for Pacificitalea manganoxidans, comb. nov., which is the type species. The genus has been separated from Celeribacter based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Pseudoprimorskyibacter gen. nov.

Pseu.do.pri.mor.sky.i.bac’ter (Gr. masc. adj. pseudês, false; N.L. n. primorsky/-yos, primorsky kray, a far-Eastern region of the Russian federation where the first strains were isolated; N.L. masc. n. bacter, a rod; N.L. masc. n. Pseudoprimorskyibacter, like Primorskyibacter, referring to the close relationship to the genus Primorskyibacter).

The description is as given for Pseudoprimorskyibacter insulae comb. nov., which is the sole and type species. The genus has been separated from Primorskyibacter based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Vannielia gen. nov.

Van.niel’i.a (N.L. fem. n. Vannielia, named to honor Cornelis Bernardus van Niel, and his many contributions to microbiology).

The description is as given for Vannielia litorea, comb. nov., which is the type species. The genus has been separated from Oceanicola based on physiology and phylogenetic analyses of genome and 16S rRNA gene sequences.

Description of Wagnerdoeblera gen. nov.

Wag.ner.doeb’ler.a (N.L. fem. n. Wagnerdoeblera, named after the German microbiologist Irene Wagner-Döbler for her work on marine Rhodobacteraceae).

Gram-negative, non-motile, non-spore-forming bacteria. Cells are irregular rod-shaped. Aerobic. Oxidase and catalase positive. The major ubiquinone is Q-10. The major polar lipids are phosphatidylethanolamine, phosphatidylglycerol and phosphatidylcholine. The major fatty acid is C_18:1 ω7c. The G+C content as calculated from genome sequences is around 64.9–66.2% while the range provided in the literature is 61.4–64.5 mol%. The type species is Wagnerdoeblera nectariphila, comb. nov.

Taxonomic Consequences: New (Combinations for) Species

Description of Acetobacter ascendens comb. nov., Change of Rank

A. as.cen’dens (L. part. adj. ascendens, ascending, climbing).

Basonym: Acetobacter pasteurianus subsp. ascendensDe Ley and Frateur, 1974 (Approved Lists 1980)

The description is as given for Acetobacter pasteurianus subsp. ascendens (De Ley and Frateur, 1974). The type strain is CCM 3612 = LMG 1590 = NCCB 51001.

Description of Acetobacter paradoxus comb. nov., Change of Rank

A. pa.ra.do’xus (L. masc. adj. paradoxus, strange, contrary to all expectation, paradoxical).

Basonym: Acetobacter pasteurianus subsp. ascendensDe Ley and Frateur, 1974 (Approved Lists 1980)

The description is as given for Acetobacter pasteurianus subsp. paradoxus (De Ley and Frateur, 1974). The type strain is LMG 1591 = NCCB 53006.

Description of Actibacterium lipolyticum comb. nov.

A. li.po.ly’ti.cum (Gr. neut. n. lipos, fat; Gr. masc. adj. lytikos, able to loosen, dissolving; N.L. neut. adj. lipolyticum, dissolving fat or lipid).

Basonym: Confluentimicrobium lipolyticumParker et al., 2019

The description is as given for Confluentimicrobium lipolyticum (Park et al., 2014d). The type strain is SSK1-4 = CECT 8621 = KCTC 42136.

Description of Actibacterium naphthalenivorans comb. nov.

A. naph.tha.le.ni.vo’rans (N.L. neut. n. naphthalenum, naphthalene; L. part. adj. vorans, devouring; N.L. part. adj. naphthalenivorans, naphthalene-consuming).

Basonym: Confluentimicrobium naphthalenivoransJeong et al., 2015

The description is as given for Confluentimicrobium naphthalenivorans (Jeong et al., 2015). The type strain is NS6 = DSM 105040 = JCM 30828.

Description of Afipia carboxidovorans comb. nov.

A. car.bo.xi.do’vo.rans (L. masc. n. carbo, charcoal, carbon; Gr. masc. adj. oxys, sour, acid; L. v. voro, devour; N.L. part. adj. carboxidovorans, carbon-acid devouring).

Basonym: Oligotropha carboxidovorans (ex Meyer and Schlegel 1978) Meyer et al. 1994.

The description is as given for Oligotropha carboxidovorans (Meyer et al., 1993). The type strain is OM5 = DSM 1227 = ATCC 49405.

Description of Albibacillus kandeliae comb. nov.

A. kan.de’li.ae (N.L. gen. n. kandeliae, of Kandelia, referring to a genus of mangrove plant).

Basonym: Ruegeria kandeliaeZhang L. et al., 2018

The description is as given for Ruegeria kandeliae (Zhang L. et al., 2018). The type strain is DSM 104293 = MCCC 1K03284.

Description of Allgaiera indica comb. nov.

A. in’di.ca (L. fem. adj. indica, referring to the Indian Ocean, where the type strain was first isolated).

Basonym: Defluviimonas indicaJiang et al., 2014

The description is as given for Defluviimonas indica (Jiang et al., 2014). The type strain is 20V17 = DSM 24802 = JCM 17871.

Description of Allorhizobium oryziradicis comb. nov.

A. o.ry.zi.ra’di.cis (L. fem. n. oryza, rice; L. fem. n. radix/-icis, root; N.L. gen. n. oryziradicis, of the rice root).

Basonym: Rhizobium oryziradicisZhao et al., 2017

The description is as given for Rhizobium oryziradicis (Zhao et al., 2017). The genomic G+C content of the type strain is 55.1%. Its approximate genome size is 5.16 Mbp. The type strain is KCTC 52413.

Description of Allorhizobium taibaishanense comb. nov.

A. tai.bai.shan.en’se (N.L. neut. adj. taibaishanense, of or belonging to the Taibaishan Mountains in the Shaanxi province of China, where the bacterium was isolated).

Basonym: Rhizobium taibaishanenseYao et al., 2012

The description is as given for Rhizobium taibaishanense (Yao et al., 2012). The type strain is DSM 100021 = HAMBI 3214.

Description of Allosediminivita pacifica comb. nov.

A. pa.ci’fi.ca (L. fem. adj. pacifica, peaceful, pertaining to the Pacific Ocean).

Basonym: Roseivivax pacificusWu et al., 2013

The description is as given for Roseivivax pacificus (Wu et al., 2013) with the following modification. The G+C content of the type-strain genome is 66.2%, its approximate size 4.84 Mbp. The type strain is 22DY03 = DSM 29329 = JCM 18866.

Description of Allosphingosinicella indica comb. nov.

A. in’di.ca (L. fem. adj. indica, of India, the origin of the type strain).

Basonym: Sphingomonas indicaNiharika et al., 2012

The description is as given for Sphingomonas indica (Niharika et al., 2012). The genomic G+C content of the type strain is 67.0%. Its approximate genome size is 2.81 Mbp. The type strain is Dd16 = CCM 7882 = DSM 25434.

Description of Allosphingosinicella vermicomposti comb. nov.

A. ver.mi.com.pos’ti (L. neut. n. vermis, worm; N.L. neut. n. compostum, compost; N.L. gen. n. vermicomposti, of vermicompost, referring to the isolation of the type strain from vermicompost).

Basonym: Sphingosinicella vermicompostiYasir et al., 2010

The description is as given for Sphingosinicella vermicomposti (Yasir et al., 2010) with the following modification. The G+C content of the type-strain genome is 62.4%, its approximate size 2.51 Mbp. The type strain is YC7378 = DSM 21593 = KCTC 22446.

Description of Aminobacter carboxidus comb. nov.

A. car.bo’xi.dus (L. masc. adj. carboxidus, intended to mean connected with carbon oxides).

Basonym: Carbophilus carboxidus (ex Nozhevnikova and Zavarzin 1974) Meyer et al. 1994.

The description is as given for Carbophilus carboxidus (Meyer et al., 1993). The type strain is ATCC 51424 = CIP 105722 = DSM 1086.

Description of Bieblia veldkampii comb. nov.

B. veld.kamp’i.i (N.L. gen. n. veldkampii, of Veldkamp, named after Hans Veldkamp, a Dutch microbiologist).

Basonym: Rhodobacter veldkampiiHansen and Imhoff, 1985

The description is as given for Rhodobacter veldkampii (Hansen and Imhoff, 1985) with the following restriction. The G+C content of the type-strain genome is 65.1%, its approximate size 3.26 Mbp. The type strain is BN 714 = DSM 11550 = ATCC 35703 = CIP 103912 = IFO 16458 = NBRC 16458.

Description of Bieblia vinaykumarii comb. nov.

B. vi.nay.ku.ma’ri.i (N.L. gen. n. vinaykumarii, of Vinaykumar, named after the late Dr. M. Vinaykumar, an Indian microbiologist and research supervisor of Ch. V. Ramana. and Ch. Sasikala, who initiated work on anoxygenic phototrophic bacteria in India).

Basonym: Rhodobacter vinaykumariiSrinivas et al., 2007a

The description is as given for Rhodobacter vinaykumarii (Srinivas et al., 2007b). The type strain is CCUG 54311 = DSM 18714 = JCM 14544.

Description of Brucella anthropi comb. nov.

B. an.thro’pi (Gr. masc. n. anthropos, a human being; N.L. gen. n. anthropi, of a human being).

Basonym: Ochrobactrum anthropiHolmes et al., 1988

The description is as given for Ochrobactrum anthropi (Holmes et al., 1988). The type strain is ATCC 49188 = CCUG 24695 = CIP 82.115 = DSMZ 6882 = IFO 15819 = JCM 21032 = LMG 3331 = NBRC 15819 = NCTC 12168.

Description of Brucella ciceri comb. nov.

B. ci.ce’ri (L. gen. n. ciceri, of chickpea (Cicer arietinum), referring to the habitat from which the type strain was isolated).

Basonym: Ochrobactrum ciceriImran et al., 2010

The description is as given for Ochrobactrum ciceri (Imran et al., 2010). The type strain is CCUG 57879 = DSM 22292.

Description of Brucella cytisi comb. nov.

B. cy.ti’si (N.L. masc. n. Cytisus, botanical genus name of the legume Cytisus scoparius; N.L. gen. n. cytisi, of Cytisus, referring to the isolation source of the first strains, nodules of Cytisus scoparius).

Basonym: Ochrobactrum cytisiZurdo-Piñeiro et al., 2007

The description is as given for Ochrobactrum cytisi (Zurdo-Piñeiro et al., 2007). The type strain is CECT 7172 = DSM 19778 = LMG 22713.

Description of Brucella daejeonensis comb. nov.

B. dae.jeon.en’sis (N.L. fem. adj. daejeonensis, of or pertaining to Daejeon, a city in South Korea, from where the type strain was isolated).

Basonym: Ochrobactrum daejeonenseWoo et al., 2011

The description is as given for Ochrobactrum daejeonense (Woo et al., 2011). The type strain is DSM 26944 = JCM 16234 = KCTC 22458.

Description of Brucella endophytica comb. nov.

B. en.do.phy.ti’ca (Gr. pref. endo, within; Gr. neut. n. phyton, plant; L. neut. suff. -icum, adjectival suffix used with the sense of belonging to; N.L. fem. adj. endophytica, within plant, endophytic, because the type strain was isolated from the interior of a plant nodule).

Basonym: Ochrobactrum endophyticumLi L. et al., 2016

The description is as given for Ochrobactrum endophyticum (Li L. et al., 2016). The type strain is CGMCC 1.15082 = DSM 29930 = KCTC 42485.

Description of Brucella gallinifaecis comb. nov.

B. gal.li.ni.fae’cis (L. fem. n. gallina, hen; L. fem. n. faex, faecis, faeces; N.L. gen. n. gallinifaecis, of the faeces of a hen).

Basonym: Ochrobactrum gallinifaecisKämpfer et al., 2003

The description is as given for Ochrobactrum gallinifaecis (Kämpfer et al., 2003). The type strain is CCUG 48291 = CIP 107753 = DSM 15295.

Description of Brucella grignonensis comb. nov.

B. gri.gnon.en’sis (N.L. fem. adj. grignonensis, pertaining to Grignon, region from which the strains were isolated).

Basonym: Ochrobactrum grignonenseLebuhn et al., 2000

The description is as given for Ochrobactrum grignonense (Lebuhn et al., 2000). The genomic G+C content of the type strain is 54.1%. Its approximate genome size is 4.84 Mbp. The type strain is CCUG 46362 = DSM 13338 = LMG 18954 = NBRC 102586.

Description of Brucella haematophila comb. nov.

B. hae.ma.to’phi.la (Gr. neut. n. haima/-atos, Latin transliteration haema/-atos, blood; N.L. adj. philus/-a/-um, from Greek adj. philos/-ê/-on, friend, loving; N.L. fem. adj. haematophila, blood-loving).

Basonym: Ochrobactrum haematophilumKämpfer et al., 2007a

The description is as given for Ochrobactrum haematophilum (Kämpfer et al., 2007a). The type strain is CCUG 38531 = CIP 109452 = DSM 22355.

Description of Brucella intermedia comb. nov.

B. in.ter.me’di.a (L. fem. adj. intermedia, that is between, intermediate).

Basonym: Ochrobactrum intermediumVelasco et al., 1998

The description is as given for Ochrobactrum intermedium (Velasco et al., 1998). The type strain is CCUG 24694 = CIP 105838 = DSM 17986 = IFO 15820 = LMG 3301 = NBRC 15820 = NCTC 12171.

Description of Brucella lupini comb. nov.

B. lu.pi’ni (L. gen. n. lupini, of a lupine, referring to the isolation source of this microorganism, nodules of Lupinus albus).

Basonym: Ochrobactrum lupini Trujillo et al. 2006.

The description is as given for Ochrobactrum lupini (Trujillo et al., 2005). The type strain is DSM 16930 = LMG 22726 = NBRC 102587.

Description of Brucella oryzae comb. nov.

B. o.ry’zae (L. gen. n. oryzae, of rice, pertaining to the habitat from which the first strains were isolated).

Basonym: Ochrobactrum oryzaeTripathi et al., 2006

The description is as given for Ochrobactrum oryzae (Tripathi et al., 2006). The type strain is DSM 17471 = MTCC 4195 = NBRC 102588.

Description of Brucella pecoris comb. nov.

B. pe.co’ris (L. gen. n. pecoris, of livestock).

Basonym: Ochrobactrum pecorisKämpfer et al., 2011

The description is as given for Ochrobactrum pecoris (Kämpfer et al., 2011). The type strain is CCM 7822 = CCUG 60088 = DSM 23868.

Description of Brucella pituitosa comb. nov.

B. pi.tu.i.to’sa (L. fem. adj. pituitosa, full of phlegm, pituilous, intended to mean slimy, referring to the consistency of the colonies after extended incubation).

Basonym: Ochrobactrum pituitosumHuber et al., 2010

The description is as given for Ochrobactrum pituitosum (Huber et al., 2010) with the following addition. The G+C content of the type-strain genome is 53.7%, its approximate size 4.28 Mbp. The type strain is CCUG 50899 = DSM 22207.

Description of Brucella pseudintermedia comb. nov.

B. pseud.in.ter.me’di.a (Gr. neut. adj. pseudês, false; L. fem. adj. intermedia, intermediate, and a specific epithet of the genus Brucella; N.L. fem. adj. pseudintermedia, a false Brucella intermedia).

Basonym: Ochrobactrum pseudintermediumTeyssier et al., 2007

The description is as given for Ochrobactrum pseudintermedium (Teyssier et al., 2007). The genomic G+C content of the type strain is 54.0%. Its approximate genome size is 5.53 Mbp. The type strain is CIP 109116 = DSM 17490.

Description of Brucella pseudogrignonensis comb. nov.

B. pseu.do.gri.gnon.en’sis (Gr. neut. adj. pseudês, false; N.L. fem. adj. grignonensis, a bacterial species epithet; N.L. fem. adj. pseudogrignonensis, a false Brucella grignonensis).

Basonym: Ochrobactrum pseudogrignonenseKämpfer et al., 2007a

The description is as given for Ochrobactrum pseudogrignonense (Kämpfer et al., 2007a) with the following addition. The G+C content of the type-strain genome is 54.0%, its approximate size 5.53 Mbp. The type strain is CCUG 30717 = CIP 109451 = DSM 22354.

Description of Brucella rhizosphaerae comb. nov.

B. rhi.zo.sphae’rae (Gr. fem. n. rhiza, root; L. fem. n. sphaera, a ball, sphere; N.L. n. rhizosphaera, rhizosphere; N.L. gen. n. rhizosphaerae, of the rhizosphere).

Basonym: Ochrobactrum rhizosphaeraeKämpfer et al., 2008

The description is as given for Ochrobactrum rhizosphaerae (Kämpfer et al., 2008). The genomic G+C content of the type strain is 53.0%. Its approximate genome size is 4.90 Mbp. The type strain is CCM 7493 = CCUG 55411 = DSM 19824.

Description of Brucella thiophenivorans comb. nov.

B. thi.o.phe.ni.vo’rans (N.L. neut. n. thiophenum, thiophene; L. pres. part. vorans, devouring; N.L. part. adj. thiophenivorans, thiophene-devouring, referring to the ability to utilize thiophene 2-carboxylate as a sole source of carbon and sulfur).

Basonym: Ochrobactrum thiophenivoransKämpfer et al., 2008

The description is as given for Ochrobactrum thiophenivorans (Kämpfer et al., 2008). The genomic G+C content of the type strain is 51.6%. Its approximate genome size is 4.36 Mbp. The type strain is CCM 7492 = CCUG 55412 = DSM 7216.

Description of Brucella tritici comb. nov.

B. tri.ti’ci (L. neut. n. triticum, wheat, and also the generic name for wheat, Triticum; L. gen. n. tritici, of wheat, of Triticum, from which the strains were isolated).

Basonym: Ochrobactrum triticiLebuhn et al., 2000

The description is as given for Ochrobactrum tritici (Lebuhn et al., 2000). The type strain is CCUG 47104 = DSM 13340 = LMG 18957 = NBRC 102585.

Description of Caenibius tardaugens comb. nov.

C. tard.au’gens (L. masc. adj. tardus, slow; L. pres. part. augens, growing; N.L. part. adj. tardaugens, slowly growing).

Basonym: Novosphingobium tardaugensFujii et al., 2003

The description is as given for Novosphingobium tardaugens (Fujii et al., 2003). The type strain is ARI-1 = DSM 16702 = JCM 11434.

Description of Celeribacter arenosi comb. nov.

C. a.re.no’si (L. gen. n. arenosi, of a sandy place, dwelling in marine sand).

Basonym: Vadicella arenosiRomanenko et al., 2011c

The description is as given for Vadicella arenosi (Romanenko et al., 2011c). The type strain is JCM 17190 = KMM 9024 = NRIC 0787.

Description of Cereibacter azotoformans comb. nov.

C. a.zo.to.for’mans (N.L. neut. n. azotum, (from Fr. n. azote), nitrogen; L. pres. part. formans, forming; N.L. part. adj. azotoformans, nitrogen forming).

Basonym: Rhodobacter azotoformans Hiraishi et al. 1997.

The description is as given for Rhodobacter azotoformans (Hiraishi et al., 1996) with the following modification. The G+C content of the type-strain genome is 68.4%, its approximate size 4.41 Mbp. The type strain is KA25 = JCM 9340 = NBRC 16436.

Description of Cereibacter johrii comb. nov.

C. joh’ri.i (N.L. gen. n. johrii, of B. N. Johri, an eminent and well-known Indian microbiologist).

Basonym: Rhodobacter johriiGirija et al., 2010

The description is as given for Rhodobacter johrii (Girija et al., 2010). The genomic G+C content of the type strain is 69.1%. Its approximate genome size is 4.51 Mbp. The type strain is JA192 = DSM 18678 = JCM 14543.

Description of Cereibacter ovatus comb. nov.

C. o.va’tus (L. masc. adj. ovatus, egg-shaped, ovate).

Basonym: Rhodobacter ovatusSrinivas et al., 2008

The description is as given for Rhodobacter ovatus (Srinivas et al., 2008) with the following modification. The G+C content of the type-strain genome is 66.5%, its approximate size 3.81 Mbp. The type strain is JA234 = CCUG 55049 = JCM 14779.

Description of Cereibacter sphaeroides comb. nov.

C. sphae.ro.i’des (L. fem. n. sphaera, sphere, globe; L. suff. -oides, from Greek suffix eides, from Greek noun eidos, that which is seen, form, shape, figure, resembling, similar; N.L. masc. adj. sphaeroides, spherical).

Basonym: Rhodopseudomonas sphaeroidesvan Niel, 1944 (Approved Lists 1980)

The description is as given for Rhodobacter sphaeroides (Imhoff et al., 1984). The genomic G+C content of the type strain is 68.8%. Its approximate genome size is 4.60 Mbp. The type strain is CECT 300 = DSM 158 = JCM 6121.

Description of Cypionkella aquatica comb. nov.

C. a.qua’ti.ca (L. fem. adj. aquatica, growing or found in water).

Basonym: Pseudorhodobacter aquaticusLi L. et al., 2016

The description is as given for Pseudorhodobacter aquaticus (Li A.-H. et al., 2016). The type strain is DC2N1-10 = CGMCC 1.14433 = KCTC 52040.

Description of Cypionkella collinsensis comb. nov.

C. col.lins.en’sis (N.L. fem. adj. collinsensis, pertaining to Collins, an icecap of Antarctic, from where the type strain was isolated).

Basonym: Pseudorhodobacter collinsensisZhang et al., 2016

The description is as given for Pseudorhodobacter collinsensis (Zhang et al., 2016). The type strain is 4-T-34 = CCTCC AB 2014005 = LMG 28256.

Description of Cypionkella psychrotolerans comb. nov.

C. psy.chro.to’le.rans (Gr. masc. adj. psychros, cold; L. pres. part. tolerans, tolerating; N.L. part. adj. psychrotolerans, cold-tolerating).

Basonym: Pseudorhodobacter psychrotoleransLee et al., 2016

The description is as given for Pseudorhodobacter psychrotolerans (Lee et al., 2016). The type strain is JCM 30764 = KCTC 42640.

Description of Cypionkella sinensis comb. nov.

C. sin.en’sis (N.L. fem. adj. sinensis, pertaining to China, referring to the geographical origin of the type strain).

Basonym: Pseudorhodobacter sinensisLi L. et al., 2016

The description is as given for Pseudorhodobacter sinensis (Li A.-H. et al., 2016). The type strain is Y1R2-4 = CGMCC 1.14435 = KCTC 52039.

Description of Elkelangia baekryungensis comb. nov.

L. baek.ryung.en’sis (N.L. fem. adj. baekryungensis, of Baekryung Island, an island of the Yellow Sea in Korea where the type strain was isolated).

Basonym: Sphingopyxis baekryungensisYoon et al., 2005a

The description is as given for Sphingopyxis baekryungensis (Yoon et al., 2005a). The type strain is DSM 16222 = KCTC 12231.

Description of Magnetospirillum chandramohanii comb. nov.

M. chan.dra.mo.han’i.i (N.L. gen. n. chandramohanii, of Chandramohan, named after Dr. D. Chandramohan, an Indian marine microbiologist, who has played a crucial role in transforming microbiological research at the National Institute of Oceanography, India, into technologically rewarding activities).

Basonym: Phaeospirillum chandramohaniiAnil Kumar et al., 2009

The description is as given for Phaeospirillum chandramohanii (Anil Kumar et al., 2009). The type strain is JA145 = JCM 14933 = KCTC 5703.

Description of Magnetospirillum fulvum comb. nov.

M. ful’vum (L. neut. adj. fulvum, deep yellow, reddish yellow, tawny).

Basonym: Rhodospirillum fulvumvan Niel, 1944 (Approved Lists 1980)

The description is as given for Phaeospirillum fulvum (Imhoff et al., 1998). The type strain is NCIMB 11762 = ATCC 15798 = DSM 113.

Description of Magnetospirillum molischianum comb. nov.

M. mo.lisch.i.a’num (N.L. neut. adj. molischianum, pertaining to Molisch, named for H. Molisch, an Austrian botanist).

Basonym: Rhodospirillum molischianumGiesberger, 1947 (Approved Lists 1980)

The description is as given for Phaeospirillum molischianum (Imhoff et al., 1998). The genomic G+C content of the type strain is 61.5%. Its approximate genome size is 3.81 Mbp. The type strain is ATCC 14031 = DSM 120 = LMG 4354.

Description of Magnetospirillum oryzae comb. nov.

M. o.ry’zae (L. gen. n. oryzae, of rice, pertaining to the isolation of the type strain from rice paddy soil).

Basonym: Phaeospirillum oryzaeLakshmi et al., 2011

The description is as given for Phaeospirillum oryzae (Lakshmi et al., 2011). The type strain is JA317 = KCTC 5704 = NBRC 104938.

Description of Magnetospirillum tilakii comb. nov.

M. ti.la’ki.i (N.L. gen. n. tilakii, of Tilak, named after Dr. K. V. B. R. Tilak, an eminent microbiologist in India).

Basonym: Phaeospirillum tilakiiRaj et al., 2012

The description is as given for Phaeospirillum tilakii (Raj et al., 2012). The type strain is JA492 = KCTC 15012 = NBRC 107650.

Description of Meinhardsimonia xiamenensis comb. nov.

M. xia.men.en’sis (N.L. fem. adj. xiamenensis, of or pertaining to Xiamen, the city where the organism was first isolated).

Basonym: Albidovulum xiamenenseYin et al., 2012

The description is as given for Albidovulum xiamenense (Yin et al., 2012). The genomic G+C content of the type strain is 68.2%. Its approximate genome size is 3.13 Mbp. The type strain is CGMCC 1.10789 = DSM 24422 = LMG 26247 = MCCC 1A06317.

Description of Mycoplana subbaraonis comb. nov.

M. sub.ba.ra.o’nis (N.L. gen. n. subbaraonis, of Subba Rao, named after Professor N. S. Subba Rao, an eminent microbiologist who significantly contributed to the knowledge of Rhizobium biofertilizers in India).

Basonym: Rhizobium subbaraonisRamana V.V. et al., 2013

The description is as given for Rhizobium subbaraonis (Ramana C.V. et al., 2013) with the following modification. The G+C content of the type-strain genome is 63.1%, its approximate size 6.58 Mbp. The type strain is DSM 24765 = KCTC 23614.

Description of Neopararhizobium haloflavum comb. nov.

N. ha.lo.fla’vum (Gr. masc. n. hals/halos, salt; L. masc. adj. flavus, yellow; N.L. neut. adj. haloflavum, salty and yellow).

Basonym: Pararhizobium haloflavumShen et al., 2018

The description is as given for Pararhizobium haloflavum (Shen et al., 2018). The type strain is KCTC 52582 = MCCC 1K03228.

Description of Neorhizobium vignae comb. nov.

N. vi’gnae (N.L. gen. n. vignae, of Vigna, referring to the fact that the majority of strains were isolated from the mung bean, Vigna radiata).

Basonym: Rhizobium vignaeRen et al., 2011

The description is as given for Rhizobium vignae (Ren et al., 2011). The type strain is HAMBI 3039 = DSM 25378 = LMG 25447.

Description of Niveispirillum irakense comb. nov.

N. i.rak.en’se (N.L. neut. adj. irakense, pertaining to the country of Iraq).

Basonym: Azospirillum irakense Khammas et al. 1991.

The description is as given for Azospirillum irakense (Khammas et al., 1989). The genomic G+C content of the type strain is 63.0%. Its approximate genome size is 5.45 Mbp. The type strain is KBC1 = ATCC 51182 = DSM 11586.

Description of Pacificitalea manganoxidans comb. nov.

P. man.gan.o’xi.dans (N.L. neut. n. manganum, manganese; N.L. pres. part. oxidans, oxidizing; N.L. part. adj. manganoxidans, manganese-oxidizing).

Basonym: Celeribacter manganoxidansWang Y.X. et al., 2015

The description is as given for Celeribacter manganoxidans (Wang L. et al., 2015). The type strain is DY2-5 = DSM 27541 = JCM 19384.

Description of Pelagimonas phthalicica comb. nov.

P. phtha.li’ci.ca (N.L. neut. n. acidum phthalicum, phthalic acid; L. fem. suff. -ica, suffix used with the sense of belonging to; N.L. fem. adj. phthalicica, belonging to phthalic acid, referring to the substrate phthalic acid that can be utilized by the species).

Basonym: Tropicibacter phthalicicusIwaki et al., 2012

The description is as given for Tropicibacter phthalicicus (Iwaki et al., 2012). The type strain is DSM 26923 = JCM 17793 = KCTC 23703.

Description of Pseudoprimorskyibacter insulae comb. nov.

P. in’su.lae (L. gen. n. insulae, of an island, referring to the source of isolation of the type strain).

Basonym: Primorskyibacter insulaeParker et al., 2019

The description is as given for Primorskyibacter insulae (Park et al., 2015a). The type strain is CECT 8871 = KCTC 42602.

Description of Pseudovibrio exalbescens comb. nov.

P. ex.al.bes’cens (L. part. adj. exalbescens, (from L. v. exalbesco) becoming white, growing white, referring to the fading color of maturing colonies).

Basonym: Nesiotobacter exalbescensDonachie et al., 2006

The description is as given for Nesiotobacter exalbescens (Donachie et al., 2006) with the following modification. The G+C content of the type-strain genome is 55.1%, its approximate size 4.15 Mbp. The type strain is LA33B = DSM 16456 = ATCC BAA-994.

Description of Rhizorhabdus histidinilytica comb. nov.

R. his.ti.di.ni.ly’ti.ca (N.L. neut. n. histidinum, histidine; N.L. fem. adj. lytica, from Greek fem. adj. lytikê, able to loose, able to dissolve; N.L. fem. adj. histidinilytica, histidine-dissolving).

Basonym: Sphingomonas histidinilyticaNigam et al., 2010

The description is as given for Sphingomonas histidinilytica (Nigam et al., 2010). The type strain is CCM 7545 = DSM 24951 = MTCC 9473.

Description of Rhizorhabdus starnbergensis comb. nov.

R. starn.berg.en’sis (N.L. fem. adj. starnbergensis, of or pertaining to Lake Starnberg, Bavaria, Germany, from where the organism was isolated).

Basonym: Sphingomonas starnbergensisChen W.-M. et al., 2013

The description is as given for Sphingomonas starnbergensis (Chen H. et al., 2013). The type strain is DSM 25077 = LMG 26763.

Description of Rhizorhabdus wittichii comb. nov.

R. wit.tich.i’i (N.L. gen. n. wittichii, of Wittich, referring to Rolf-Michael Wittich, the German bacteriologist who first isolated this potent metabolizer of dibenzo-p-dioxin from the water of the river Elbe and described the metabolism of the compound by this organism).

Basonym: Sphingomonas wittichiiYabuuchi et al., 2001

The description is as given for Sphingomonas wittichii (Kim M.C. et al., 2016). The type strain is CCUG 31198 = DSM 6014 = JCM 10273.

Description of Roseibium aggregatum comb. nov.

R. ag.gre.ga’tum (L. neut. adj. aggregatum, joined together).

Basonym: Stappia aggregata (ex Ahrens 1968) Uchino et al. 1999.

The description is as given for Labrenzia aggregata (Biebl et al., 2007). The type strain is ATCC 25650 = DSM 13394 = IFO 16684 = JCM 20685 = LMG 122 = NBRC 16684 = NCIMB 2208.

Description of Roseibium album comb. nov.

R. al’bum (L. neut. adj. album, white).

Basonym: Stappia alba Pujalte et al. 2006.

The description is as given for Labrenzia alba (Biebl et al., 2007) with the following addition. The genomic G+C content of the type strain is 56.4%. Its approximate genome size is 6.90 Mbp. The type strain is CECT 5095 = CIP 108402 = DSM 18320 = DSM 18380.

Description of Roseibium alexandrii comb. nov.

R. a.le.xan’dri.i (N.L. gen. n. alexandrii, of Alexandrium, the genus name of the dinoflagellate Alexandrium lusitanicum, the source of isolation of the type strain).

Basonym: Labrenzia alexandriiBiebl et al., 2007

The description is as given for Labrenzia alexandrii (Biebl et al., 2007). The type strain is DSM 17067 = NCIMB 14079.

Description of Roseibium marinum comb. nov.

R. ma.ri’num (L. neut. adj. marinum, of the sea, marine).

Basonym: Stappia marina Kim et al. 2006.

The description is as given for Labrenzia marina (Biebl et al., 2007). The type strain is DSM 17023 = KCTC 12288.

Description of Roseibium salinum comb. nov.

R. sa.li’num (N.L. neut. adj. salinum, salted, referring to the saline habitat of the micro-organism).

Basonym: Labrenzia salinaCamacho et al., 2016

The description is as given for Labrenzia salina (Camacho et al., 2016). The type strain is CECT 8816 = DSM 29163.

Description of Roseibium suaedae comb. nov.

R. su.ae’dae (N.L. gen. n. suaedae, of the plant Suaeda corniculata, referring to the isolation of the type strain from the roots of Suaeda corniculata).

Basonym: Labrenzia suaedaeBibi et al., 2014

The description is as given for Labrenzia suaedae (Bibi et al., 2014) with the following modification. The G+C content of the type-strain genome is 60.2%, its approximate size 5.14 Mbp. The type strain is DSM 22153 = KACC 13772.

Description of Roseovarius litorisediminis comb. nov.

R. li.to.ri.se.di’mi.nis (L. neut. n. litus/-oris, the seashore, coast; L. n. sedimen/-inis, sediment; N.L. gen. n. litorisediminis, of a coastal sediment, tidal flat sediment).

Basonym: Pelagicola litorisediminis Park et al. 2013.

The description is as given for Pelagicola litorisediminis (Park et al., 2013a). The type strain is D1-W8 = CECT 8287 = KCTC 32327.

Description of Ruegeria meteori comb. nov.

R. me.te’or.i (N.L. gen. n. meteori, of meteor, after the German research vessel Meteor).

Basonym: Agrobacterium meteoriRüger and Höfle, 1992

Agrobacterium meteori (Rüger and Höfle, 1992) with the following modification. The G+C content of the type-strain genome is 56.5%, its approximate size 4.83 Mbp. The type strain is ATCC 700001 = CECT 4293 = DSM 5824.

Description of Salipiger pacificus comb. nov.

S. pa.ci’fi.cus (L. masc. adj. pacificus, peacemaking, pacific, and by extension pertaining to the Pacific Ocean, the origin of the type strain).

Basonym: Yangia pacificaDai et al., 2006

The description is as given for Yangia pacifica (Dai et al., 2006). The genomic G+C content of the type strain is 66.3%. Its approximate genome size is 6.14 Mbp. The type strain is CGMCC 1.3455 = DSM 26894 = JCM 12573.

Description of Tabrizicola blastica comb. nov.

T. bla’sti.cus (Gr. adj. blastikos/-ê/-on, budding, sprouting; N.L. fem. adj. blastica, budding, apt to bud).

Basonym: Rhodopseudomonas blastica Eckersley and Dow 1981.

The description is as given for Rhodobacter blasticus (Kawasaki et al., 1993). The type strain is ATCC 33485 = CIP 104374 = DSM 2131 = DSM 26431 = IFO 16437 = LMG 4305 = NBRC 16437 = NCIMB 11576.

Description of Thalassobius litoralis comb. nov.

T. li.to.ra’lis (L. masc. adj. litoralis, of or belonging to the seashore, referring to the supralitoral habitat from which the type strain was isolated).

Basonym: Lutimaribacter litoralisIwaki et al., 2013

The description is as given for Lutimaribacter litoralis (Iwaki et al., 2013). The type strain is DSM 29506 = JCM 17792 = KCTC 23660.

Description of Thalassobius taeanensis comb. nov.

T. tae.an.en’sis (N.L. masc. adj. taeanensis, of or belonging to Taean, from where the organism was isolated).

Basonym: Litorimicrobium taeanenseJin et al., 2011

The description is as given for Litorimicrobium taeanense (Jin et al., 2011). The genomic G+C content of the type strain is 60.5%. Its approximate genome size is 4.02 Mbp. The type strain is DSM 22007 = KACC 13706.

Description of Tranquillimonas rosea comb. nov.

T. ro’se.a (L. fem. adj. rosea, rose-colored, pink).

Basonym: Roseivivax roseusZhang et al., 2014a.

The description is as given for Roseivivax roseus (Zhang et al., 2014c) with the following modification. The G+C content of the type-strain genome is 67.8%, its approximate size 4.23 Mbp. The type strain is BH87090 = DSM 23042 = KCTC 22650.

Description of Tritonibacter mobilis comb. nov.

T. mo.bi’lis (L. masc. adj. mobilis, movable, motile).

Basonym: Ruegeria mobilisMuramatsu et al., 2007

The description is as given for Epibacterium mobile (Wirth and Whitman, 2018). The type strain is CIP 109181 = DSM 23403 = NBRC 101030.

Description of Tritonibacter multivorans comb. nov.

T. mul.ti.vo’rans (L. masc. adj. multus, many, numerous; L. v. vorare, to devour, swallow; N.L. part. adj. multivorans, devouring many, referring to the utilization of numerous different substrates for growth).

Basonym: Tropicibacter multivoransLucena et al., 2012

The description is as given for Epibacterium multivorans (Wirth and Whitman, 2018). The genomic G+C content of the type strain is 59.7%. Its approximate genome size is 4.15 Mbp. The type strain is CECT 7557 = DSM 26470 = KCTC 23350.

Description of Tritonibacter scottomollicae comb. nov.

T. scot.to.mol’li.cae (N.L. gen. n. scottomollicae, of Scotto-Mollica, in honor of Dr. Victoria Scotto-Mollica and Dr. Alfonso Mollica, both of whom were pioneers in the field of microbe-induced corrosion of steels and the generation of electroactive seawater biofilms).

Basonym: Ruegeria scottomollicae Vandecandelaere et al. 2008

The description is as given for Epibacterium scottomollicae (Wirth and Whitman, 2018). The type strain is CCUG 55858 = DSM 25328 = LMG 24367.

Description of Tritonibacter ulvae comb. nov.

T. ul’vae (N.L. gen. n. ulvae, of Ulva, the name of a genus of green algae, as the type strain was isolated from a frond surface of Ulva pertusa Kjellman).

Basonym: Epibacterium ulvaePenesyan et al., 2013

The description is as given for Epibacterium ulvae (Penesyan et al., 2013). The type strain is DSM 24752 = LMG 26464.

Description of Vannielia litorea comb. nov.

V. li.to.re’a (L. fem. adj. litorea, living near the sea, of or belonging to the seashore).

Basonym: Oceanicola litoreusPark et al., 2013a

The description is as given for Oceanicola litoreus (Park et al., 2013c). The type strain is CCUG 62794 = DSM 29440 = KCTC 32083.

Description of Wagnerdoeblera megaterium comb. nov.

W. me.ga.te’ri.um (Gr. masc. adj. megas, large; Gr. neut. n. teras/-atos, monster, beast; N.L. neut. n. megaterium, big beast).

Basonym: Gemmobacter megateriumLiu K. et al., 2014

The description is as given for Gemmobacter megaterium (Liu J.-J. et al., 2014) with the following modification. The G+C content of the type-strain genome is 64.9%, its approximate size 4.17 Mbp. The type strain is CF17 = DSM 26375 = JCM 18498.

Description of Wagnerdoeblera nectariphila comb. nov.

W. nec.ta.ri’phi.la (L. n. nectar, nectar; N.L. fem. adj. phila, friend, loving (from Gr. masc. adj. philos); N.L. fem. adj. nectariphila, loving nectar, referring to the stimulation of growth by excretions of other bacteria).

Basonym: Gemmobacter nectariphilus (Tanaka et al., 2004) Chen W.-M. et al., 2013

The description is as given for Gemmobacter nectariphilus (Chen W.-M. et al., 2013) with the following modification. The G+C content of the type-strain genome is 66.2%, its approximate size 4.52 Mbp. The type strain is AST4 = DSM 15620 = JCM 11959.

Taxonomic Consequences: New Subspecies

Description of Borrelia garinii subsp. bavariensis subsp. nov.

B. ga.ri’ni.i subsp. ba.va.ri.en’sis (N.L. fem. adj. bavariensis, of or belonging to Bavaria, from where the type strain was isolated).

The description is as given for Borrelia bavariensis (Margos et al., 2013). The type strain is PBi = ATCC BAA-2496 = DSM 23469.

Description of Gluconobacter japonicus subsp. nephelii subsp. nov.

G. ja.po.ni’cus subsp. ne.phe’li.i (N.L. neut. n. Nephelium, the generic name of rambutan, Nephelium lappaceum, a tropical fruit; N.L. gen. n. nephelii, of Nephelium, from which the type strain was isolated).

The description is as given for Gluconobacter nephelii (Kommanee et al., 2011). The type strain is RBY-1 = BCC 36733 = NBRC 106061.

Description of Methylobacterium dichloromethanicum subsp. chloromethanicum subsp. nov.

M. di.chlo.ro.me.tha’ni.cum subsp. chlo.ro.me.tha’ni.cum (N.L. neut. n. chloromethanicum, chloromethane-utilizing).

The description is as given for Methylobacterium chloromethanicum (McDonald et al., 2001). The type strain is CM4 = NCIMB 13688 = VKM B-2223.

Description of Rhizobium marinum subsp. pelagicum subsp. nov.

R. ma.ri’num subsp. pe.la.gi’cum (L. neut. adj. pelagicum, of or belonging to the sea).

The description is as given for Pseudorhizobium pelagicum (Kimes et al., 2015). The type strain is R1-200B4 = CECT 8629 = LMG 28314.

Description of Rhizobium mongolense subsp. loessense subsp. nov.

R. mon.go.len’se subsp. loess.en’se (N.L. neut. adj. loessense, referring to the Loess Plateau of China, where the bacterium was isolated).

The description is as given for Rhizobium loessense (Wei et al., 2003). The type strain is CGMCC 1.3401 = CIP 108030 = LMG 21975.

Description of Rickettsia conorii subsp. gravesii subsp. nov.

R. co.no’ri.i subsp. gra.ves’i.i (N.L. gen. n. gravesii, of Graves, named after Professor Stephen Graves, founder of the Australian Rickettsial Reference Laboratory and a major contributor to rickettsial research in Australia).

The description is as given for Rickettsia gravesii (Abdad et al., 2017). The type strain is BWI-1 = ATCC VR-1664 = CSUR R172.

Description of Rickettsia conorii subsp. heilongjiangensis subsp. nov.

R. co.no’ri.i subsp. hei.long.jiang.en’sis (N.L. fem. adj. heilongjiangensis, from Heilongjiang, the Chinese province where the D. silvarum tick providing the first isolate was collected).

The description is as given for Rickettsia heilongjiangensis (Fournier et al., 2003). The genomic G+C content of the type strain is 32.3%. Its approximate genome size is 1.28 Mbp. The type strain is 054 = ATCC VR-1524 = CSUR 054.

Description of Rickettsia conorii subsp. raoultii subsp. nov.

R. co.no’ri.i subsp. ra.oult’i.i (N.L. gen. n. raoultii, of Raoult, named after Professor Didier Raoult, founder of the WHO-Collaborative Centre for Rickettsioses, Borrelioses and Tick-borne Infections in Marseilles, France, and a major contributor to the study of rickettsiae).

The description is as given for Rickettsia raoultii (Mediannikov et al., 2008). The type strain is Khabarovsk = ATCC VR-1596 = CSUR R3.

Description of Rickettsia tamurae subsp. buchneri subsp. nov.

R. ta.mu’rae subsp. buch’ner.i (N.L. gen. n. buchneri, of Buchner, named in honor of Dr. Paul Buchner, a German biologist who made pioneering contributions to the identification of non-pathogenic tick endosymbionts that are transovarially transmitted).

The description is as given for Rickettsia buchneri (Kurtti et al., 2015). The type strain is ISO-7 = DSM 29016 = ATCC VR-1814.

Description of Tritonibacter mobilis subsp. pelagius subsp. nov.

T. mo’bi.lis subsp. pe.la’gi.us (L. masc. adj. pelagius, of the sea).

The description is as given for Ruegeria pelagia (Lee et al., 2007c). The type strain is HTCC2662 = KCCM 42378 = NBRC 102038.

Taxonomic Consequences: Emendations of Orders

Emended Description of Hyphomicrobiales Douglas et al. 1957 (Approved Lists 1980)

The description is as given for Rhizobiales (Kuykendall, 2005), which is an illegitimate synonym, with the following additions. The order consists of the families Ancalomicrobiaceae, Aurantimonadaceae, Bartonellaceae, Beijerinckiaceae, Nitrobacteraceae, Brucellaceae, Chelatococcaceae, Cohaesibacteraceae, Hyphomicrobiaceae, Methylobacteriaceae, Methylocystaceae, Notoacmeibacteraceae, Phyllobacteriaceae, Rhizobiaceae, Rhodobiaceae, Xanthobacteraceae. In addition to the new families Acuticoccaceae fam. nov., Afifellaceae fam nov., Ahrensiaceae fam. nov., Amorphaceae fam. nov., Blastochloridaceae fam. nov., Breoghaniaceae fam. nov., Devosiaceae fam. nov., Kaistiaceae fam. nov., Parvibaculaceae fam. nov., Phreatobacteraceae fam. nov., Pleomorphomonadaceae fam. nov., Pseudoxanthobacteraceae fam. nov., Stappiaceae fam. nov. and Tepidamorphaceae fam. nov. The type genus is Hyphomicrobium.

Emended Description of Kordiimonadales Kwon et al. 2005

The description is as given before (Kwon et al., 2005) with the following modifications. Cells are Gram-negative, rod-shaped, motile, aerobic heterotrophs. This order houses Kordiimonadaceae fam. nov., which is currently the sole family of the order. The type genus is Kordiimonas.

Emended Description of Rhodobacterales Garrity et al. 2006

The description is as given before (Garrity et al., 2005d) with the following modifications. This order houses Neomegalonemataceae fam. nov. in addition to the previously included families. The type genus is Rhodobacter.

Emended Description of Rhodospirillales Pfennig and Trüper 1971 (Approved Lists 1980)

The description is as given before (Pfennig and Trüper, 1971) with the following modification. The order contains the families Acetobacteraceae and Kiloniellaceae. In addition to the new families Azospirillaceae, fam. nov., Reyranellaceae fam. nov., Rhodospirillaceae, Rhodovibrionaceae, fam. nov., Stellaceae fam. nov., Terasakiellaceae fam. nov., Thalassobaculaceae fam. nov., Thalassospiraceae fam. nov. and Zavarziniaceae fam. nov. The order contains heterotrophs as well as autotrophs. The type genus is Rhodospirillum.

Emended Description of Sphingomonadales Yabuuchi and Kosako 2006

The description is as given before (Garrity et al., 2005d) with the following modification. This order houses Zymomonadaceae, fam. nov. and Sphingosinicellaceae, fam. nov., in addition to the previously included families. The type genus is Sphingomonas.

Emended Description of Spirochaetales Buchanan 1917 (Approved Lists 1980) emend. Gupta et al., 2013

The description is as given before (Gupta et al., 2013) with the following modification. This order contains the families Sphaerochaetaceae, Spirochaetaceae, and Treponemataceae. The order comprises motile as well as non-motile bacteria. The type genus is Spirochaeta.

Taxonomic Consequences: Emendations of Families

Emended Description of Acetobacteraceae(ex Henrici 1939) Gillis and De Ley 1980

The description is as given before (Gillis and De Ley, 1980) with the following modification. This family houses Acidicaldus, Gluconacetobacter, Neoasaia, Roseococcus, Swaminathania, Rubritepida, Saccharibacter, Swingsia, Teichococcus, Dankookia, Crenalkalicoccus, Acetobacter, Ameyamaea, Asaia, Gluconobacter, Granulibacter, Humitalea, Kozakia, Muricoccus, Paracraurococcus, Tanticharoenia, Acidiphilium, Acidisoma, Acidisphaera, Acidocella, Acidomonas, Belnapia, Bombella, Craurococcus, Endobacter, Komagataeibacter, Neokomagataea, Nguyenibacter, Rhodopila, Rhodovarius, Roseomonas (including Muricoccus and Teichococcus), Caldovatus, Elioraea, Siccirubricoccus and Rhodovastum. The type genus is Acetobacter.

Emended Description of Beijerinckiaceae Garrity et al. 2006 emend. Dedysh et al. 2016

The description is as given before (Garrity et al., 2005d; Dedysh et al., 2016), with the following modification. This family houses Beijerinckia (the type genus), Methylocapsa, Methylocella, Methyloferula, Methylorosula, and Methylovirgula.

Emended Description of Cohaesibacteraceae Hwang and Cho 2018 emend. Gallego et al. 2010

The description is as given before (Hwang and Cho, 2008a; Gallego et al., 2010) with the following modification. This family houses Cohaesibacter, which is the type and currently the sole genus of the family.

Emended Description of Ehrlichiaceae Moshkovski 1945 (Approved Lists 1980)

The description is as given for Anaplasmataceae (Philip, 1957; Dumler et al., 2001) with the following additions. This family houses Anaplasma, Ehrlichia (the type genus), Lyticum and Neorickettsia. The genera Aegyptianella, Cowdria and Wolbachia (for which cultures and 16S rRNA gene sequences of type strains are missing) are also tentatively assigned to this family.

Emended Description of Erythrobacteraceae Lee et al. 2005 emend. Xu et al. 2009

The description is as given before (Lee et al., 2005; Xu X.-W. et al., 2009) with the following modification. Some genera of this family contain carotenoids and bacteriochlorophyll α. This family houses Novosphingobium, in addition to the previously included genera. The type genus is Erythrobacter.

Emended Description of Geminicoccaceae Proença et al. 2018

The description is as given before (Proença et al., 2018) with the following modification. Cells are cocci, diplococci or rods. This family houses Arboricoccus and Geminicoccus (the type genus). Defluviicoccus and Tistrella are tentatively assigned to this family as well.

Emended Description of Hyphomicrobiaceae Babudieri 1950 (Approved Lists 1980)

Cells are Gram negative, mostly rod-shaped or ovoid, usually non-motile, predominantly aerobic. The family contains heterotrophic as well as phototrophic genera. The major ubiquinone is Q-10. This family houses Caenibius, Dichotomicrobium, Filomicrobium, Hyphomicrobium (the type genus), Methyloceanibacter, Methyloligella, Pedomicrobium, Rhodomicrobium, and Seliberia.

Emended Description of Kiloniellaceae Wiese et al. 2009

The description is as given before (Wiese et al., 2009) with the following modification. Cells are rod-, spiral- or vibrio-shaped. The G+C content of the DNA is 50-61%. This family houses Kiloniella (the type genus) and Aestuariispira. Marivibrio and Thalassocola are tentatively assigned to this family as well.

Emended Description of Methylobacteriaceae Garrity et al. 2006

The description is as before (Garrity et al., 2005c) with the following modification. Colonies are usually pink or cream colored. Cells are Gram-negative, are rod-shaped, non-motile or motile. Aerobic with Chemoorganoheterotrophic or chemolithoheterotrophic metabolism. The major cellular fatty acids are usually C_18:1 ω7c and C_16:1 ω7c. The major ubiquinone is Q-10. The G+C content is 60-70 mol%. The family currently comprises the genera Microvirga, Methylobacterium (the type genus), Protomonas and Psychroglaciecola. Enterovirga is tentatively assigned to this family.

Emended Description of Methylocystaceae Bowman 2006

The description is as given before (Bowman, 2005) with the following modification. This family houses Methylocystis (the type genus) and Methylosinus.

Emended Description of Nitrobacteraceae Buchanan 1917 (Approved Lists 1980)

The description is as given for Bradyrhizobiaceae (Garrity et al., 2005e), which is an illegitimate synonym of the family, with the following additions. This family houses Afipia, Blastobacter, Bradyrhizobium (including Agromonas), Nitrobacter (the type genus), Pseudolabrys, Pseudorhodoplanes, Rhodopseudomonas, Tardiphaga and Variibacter. Rhodoplanes is tentatively assigned to this family as well.

Emended Description of Notoacmeibacteraceae Huang et al. 2017

The description is as before (Huang et al., 2017), with the following modification. Gram-stain-negative, oxidase- and catalase-positive, aerobic or facultatively anaerobic heterotrophs. The predominant respiratory quinone is Q-10. The major fatty acids are usually C_18:1 ω7c and C_18:1 ω6c. The family contains Mabikibacter, Notoacmeibacter (the type genus), and Zhengella.

Emended Description of Phyllobacteriaceae Mergaert and Swings 2006

The description is as given for Phyllobacteriaceae (Mergaert and Swings, 2005) with the following additions. Cells are Gram-negative, rod, ovoid or coccoid shaped and usually motile by flagella. Predominently aerobic heterotrophs. Generally catalyze and oxidase positive. The predominant respiratory quinone is Q-10. The major polar lipids are phosphatidylcholin, phosphatidylethanolamine and phosphatidylglycerol. The major fatty acids is usually C_18:1 ω7c. The G+C content of the DNA is 56–64%. This family houses Aminobacter, Aquamicrobium, Chelativorans (including Thermovum), Chelatobacter, Corticibacterium, Defluvibacter, Mesorhizobium, Nitratireductor, Oricola, Phyllobacterium (the type genus), Pseudaminobacter, Pseudohoeflea, Roseitalea, and Tianweitania.

Emended Description of Rhizobiaceae Conn 1938 (Approved Lists 1980)

The description is as before (Conn, 1938) with the following modification updated in accordance to later findings (Peix et al., 2005; Rahul et al., 2015). Cells are Gram-negative, mostly rod-shaped, usually motile, aerobic and generally heterotrophic with some autotrophic genera. The major respiratory quinone is Q-10. The G+C content is 49–68%. The family contains Hoeflea, Lentilitoribacter, Martelella, Mycoplana, and Neopararhizobium along with the previously included genera except for Kaistia, which has been removed from the family. The type genus is Rhizobium.

Emended Description of Rhodobacteraceae Garrity et al. 2006

The description is as given before (Garrity et al., 2005b) with the following modification. Rhodobacteraceae houses multiple and diverse genera but not Ahrensia, Gemmobacter, Hyphomonas, Maricaulis, Methylarcula, Pannonibacter, Roseibium, and Stappia as stated in the initial description. Additionally the following new genera are included to this family: Albibacillus, Allgaiera, Allosediminivita, Bieblia, Cypionkella, Meinhardsimonia, Pacificitalea, Pseudoprimorskyibacter, Vannielia, and Wagnerdoeblera. The type genus is Rhodobacter.

Emended Description of Rhodobiaceae Garrity et al. 2006

The description is as given before (Garrity et al., 2005f) with the following modification. This family houses Rhodobium which is the type and currently the sole genus of the family.

Emended Description of Rhodospirillaceae Pfennig and Trueper 1971 (Approved Lists 1980)

The description is as given before (Pfennig and Trüper, 1971), with the following modification. This family houses Caenispirillum, Conglomeromonas, Haematospirillum, Insolitispirillum, Magnetospirillum, Marispirillum, Novispirillum, Pararhodospirillum, Phaeospirillum, Phaeovibrio, Rhodospira, Rhodospirillum (the type genus), Roseospira, Roseospirillum, and Telmatospirillum.

Emended Description of Roseiarcaceae Kulichevskaya et al. 2014

The description is as before (Kulichevskaya et al., 2014) with the following modification after inclusion of Rhodoblastus. The major cellular fatty acids are usually C_16:0, C_16:1, C_18:1 ω7c and sometimes C_19:0 ω8c. The family contains Roseiarcus (the type genus) and Rhodoblastus.

Emended Description of Sneathiellaceae Kurahashi et al. 2008

The description is as before (Kurahashi et al., 2008) with the following modification. This family houses Oceanibacterium and Sneathiella (the type genus). Ferrovibrio, Taonella and Marinibaculum are tentatively assigned to this family.

Emended Description of Sphingomonadaceae Kosako et al. 2000

The description is as given before (Kosako et al., 2000) with the following modification. This family houses Blastomonas (including Erythromonas), Hephaestia, Parablastomonas, Parasphingopyxis, Rhizorhabdus, Rhizorhapis (including Rhizomonas), Sphingobium, Sphingomicrobium, Sphingomonas (the type genus), Sphingopyxis, Sphingorhabdus and Stakelama. Additionally the following new genera are included as well: Allosphingosinicella and Elkelangia.

Emended Description of Spirochaetaceae Swellengrebel 1907 (Approved Lists 1980) emend. Abt et al. 2012 emend. Gupta et al. 2013

The description is as given before (Gupta et al., 2013), with the following modification. This family houses Alkalispirochaeta, Marispirochaeta, Oceanispirochaeta, Salinispira, Sediminispirochaeta, and Spirochaeta (the type genus). The genera Clevelandina, Diplocalyx, Hollandina and Pillotina which lack published 16S rRNA gene sequences are also tentatively assigned to this family.

Emended Description of Temperatibacteraceae Teramoto and Nishijima 2014

Cells are Gram-negative, rod-shaped, motile, aerobic heterotrophs. The major ubiquinone is predominantly Q-10. The major polar lipids are phosphatidylethanolamine and phosphatidylglycerol. The major fatty acids are usually iso-C_19:1 ω9c, iso-C_17:0, C_18:1 ω7c, C_17:1 ω6c, and iso-C_15:0. The family currently comprises the genera Eilatimonas, Kordiimonas and Temperatibacter (the type genus). The family Kordiimonadaceae was proposed earlier on, too (Xu et al., 2014), but has not been validly published yet. The emended family is based on phylogenetic analyses of genome and 16S rRNA gene sequences.

Emended Description of Treponemataceae Robinson 1948 (Approved Lists 1980)

The description is as given before (Robinson, 1948) with the following modification. Cells are mostly motile and of typical spirochaete-like, helical shape or rarely (in Rectinema) rod-shaped to spherical. The family contains Rectinema and Treponema (the type genus).

Emended Description of Xanthobacteraceae Lee et al. 2005

The description is as given before (Lee et al., 2005) with the following modification. Cells are rod-shaped, coccoid or ellipsoidal, aerobic and generally heterotrophic with some autotrophic genera. This family houses Ancylobacter (including Microcyclus), Angulomicrobium, Aquabacter, Azorhizobium, Labrys, Methylorhabdus, Starkeya and Xanthobacter (the type genus).

Taxonomic Consequences: Emendations of Genera

Emended Description of Actibacterium Lucena et al. 2012 emend. Guo et al. 2017

The description is as given before (Lucena et al., 2012; Guo et al., 2017) with additions following the inclusion of Confluentimicrobium lipolyticum and C. naphthalenivorans. Cells can be motile. Additionally another major polar lipid is frequently sulphoquinovosyldiacylglyceride. The type species is Actibacterium mucosum.

Emended Description of Allorhizobium De Lajudie et al. 1998

The description is as given before (de Lajudie et al., 1998) with additions following the inclusion of Rhizobium oryziradicis and Rhizobium taibaishanense. The genomic G+C content is 55.1–62.8%. The type species is Allorhizobium undicola.

Emended Description of Celeribacter Ivanova et al. 2010 emend. Lee et al. 2012

The description is as given before (Ivanova et al., 2010; Lee et al., 2012) with additions that reflect developments in the composition of the genus, notably the removal of Celeribacter manganoxidans and addition of Vadicella arenosi. The genomic G+C content is 56.7–60.9%. The type species is Celeribacter neptunius.

Emended Description of Cereibacter Suresh et al. 2015

The description is as given before (Suresh et al., 2015) with additions following the inclusion of Rhodobacter azotoformans, Rhodobacter johrii, Rhodobacter ovatus and Rhodobacter sphaeroides. Consists of motile (via flagella) as well as non-motile species. The genomic G+C content is 66.5–69.1%. The type species is Cereibacter changlensis.

Emended Description of Cognatishimia Wirth and Whitman 2018

After the inclusion of Thalassobius activus (Arahal et al., 2019) the original description (Wirth and Whitman, 2018) needs to be modified as follows. Cells are Gram-negative, motile or non-motile, rod- or coccus-shaped, aerobic chemotrophs. Some species require sodium chloride for growth. The predominant ubiquinone is Q-10. The major polar lipids are phosphatidylcholine and phosphatidylglycerol. The major fatty acid is C_18:1 ω7c and in some species also C_18:1 ω6c as well as C_10:0 3OH. The G+C content is 54.4–56.3%. The type species is Cognatishimia maritima.

Emended Description of Magnetospirillum Schleifer et al. 1991

The description is as given before (Schleifer et al., 1991) with additions following the inclusion of Phaeospirillum molischianum, Phaeospirillum chandramohanii, Phaeospirillum fulvum, Phaeospirillum oryzae, and Phaeospirillum tilakii. Cells are vibrio-, spiral- or helix-shaped and motile by means of flagella. Includes photoorganoheterotrophic anaerobic and microaerophilic chemoorganotrophic species. Some species are magnetotactic and contain enveloped magnetosomes which are arranged in a chain within the cytoplasm. Mobility and magnetic behavior can be diminished or lost after several subcultivations. The G+C content is 60–71%. The type species is Magnetospirillum gryphiswaldense.

Emended Description of Pelagimonas Hahnke et al. 2013

The description is as given before (Hahnke et al., 2013) with additions following the inclusion of Tropicibacter phthalicicus. Oxidase and catalase variable. The G+C content is 55.2–57.9%. The type species is Pelagimonas varians.

Emended Description of Pseudorhodobacter Uchino et al. 2002 emend. Lee et al. 2013 emend. Lee et al. 2016

The description is as given before (Uchino et al., 2002; Lee et al., 2013, 2016) with additions that reflect developments in the composition of the genus, notably the removal of Pseudorhodobacter aquaticus, P. collinsensis, P. psychrotolerans, and P. sinensis. The G+C content is 58.0–63.1%. The type species is Pseudorhodobacter ferrugineus.

Emended Description of Pseudovibrio Shieh et al. 2004

The description is as given before (Shieh et al., 2004) with additions following the inclusion of Nesiotobacter exalbescens. Mesophilic, growing at 20–35°C, some species able to grow at 45°C. Halophilic, no growth in the absence of NaCl. The G+C content is 47.0–55.8%. The type species is Pseudovibrio denitrificans.

Emended Description of Salipiger Martínez-Cánovas et al. 2004 emend. Wirth and Whitman 2018

The description is as given before (Martínez-Cánovas et al., 2004; Wirth and Whitman, 2018) with additions following the inclusion of Yangia pacifica. Cells are rod-shaped and of variable sizes with widths up to 1 μm and lengths in the range of 1–2.5 μm. The G+C content is 64.3–67.3%. The type species is Salipiger mucosus.

Emended Description of Sphingopyxis Takeuchi et al. 2001 emend. Baik et al. 2013

The description is as given before (Takeuchi et al., 2001; Baik et al., 2013) with additions that reflect developments in the composition of the genus, notably the removal of Sphingopyxis baekryungensis. The G+C content is 63.3–66.4%. The type species is Sphingopyxis macrogoltabida.

Emended Description of Tabrizicola Tarhriz et al. 2013

The description is as given before (Tarhriz et al., 2013) with additions following the inclusion of Rhodobacter blasticus. Cells are ovoid to rod-shaped, aerobic or facultatively anaerobic. Includes chemoheterotrophic as well as photoorganotrophic species. The G+C content is 66.4–66.5%. The type species is Tabrizicola aquatica.

Emended Description of Thalassobius Arahal et al. 2005

The description is as given before (Arahal et al., 2005) with additions that reflect developments in the composition of the genus, notably the removal of Thalassobius activus and addition of Litorimicrobium taeanense as well as Lutimaribacter litoralis. Cells are coccoid, rod-shaped or ovoid. The G+C content is 58.5–66.5%. The type species is Thalassobius mediterraneus.

Emended Description of Tranquillimonas Harwati et al. 2008

The description is as given before (Harwati et al., 2008) with additions following the inclusion of Roseivivax roseus. Includes motile as well as non-motile species. Oxidase variable. Cells are rod-shaped and of variable sizes, 0.2–1.0 μm width and 1.1–2.8 μm in length. The G+C content is 67.3–67.8%. The type species is Tranquillimonas alkanivorans.

Taxonomic Consequences: Emendations of Species

Emended Description of Acetobacter aceti (Pasteur 1864) Beijerinck 1898 (Approved Lists 1980)

The description is as before (Beijerinck, 1898) with the following addition. The G+C content of the type-strain genome is 57.1%, its approximate size 3.63 Mbp.

Emended Description of Acetobacter cerevisiae Cleenwerck et al. 2002

The description is as before (Cleenwerck et al., 2002) with the following restriction. The genomic G+C content of the type strain is 58.0%. Its approximate genome size is 3.09 Mbp.

Emended Description of Acetobacter nitrogenifigens Dutta and Gachhui 2006

The description is as before (Dutta and Gachhui, 2006) with the following modification. The genomic G+C content of the type strain is 60.5%. Its approximate genome size is 4.27 Mbp.

Emended Description of Acetobacter okinawensis Iino et al. 2013

The description is as before (Iino et al., 2012) with the following modification. The genomic G+C content of the type strain is 57.6%. Its approximate genome size is 3.17 Mbp.

Emended Description of Acetobacter orleanensis (Henneberg 1906) Lisdiyanti et al. 2001

The description is as before (Lisdiyanti et al., 2000) with the following addition. The genomic G+C content of the type strain is 56.4%. Its approximate genome size is 3.01 Mbp.

Emended Description of Acetobacter papayae Iino et al. 2013

The description is as before (Iino et al., 2012) with the following modification. The genomic G+C content of the type strain is 59.3%. Its approximate genome size is 3.04 Mbp.

Emended Description of Acetobacter pasteurianus (Hansen 1879) Beijerinck and Folpmers 1916 (Approved Lists 1980)

The description is as before (Beijerinck and Folpmers, 1916) with the following addition. The genomic G+C content of the type strain is 53.1%. Its approximate genome size is 2.98 Mbp.

Emended Description of Acetobacter peroxydans Visser’t Hooft 1925 (Approved Lists 1980)

The description is as before (Visser’t Hooft, 1925) with the following addition. The genomic G+C content of the type strain is 60.5%. Its approximate genome size is 2.71 Mbp.

Emended Description of Acetobacter persici Iino et al. 2013

The description is as before (Iino et al., 2012) with the following modification. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 3.70 Mbp.

Emended Description of Acetobacter syzygii Lisdiyanti et al. 2002

The description is as before (Lisdiyanti et al., 2001) with the following restriction. The G+C content of the type-strain genome is 55.5%, its approximate size 2.67 Mbp.

Emended Description of Acidiphilium angustum Wichlacz et al. 1986

The description is as before (Wichlacz et al., 1986) with the following modification. The genomic G+C content of the type strain is 63.6%. Its approximate genome size is 4.18 Mbp.

Emended Description of Acidiphilium multivorum Wakao et al. 1995

The description is as before (Wakao et al., 1994) with the following restriction. The genomic G+C content of the type strain is 67.0%. Its approximate genome size is 4.21 Mbp.

Emended Description of Acidocella facilis (Wichlacz et al. 1986) Kishimoto et al. 1996

The description is as before (Kishimoto et al., 1995) with the following restriction. The genomic G+C content of the type strain is 64.5%. Its approximate genome size is 3.40 Mbp.

Emended Description of Acidomonas methanolica (Uhlig et al. 1986) Urakami et al. 1989 emend. Yamashita et al. 2004

The description is as before (Yamashita et al., 2004) with the following modification. The G+C content of the type-strain genome is 64.7%, its approximate size 3.68 Mbp.

Emended Description of Actibacterium ureilyticum Lin et al. 2016

The description is as before (Lin et al., 2016) with the following modification. The genomic G+C content of the type strain is 64.3%. Its approximate genome size is 4.15 Mbp.

Emended Description of Acuticoccus yangtzensis Hou et al. 2017

The description is as before (Hou et al., 2015) with the following modification. The genomic G+C content of the type strain is 68.6%. Its approximate genome size is 5.04 Mbp.

Emended Description of Aestuariivita boseongensis Park et al. 2014

The description is as before (Park et al., 2014f) with the following modification. The G+C content of the type-strain genome is 61.1%, its approximate size 3.94 Mbp.

Emended Description of Afifella marina (Imhoff 1984) Urdiain et al. 2009

The description is as before (Urdiain et al., 2008) with the following restriction. The genomic G+C content of the type strain is 63.2%. Its approximate genome size is 3.96 Mbp.

Emended Description of Afipia birgiae La Scola et al. 2002

The description is as before (La Scola et al., 2002) with the following modification. The genomic G+C content of the type strain is 60.8%. Its approximate genome size is 5.30 Mbp.

Emended Description of Afipia clevelandensis Brenner et al. 1992

The description is as before (Brenner et al., 1991) with the following modification. The genomic G+C content of the type strain is 61.7%. Its approximate genome size is 4.39 Mbp.

Emended Description of Afipia felis Brenner et al. 1992

The description is as before (Brenner et al., 1991) with the following modification. The genomic G+C content of the type strain is 60.7%. Its approximate genome size is 4.20 Mbp.

Emended Description of Agrobacterium larrymoorei Bouzar and Jones 2001

The description is as before (Bouzar and Jones, 2001) with the following addition. The G+C content of the type-strain genome is 57.2%, its approximate size 5.16 Mbp.

Emended Description of Agrobacterium nepotum (Pulawska et al. 2012) Mousavi et al. 2016

The description is as before (Mousavi et al., 2015) with the following modification. The genomic G+C content of the type strain is 59.1%. Its approximate genome size is 5.33 Mbp.

Emended Description of Agrobacterium radiobacter (Beijerinck and van Delden 1902) Conn 1942 (Approved Lists 1980) emend. Zhang et al. 2014

The description is as before (Zhang et al., 2014a) with the following restriction. The genomic G+C content of the type strain is 59.4%. Its approximate genome size is 5.50 Mbp.

Emended Description of Agrobacterium tumefaciens (Smith and Townsend 1907) Conn 1942

The description is as before (Conn, 1942) with the following addition. The genomic G+C content of the type strain is 59.3%. Its approximate genome size is 5.66 Mbp.

Emended Description of Albidovulum inexpectatum Albuquerque et al. 2003

The description is as before (Albuquerque et al., 2002) with the following modification. The genomic G+C content of the type strain is 64.8%. Its approximate genome size is 3.00 Mbp.

Emended Description of Albimonas pacifica Li et al. 2013

The description is as before (Li G.-W. et al., 2013) with the following modification. The genomic G+C content of the type strain is 72.9%. Its approximate genome size is 6.03 Mbp.

Emended Description of Aliiroseovarius halocynthiae (Kim et al. 2012) Park et al. 2015

The description is as before (Park et al., 2015c) with the following modification. The G+C content of the type-strain genome is 57.1%, its approximate size 3.40 Mbp.

Emended Description of Aliiroseovarius sediminilitoris (Park and Yoon 2013) Park et al. 201

The description is as before (Park et al., 2015c) with the following modification. The G+C content of the type-strain genome is 58.7%, its approximate size 3.41 Mbp.

Emended Description of Alkalimicrobium pacificum Zhang et al. 2015

The description is as before (Zhang G. et al., 2015) with the following modification. The genomic G+C content of the type strain is 65.0%. Its approximate genome size is 5.79 Mbp.

Emended Description of Alkalispirochaeta alkalica (Zhilina et al. 1996) Sravanthi et al. 2016

The description is as before (Sravanthi et al., 2016) with the following modification. The genomic G+C content of the type strain is 60.5%. Its approximate genome size is 3.34 Mbp.

Emended Description of Alkalispirochaeta americana (Hoover et al. 2003) Sravanthi et al. 2016

The description is as before (Sravanthi et al., 2016) with the following modification. The genomic G+C content of the type strain is 57.5%. Its approximate genome size is 3.31 Mbp.

Emended Description of Alkalispirochaeta sphaeroplastigenens (Vishnuvardhan Reddy et al. 2013) Sravanthi et al. 2016

The description is as before (Sravanthi et al., 2016) with the following addition. The G+C content of the type-strain genome is 60.5%, its approximate size 3.35 Mbp.

Emended Description of Allorhizobium oryzae (Peng et al. 2008) Mousavi et al. 2016

The description is as before (Mousavi et al., 2015) with the following modification. The genomic G+C content of the type strain is 62.8%. Its approximate genome size is 5.39 Mbp.

Emended Description of Allorhizobium vitis (Ophel and Kerr 1990) Mousavi et al. 2016

The description is as before (Mousavi et al., 2015) with the following modification. The genomic G+C content of the type strain is 57.6%. Its approximate genome size is 5.74 Mbp.

Emended Description of Altererythrobacter atlanticus Wu et al. 2014

The description is as before (Wu Y.-H. et al., 2014) with the following modification. The genomic G+C content of the type strain is 61.9%. Its approximate genome size is 3.48 Mbp.

Emended Description of Altererythrobacter ishigakiensis Matsumoto et al. 2011

The description is as before (Matsumoto et al., 2011) with the following modification. The genomic G+C content of the type strain is 56.9%. Its approximate genome size is 2.67 Mbp.

Emended Description of Altererythrobacter marensis Seo and Lee 2010

The description is as before (Seo and Lee, 2010) with the following modification. The genomic G+C content of the type strain is 64.7%. Its approximate genome size is 2.89 Mbp.

Emended Description of Altererythrobacter namhicola Park et al. 2011

The description is as before (Park S.C. et al., 2011) with the following modification. The genomic G+C content of the type strain is 65.0%. Its approximate genome size is 2.59 Mbp.

Emended Description of Amylibacter kogurei Wong et al. 2018

The description is as before (Wong et al., 2018) with the following modification. The G+C content of the type-strain genome is 48.8%, its approximate size 2.98 Mbp.

Emended Description of Ancylobacter aquaticus (Ørskov 1928) Raj 1983

The description is as before (Raj, 1983) with the following addition. The G+C content of the type-strain genome is 67.0%, its approximate size 4.83 Mbp.

Emended Description of Aquamicrobium aerolatum Kämpfer et al. 2009

The description is as before (Kämpfer et al., 2009b) with the following addition. The genomic G+C content of the type strain is 60.1%. Its approximate genome size is 3.64 Mbp.

Emended Description of Aquamicrobium defluvii Bambauer et al. 1998

The description is as before (Bambauer et al., 1998) with the following modification. The G+C content of the type-strain genome is 63.2%, its approximate size 4.52 Mbp.

Emended Description of Arboricoccus pini Proença et al. 2018

The description is as before (Proença et al., 2018) with the following modification. The G+C content of the type-strain genome is 63.2%.

Emended Description of Asaia prunellae Suzuki et al. 2012

The description is as before (Suzuki et al., 2010) with the following modification. The genomic G+C content of the type strain is 55.8%. Its approximate genome size is 3.18 Mbp.

Emended Description of Asticcacaulis benevestitus Vasilyeva et al. 2006

The description is as before (Vasilyeva et al., 2006) with the following modification. The genomic G+C content of the type strain is 58.4%. Its approximate genome size is 4.99 Mbp.

Emended Description of Asticcacaulis excentricus Poindexter 1964 (Approved Lists 1980)

The description is as before (Poindexter, 1964) with the following addition. The genomic G+C content of the type strain is 59.5%. Its approximate genome size is 4.31 Mbp.

Emended Description of Aurantimonas coralicida Denner et al. 2003 emend. Rathsack et al. 2011

The description is as before (Rathsack et al., 2011) with the following addition. The genomic G+C content of the type strain is 66.7%. Its approximate genome size is 4.62 Mbp.

Emended Description of Aureimonas altamirensis (Jurado et al. 2006) Rathsack et al. 2011

The description is as before (Rathsack et al., 2011) with the following modification. The genomic G+C content of the type strain is 64.8%. Its approximate genome size is 4.19 Mbp.

Emended Description of Aureimonas frigidaquae (Kim et al. 2008) Rathsack et al. 2011

The description is as before (Rathsack et al., 2011) with the following modification. The G+C content of the type-strain genome is 66.1%, its approximate size 4.10 Mbp.

Emended Description of Azorhizobium caulinodans Dreyfus et al. 1988

The description is as before (Dreyfus et al., 1988) with the following modification. The genomic G+C content of the type strain is 67.3%. Its approximate genome size is 5.37 Mbp.

Emended Description of Azorhizobium doebereinerae Moreira et al. 2006

The description is as before (Moreira et al., 2006) with the following addition. The genomic G+C content of the type strain is 68.9%. Its approximate genome size is 5.82 Mbp.

Emended Description of Azospirillum brasilense Tarrand et al. 1979

The description is as before (Tarrand et al., 1978)with the following addition. The genomic G+C content of the type strain is 68.3%. Its approximate genome size is 7.14 Mbp.

Emended Description of Azospirillum thiophilum Lavrinenko et al. 2010

The description is as before (Lavrinenko et al., 2010) with the following modification. The genomic G+C content of the type strain is 68.2%. Its approximate genome size is 7.61 Mbp.

Emended Description of Bartonella bacilliformis (Strong et al. 1913) Strong et al. 1915 (Approved Lists 1980)

The description is as before (Strong et al., 1915) with the following addition. The genomic G+C content of the type strain is 38.2%. Its approximate genome size is 1.45 Mbp.

Emended Description of Bartonella clarridgeiae Lawson and Collins 1996

The description is as before (Lawson and Collins, 1996) with the following addition. The genomic G+C content of the type strain is 35.7%. Its approximate genome size is 1.49 Mbp.

Emended Description of Bartonella doshiae Birtles et al. 1995

The description is as before (Birtles et al., 1995) with the following modification. The genomic G+C content of the type strain is 37.9%. Its approximate genome size is 1.77 Mbp.

Emended Description of Bartonella elizabethae (Daly et al. 1993) Brenner et al. 1993

The description is as before (Brenner et al., 1993) with the following modification. The genomic G+C content of the type strain is 38.3%. Its approximate genome size is 1.96 Mbp.

Emended Description of Bartonella grahamii Birtles et al. 1995

The description is as before (Birtles et al., 1995) with the following modification. The genomic G+C content of the type strain is 38.0%. Its approximate genome size is 2.19 Mbp.

Emended Description of Bartonella henselae (Regnery et al. 1992) Brenner et al. 1993

The description is as before (Brenner et al., 1993) with the following modification. The genomic G+C content of the type strain is 38.2%. Its approximate genome size is 1.93 Mbp.

Emended Description of Bartonella koehlerae Droz et al. 2000

The description is as before (Droz et al., 1999) with the following addition. The genomic G+C content of the type strain is 37.6%. Its approximate genome size is 1.74 Mbp.

Emended Description of Bartonella rattaustraliani Gundi et al. 2009

The description is as before (Gundi et al., 2009) with the following addition. The genomic G+C content of the type strain is 38.8%. Its approximate genome size is 2.16 Mbp.

Emended Description of Bartonella rochalimae Eremeeva et al. 2012

The description is as before (Eremeeva et al., 2007) with the following addition. The genomic G+C content of the type strain is 35.8%. Its approximate genome size is 1.54 Mbp.

Emended Description of Bartonella schoenbuchensis Dehio et al. 2001

The description is as before (Dehio et al., 2001) with the following addition. The genomic G+C content of the type strain is 37.6%. Its approximate genome size is 1.74 Mbp.

Emended Description of Bauldia litoralis (Bauld et al. 1983) Yee et al. 2010

The description is as before (Yee et al., 2010) with the following modification. The genomic G+C content of the type strain is 65.8%. Its approximate genome size is 5.09 Mbp.

Emended Description of Beijerinckia indica (Starkey and De 1939) Derx 1950 (Approved Lists 1980)

The description is as before (Derx, 1950) with the following addition. The genomic G+C content of the type strain is 57.0%. Its approximate genome size is 4.42 Mbp.

Emended Description of Beijerinckia mobilis Derx 1950 (Approved Lists 1980)

The description is as before (Derx, 1950) with the following addition. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 4.32 Mbp.

Emended Description of Belnapia moabensis Reddy et al. 2006

The description is as before (Reddy et al., 2006) with the following modification. The genomic G+C content of the type strain is 68.9%. Its approximate genome size is 6.72 Mbp.

Emended Description of Blastochloris viridis (Drews and Giesbrecht 1966) Hiraishi 1997

The description is as before (Hiraishi, 1997) with the following addition. The genomic G+C content of the type strain is 67.9%. Its approximate genome size is 3.72 Mbp.

Emended Description of Blastomonas natatoria (Sly 1985) Sly and Cahill 1997 emend. Hiraishi et al. 2000

The description is as before (Hiraishi et al., 2000) with the following modification. The G+C content of the type-strain genome is 63.4%, its approximate size 4.05 Mbp.

Emended Description of Borrelia bavariensis Margos et al. 2013

The description is as before (Margos et al., 2013) with the following addition. The G+C content of the type-strain genome is 28.1%, its approximate size 0.99 Mbp.

Emended Description of Borrelia bissettiae Margos et al. 2016

The description is as before (Margos et al., 2016) with the following modification. The genomic G+C content of the type strain is 28.3%. Its approximate genome size is 1.40 Mbp.

Emended Description of Borrelia burgdorferi Johnson et al. 1984 emend. Baranton et al. 1992

The description is as before (Baranton et al., 1992) with the following restriction. The genomic G+C content of the type strain is 28.2%. Its approximate genome size is 1.52 Mbp.

Emended Description of Borrelia coriaceae Johnson et al. 1987

The description is as before (Johnson et al., 1987) with the following modification. The genomic G+C content of the type strain is 29.3%. Its approximate genome size is 1.57 Mbp.

Emended Description of Borrelia mayonii Pritt et al. 2016

The description is as before (Pritt et al., 2016) with the following addition. The genomic G+C content of the type strain is 27.9%. Its approximate genome size is 1.31 Mbp.

Emended Description of Borrelia valaisiana Wang et al. 1997

The description is as before (Wang et al., 1997) with the following addition. The genomic G+C content of the type strain is 27.5%. Its approximate genome size is 1.26 Mbp.

Emended Description of Bosea lathyri De Meyer and Willems 2012

The description is as before (De Meyer and Willems, 2012) with the following modification. The genomic G+C content of the type strain is 64.8%. Its approximate genome size is 5.92 Mbp.

Emended Description of Bosea robiniae De Meyer and Willems 2012

The description is as before (De Meyer and Willems, 2012) with the following modification. The genomic G+C content of the type strain is 66.3%. Its approximate genome size is 5.28 Mbp.

Emended Description of Brachyspira alvinipulli Stanton et al. 1998

The description is as before (Stanton et al., 1998) with the following modification. The genomic G+C content of the type strain is 26.9%. Its approximate genome size is 3.41 Mbp.

Emended Description of Brachyspira hyodysenteriae (Harris et al. 1972) Ochiai et al. 1998

The description is as before (Ochiai et al., 1997) with the following modification. The genomic G+C content of the type strain is 27.0%. Its approximate genome size is 3.05 Mbp.

Emended Description of Brachyspira innocens (Kinyon and Harris 1979) Ochiai et al. 1998

The description is as before (Ochiai et al., 1997) with the following modification. The genomic G+C content of the type strain is 27.7%. Its approximate genome size is 3.28 Mbp.

Emended Description of Brachyspira intermedia (Stanton et al. 1997) Hampson and La 2006

The description is as before (Hampson and La, 2006) with the following modification. The genomic G+C content of the type strain is 27.2%. Its approximate genome size is 3.31 Mbp.

Emended Description of Brachyspira pilosicoli (Trott et al. 1996) Ochiai et al. 1998

The description is as before (Ochiai et al., 1997) with the following modification. The genomic G+C content of the type strain is 27.9%. Its approximate genome size is 2.56 Mbp.

Emended Description of Bradyrhizobium arachidis Wang et al. 2015

The description is as before (Wang et al., 2013) with the following modification. The genomic G+C content of the type strain is 63.6%. Its approximate genome size is 9.79 Mbp.

Emended Description of Bradyrhizobium elkanii Kuykendall et al. 1993

The description is as before (Kuykendall et al., 1992) with the following addition. The genomic G+C content of the type strain is 63.7%. Its approximate genome size is 9.48 Mbp.

Emended Description of Bradyrhizobium embrapense Delamuta et al. 2015

The description is as before (Delamuta et al., 2015) with the following modification. The genomic G+C content of the type strain is 64.0%. Its approximate genome size is 8.27 Mbp.

Emended Description of Bradyrhizobium japonicum (Kirchner 1896) Jordan 1982

The description is as before (Jordan, 1982) with the following restriction. The genomic G+C content of the type strain is 63.7%. Its approximate genome size is 9.21 Mbp.

Emended Description of Bradyrhizobium jicamae Ramírez-Bahena et al. 2009

The description is as before (Ramírez-Bahena et al., 2009) with the following modification. The genomic G+C content of the type strain is 62.4%. Its approximate genome size is 8.71 Mbp.

Emended Description of Bradyrhizobium lablabi Chang et al. 2011

The description is as before (Chang et al., 2011) with the following modification. The genomic G+C content of the type strain is 62.6%. Its approximate genome size is 8.80 Mbp.

Emended Description of Bradyrhizobium ottawaense Yu et al. 2014

The description is as before (Yu et al., 2014) with the following modification. The genomic G+C content of the type strain is 63.8%. Its approximate genome size is 8.68 Mbp.

Emended Description of Brevinema andersonii Defosse et al. 1995

The description is as before (Defosse et al., 1995) with the following modification. The genomic G+C content of the type strain is 35.2%. Its approximate genome size is 1.50 Mbp.

Emended Description of Brevirhabdus pacifica Wu et al. 2015

The description is as before (Wu et al., 2015) with the following modification. The genomic G+C content of the type strain is 66.2%. Its approximate genome size is 3.30 Mbp.

Emended Description of Brevundimonas aveniformis Ryu et al. 2007

The description is as before (Ryu et al., 2007) with the following modification. The genomic G+C content of the type strain is 65.1%. Its approximate genome size is 2.58 Mbp.

Emended Description of Brevundimonas bacteroides (Poindexter 1964) Abraham et al. 1999

The description is as before (Abraham et al., 1999) with the following addition. The genomic G+C content of the type strain is 68.2%. Its approximate genome size is 3.22 Mbp.

Emended Description of Brevundimonas diminuta (Leifson and Hugh 1954) Segers et al. 1994

The description is as before (Segers et al., 1994) with the following restriction. The genomic G+C content of the type strain is 67.1%. Its approximate genome size is 3.24 Mbp.

Emended Description of Brevundimonas subvibrioides (Poindexter 1964) Abraham et al. 1999

The description is as before (Abraham et al., 1999) with the following addition. The genomic G+C content of the type strain is 68.4%. Its approximate genome size is 3.45 Mbp.

Emended Description of Brevundimonas vesicularis (Büsing et al. 1953) Segers et al. 1994

The description is as before (Segers et al., 1994) with the following modification. The genomic G+C content of the type strain is 66.3%. Its approximate genome size is 3.36 Mbp.

Emended Description of Brevundimonas viscosa Wang et al. 2012

The description is as before (Wang et al., 2012) with the following modification. The genomic G+C content of the type strain is 70.4%. Its approximate genome size is 3.00 Mbp.

Emended Description of Brucella abortus (Schmidt 1901) Meyer and Shaw 1920 (Approved Lists 1980)

The description is as before (Meyer and Shaw, 1920) with the following addition. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 3.29 Mbp.

Emended Description of Brucella canis Carmichael and Bruner 1968 (Approved Lists 1980)

The description is as before (Carmichael and Bruner, 1968) with the following addition. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 3.31 Mbp.

Emended Description of Brucella ceti Foster et al. 2007

The description is as before (Foster et al., 2007) with the following addition. The genomic G+C content of the type strain is 57.3%. Its approximate genome size is 3.27 Mbp.

Emended Description of Brucella inopinata Scholz et al. 2010

The description is as before (Scholz et al., 2010) with the following addition. The genomic G+C content of the type strain is 57.1%. Its approximate genome size is 3.37 Mbp.

Emended Description of Brucella melitensis (Hughes 1893) Meyer and Shaw 1920 (Approved Lists 1980) emend. Verger et al. 1985

The description is as before (Verger et al., 1985) with the following addition. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 3.29 Mbp.

Emended Description of Brucella microti Scholz et al. 2008

The description is as before (Scholz et al., 2008) with the following addition. The genomic G+C content of the type strain is 57.3%. Its approximate genome size is 3.34 Mbp.

Emended Description of Brucella neotomae Stoenner and Lackman 1957 (Approved Lists 1980)

The description is as before (Stoenner and Lackman, 1957) with the following addition. The genomic G+C content of the type strain is 57.3%. Its approximate genome size is 3.30 Mbp.

Emended Description of Brucella ovis Buddle 1956 (Approved Lists 1980)

The description is as before (Buddle, 1956) with the following addition. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 3.28 Mbp.

Emended Description of Brucella suis Huddleson 1929 (Approved Lists 1980)

The description is as before (Huddleson, 1929) with the following addition. The genomic G+C content of the type strain is 57.3%. Its approximate genome size is 3.32 Mbp.

Emended Description of Caenispirillum salinarum Ritika et al. 2012

The description is as before (Ritika et al., 2012) with the following modification. The genomic G+C content of the type strain is 68.8%. Its approximate genome size is 4.95 Mbp.

Emended Description of Camelimonas lactis Kämpfer et al. 2010

The description is as before (Kämpfer et al., 2010b) with the following modification. The G+C content of the type-strain genome is 66.3%, its approximate size 4.17 Mbp.

Emended Description of Caulobacter crescentus Poindexter 1964 (Approved Lists 1980)

The description is as before (Poindexter, 1964) with the following addition. The G+C content of the type-strain genome is 67.2%, its approximate size 4.12 Mbp.

Emended Description of Caulobacter henricii Poindexter 1964 (Approved Lists 1980)

The description is as before (Poindexter, 1964) with the following modification. The genomic G+C content of the type strain is 65.8%. Its approximate genome size is 3.96 Mbp.

Emended Description of Caulobacter mirabilis Abraham et al. 2008

The description is as before (Abraham et al., 2008) with the following modification. The G+C content of the type-strain genome is 69.3%, its approximate size 4.58 Mbp.

Emended Description of Caulobacter vibrioides Henrici and Johnson 1935 (Approved Lists 1980)

The description is as before (Henrici and Johnson, 1935) with the following addition. The G+C content of the type-strain genome is 67.2%, its approximate size 3.97 Mbp.

Emended Description of Celeribacter baekdonensis Lee et al. 2012

The description is as before (Lee et al., 2012) with the following modification. The genomic G+C content of the type strain is 58.1%. Its approximate genome size is 4.44 Mbp.

Emended Description of Celeribacter halophilus (Wang et al. 2012) Lai et al. 2014

The description is as before (Lai et al., 2014) with the following modification. The genomic G+C content of the type strain is 58.1%. Its approximate genome size is 3.87 Mbp.

Emended Description of Celeribacter marinus Baek et al. 2014

The description is as before (Baek et al., 2014) with the following modification. The genomic G+C content of the type strain is 56.2%. Its approximate genome size is 3.10 Mbp.

Emended Description of Celeribacter neptunius Ivanova et al. 2010

The description is as before (Ivanova et al., 2010) with the following modification. The genomic G+C content of the type strain is 61.7%. Its approximate genome size is 4.40 Mbp.

Emended Description of Cereibacter changlensis (Anil Kumar et al. 2007) Suresh et al. 2015

The description is as before (Suresh et al., 2015) with the following modification. The genomic G+C content of the type strain is 68.1%. Its approximate genome size is 4.92 Mbp.

Emended Description of Ciceribacter lividus Kathiravan et al. 2013

The description is as before (Kathiravan et al., 2013) with the following modification. The G+C content of the type-strain genome is 63.2%, its approximate size 4.52 Mbp.

Emended Description of Citreicella aestuarii Park et al. 2011

The description is as before (Park M.S. et al., 2011) with the following modification. The genomic G+C content of the type strain is 64.3%. Its approximate genome size is 4.66 Mbp.

Emended Description of Citreicella thiooxidans Sorokin et al. 2006

The description is as before (Sorokin et al., 2005a) with the following restriction. The genomic G+C content of the type strain is 67.3%. Its approximate genome size is 5.87 Mbp.

Emended Description of Cognatiyoonia koreensis (Weon et al. 2006) Wirth and Whitman 2018

The description is as before (Wirth and Whitman, 2018) with the following modification. The G+C content of the type-strain genome is 57.2%, its approximate size 3.65 Mbp.

Emended Description of Cognatiyoonia sediminum (Liang et al. 2015) Wirth and Whitman 2018

The description is as before (Wirth and Whitman, 2018) with the following modification. The G+C content of the type-strain genome is 54.4%, its approximate size 3.26 Mbp.

Emended Description of Cohaesibacter gelatinilyticus Hwang and Cho 2008 emend. Gallego et al. 2010

The description is as before (Gallego et al., 2010) with the following modification. The genomic G+C content of the type strain is 50.5%. Its approximate genome size is 5.20 Mbp.

Emended Description of Cohaesibacter haloalkalitolerans Sultanpuram et al. 2013

The description is as before (Sultanpuram et al., 2013) with the following modification. The G+C content of the type-strain genome is 57.1%, its approximate size 5.01 Mbp.

Emended Description of Cohaesibacter marisflavi Qu et al. 2011

The description is as before (Qu et al., 2011) with the following modification. The genomic G+C content of the type strain is 53.8%. Its approximate genome size is 5.34 Mbp.

Emended Description of Cribrihabitans marinus Chen et al. 2014

The description is as before (Chen et al., 2014) with the following modification. The genomic G+C content of the type strain is 66.0%. Its approximate genome size is 4.18 Mbp.

Emended Description of Defluviimonas alba Pan et al. 2015

The description is as before (Pan et al., 2015) with the following restriction. The genomic G+C content of the type strain is 66.5%. Its approximate genome size is 4.99 Mbp.

Emended Description of Defluviimonas aquaemixtae Jung et al. 2014

The description is as before (Jung et al., 2014b) with the following modification. The G+C content of the type-strain genome is 64.4%, its approximate size 4.24 Mbp.

Emended Description of Devosia chinhatensis Kumar et al. 2008

The description is as before (Kumar et al., 2008) with the following addition. The genomic G+C content of the type strain is 62.4%. Its approximate genome size is 3.50 Mbp.

Emended Description of Devosia crocina Verma et al. 2009

The description is as before (Verma et al., 2009) with the following addition. The genomic G+C content of the type strain is 61.3%. Its approximate genome size is 3.72 Mbp.

Emended Description of Devosia elaeis Mohd Nor et al. 2017

The description is as before (Mohd Nor et al., 2017) with the following modification. The G+C content of the type-strain genome is 64.1%, its approximate size 3.88 Mbp.

Emended Description of Devosia epidermidihirudinis Galatis et al. 2013

The description is as before (Galatis et al., 2013) with the following addition. The genomic G+C content of the type strain is 61.1%. Its approximate genome size is 3.86 Mbp.

Emended Description of Devosia geojensis Ryu et al. 2008

The description is as before (Ryu et al., 2008) with the following modification. The genomic G+C content of the type strain is 65.9%. Its approximate genome size is 4.47 Mbp.

Emended Description of Devosia soli Yoo et al. 2006

The description is as before (Yoo et al., 2006) with the following modification. The genomic G+C content of the type strain is 61.0%. Its approximate genome size is 4.14 Mbp.

Emended Description of Devosia submarina Romanenko et al. 2013

The description is as before (Romanenko et al., 2013) with the following addition. The G+C content of the type-strain genome is 60.4%, its approximate size 3.98 Mbp.

Emended Description of Dichotomicrobium thermohalophilum Hirsch and Hoffmann 1989

The description is as before (Hirsch and Hoffmann, 1989) with the following restriction. The genomic G+C content of the type strain is 64.3%. Its approximate genome size is 2.99 Mbp.

Emended Description of Ehrlichia chaffeensis Anderson et al. 1992 emend. Dumler et al. 2001

The description is as before (Dumler et al., 2001) with the following addition. The genomic G+C content of the type strain is 30.1%.

Emended Description of Ehrlichia muris Wen et al. 1995 emend. Dumler et al. 2001

The description is as before (Dumler et al., 2001) with the following addition. The genomic G+C content of the type strain is 29.7%. Its approximate genome size is 1.20 Mbp.

Emended Description of Ehrlichia ruminantium (Cowdry 1925) Dumler et al. 2001

The description is as before (Dumler et al., 2001) with the following addition. The genomic G+C content of the type strain is 27.5%. Its approximate genome size is 1.51 Mbp.

Emended Description of Elioraea tepidiphila Albuquerque et al. 2008

The description is as before (Albuquerque et al., 2008) with the following restriction. The genomic G+C content of the type strain is 71.3%. Its approximate genome size is 4.30 Mbp.

Emended Description of Ensifer adhaerens Casida 1982

The description is as before (Casida, 1982) with the following restriction. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 7.28 Mbp.

Emended Description of Ensifer americanus (Toledo et al. 2004) Wang et al. 2015

The description is as before (Wang Y.C. et al., 2013) with the following addition. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 6.65 Mbp.

Emended Description of Ensifer arboris (Nick et al. 1999) Young 2003

The description is as before (Young, 2003) with the following restriction. The genomic G+C content of the type strain is 62.0%. Its approximate genome size is 6.85 Mbp.

Emended Description of Ensifer fredii (Scholla and Elkan 1984) Young 2003

The description is as before (Young, 2003) with the following addition. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 6.58 Mbp.

Emended Description of Ensifer saheli (de Lajudie et al. 1994) Young 2003

The description is as before (Young, 2003) with the following modification. The genomic G+C content of the type strain is 63.6%. Its approximate genome size is 5.99 Mbp.

Emended Description of Ensifer shofinae Chen et al. 2017

The description is as before (Chen et al., 2017c) with the following modification. The G+C content of the type-strain genome is 61.6%.

Emended Description of Ensifer sojae Li et al. 2011

The description is as before (Li et al., 2011) with the following modification. The genomic G+C content of the type strain is 62.0%. Its approximate genome size is 5.96 Mbp.

Emended Description of Epibacterium multivorans (Lucena et al. 2012) Wirth and Whitman 2018

The description is as before (Wirth and Whitman, 2018) with the following modification. The G+C content of the type-strain genome is 59.7%, its approximate size 4.15 Mbp.

Emended Description of Erythrobacter gangjinensis Lee et al. 2010

The description is as before (Lee et al., 2010) with the following modification. The genomic G+C content of the type strain is 62.7%. Its approximate genome size is 2.72 Mbp.

Emended Description of Erythrobacter litoralis Yurkov et al. 1994

The description is as before (Yurkov et al., 1994) with the following modification. The genomic G+C content of the type strain is 65.2%. Its approximate genome size is 3.21 Mbp.

Emended Description of Erythrobacter longus Shiba and Simidu 1982

The description is as before (Shiba and Smidu, 1982) with the following modification. The genomic G+C content of the type strain is 57.4%. Its approximate genome size is 3.60 Mbp.

Emended Description of Erythrobacter marinus Jung et al. 2012

The description is as before (Jung et al., 2012b) with the following modification. The genomic G+C content of the type strain is 59.1%. Its approximate genome size is 2.84 Mbp.

Emended Description of Erythrobacter nanhaisediminis Xu et al. 2010

The description is as before (Xu et al., 2010) with the following modification. The genomic G+C content of the type strain is 62.0%. Its approximate genome size is 2.90 Mbp.

Emended Description of Erythrobacter odishensis Subhash et al. 2013

The description is as before (Subhash et al., 2013) with the following modification. The G+C content of the type-strain genome is 63.7%, its approximate size 3.19 Mbp.

Emended Description of Falsochrobactrum ovis Kämpfer et al. 2013

The description is as before (Kämpfer et al., 2013b) with the following modification. The G+C content of the type-strain genome is 50.7%, its approximate size 3.27 Mbp.

Emended Description of Filomicrobium insigne Wu et al. 2009

The description is as before (Wu et al., 2009) with the following modification. The genomic G+C content of the type strain is 57.4%. Its approximate genome size is 3.85 Mbp.

Emended Description of Fodinicurvata fenggangensis Wang et al. 2009

The description is as before (Wang Y. et al., 2009) with the following modification. The genomic G+C content of the type strain is 61.0%. Its approximate genome size is 3.77 Mbp.

Emended Description of Fulvimarina manganoxydans Ren et al. 2014

The description is as before (Ren et al., 2014) with the following modification. The genomic G+C content of the type strain is 62.9%. Its approximate genome size is 4.83 Mbp.

Emended Description of Fulvimarina pelagi Cho and Giovannoni 2003 emend. Rathsack et al. 2011

The description is as before (Rathsack et al., 2011) with the following modification. The genomic G+C content of the type strain is 61.2%. Its approximate genome size is 3.80 Mbp.

Emended Description of Geminicoccus roseus Foesel et al. 2008

The description is as before (Foesel et al., 2007) with the following modification. The genomic G+C content of the type strain is 68.5%. Its approximate genome size is 5.70 Mbp.

Emended Description of Gemmobacter aquatilis Rothe et al. 1988 emend. Chen et al. 2013

The description is as before (Chen W.-M. et al., 2013) with the following modification. The genomic G+C content of the type strain is 65.1%. Its approximate genome size is 3.96 Mbp.

Emended Description of Gemmobacter caeni (Zheng et al. 2011) Chen et al. 2013

The description is as before (Chen W.-M. et al., 2013) with the following modification. The genomic G+C content of the type strain is 64.7%. Its approximate genome size is 5.13 Mbp.

Emended Description of Gluconacetobacter diazotrophicus (Gillis et al. 1989) Yamada et al. 1998

The description is as before (Yamada et al., 1997) with the following restriction. The genomic G+C content of the type strain is 66.3%. Its approximate genome size is 3.91 Mbp.

Emended Description of Gluconacetobacter entanii Schüller et al. 2000

The description is as before (Schüller et al., 2000) with the following modification. The G+C content of the type-strain genome is 62.6%, its approximate size 3.59 Mbp.

Emended Description of Gluconacetobacter liquefaciens (Asai 1935) Yamada et al. 1998

The description is as before (Yamada et al., 1997) with the following restriction. The G+C content of the type-strain genome is 64.4%, its approximate size 4.18 Mbp.

Emended Description of Gluconobacter cerinus (ex Asai 1935) Yamada and Akita 1984 emend. Katsura et al. 2002

The description is as before (Katsura et al., 2002) with the following restriction. The G+C content of the type-strain genome is 55.6%, its approximate size 3.59 Mbp.

Emended Description of Gluconobacter frateurii Mason and Claus 1989

The description is as before (Mason and Claus, 1989) with the following restriction. The G+C content of the type-strain genome is 56.1%, its approximate size 3.31 Mbp.

Emended Description of Gluconobacter kondonii Malimas et al. 2008

The description is as before (Malimas et al., 2007) with the following modification. The G+C content of the type-strain genome is 58.3%, its approximate size 3.27 Mbp.

Emended Description of Gluconobacter nephelii Kommanee et al. 2011

The description is as before (Kommanee et al., 2011) with the following modification. The G+C content of the type-strain genome is 55.8%, its approximate size 3.16 Mbp.

Emended Description of Gluconobacter oxydans (Henneberg 1897) De Ley 1961 (Approved Lists 1980) emend. Mason and Claus 1989

The description is as before (Mason and Claus, 1989) with the following restriction. The G+C content of the type-strain genome is 60.8%, its approximate size 2.91 Mbp.

Emended Description of Haematobacter massiliensis (Greub and Raoult 2006) Helsel et al. 2007

The description is as before (Helsel et al., 2007) with the following addition. The genomic G+C content of the type strain is 64.6%. Its approximate genome size is 4.13 Mbp.

Emended Description of Haematobacter missouriensis Helsel et al. 2007

The description is as before (Helsel et al., 2007) with the following modification. The genomic G+C content of the type strain is 64.3%. Its approximate genome size is 3.96 Mbp.

Emended Description of Haematospirillum jordaniae Humrighouse et al. 2016

The description is as before (Humrighouse et al., 2016) with the following addition. The genomic G+C content of the type strain is 55.4%. Its approximate genome size is 2.47 Mbp.

Emended Description of Hartmannibacter diazotrophicus Suarez et al. 2014

The description is as before (Suarez et al., 2014) with the following restriction. The G+C content of the type-strain genome is 64.0%, its approximate size 5.45 Mbp.

Emended Description of Hellea balneolensis Alain et al. 2008

The description is as before (Alain et al., 2008) with the following modification. The genomic G+C content of the type strain is 48.4%. Its approximate genome size is 3.21 Mbp.

Emended Description of Henriciella algicola Abraham et al. 2017

The description is as before (Abraham et al., 2017) with the following modification. The G+C content of the type-strain genome is 60.4%, its approximate size 3.20 Mbp.

Emended Description of Henriciella aquimarina Lee et al. 2011

The description is as before (Lee et al., 2011) with the following modification. The genomic G+C content of the type strain is 62.2%. Its approximate genome size is 4.34 Mbp.

Emended Description of Henriciella barbarensis Abraham et al. 2017

The description is as before (Abraham et al., 2017) with the following modification. The G+C content of the type-strain genome is 59.2%, its approximate size 3.33 Mbp.

Emended Description of Henriciella litoralis Lee et al. 2011

The description is as before (Lee et al., 2011) with the following modification. The genomic G+C content of the type strain is 58.9%. Its approximate genome size is 3.78 Mbp.

Emended Description of Henriciella marina Quan et al. 2009

The description is as before (Quan et al., 2009) with the following modification. The genomic G+C content of the type strain is 59.9%. Its approximate genome size is 3.28 Mbp.

Emended Description of Hoeflea halophila Jung et al. 2013

The description is as before (Jung et al., 2013) with the following modification. The G+C content of the type-strain genome is 61.1%, its approximate size 4.19 Mbp.

Emended Description of Hoeflea marina Peix et al. 20055

The description is as before (Peix et al., 2005) with the following modification. The G+C content of the type-strain genome is 65.0%, its approximate size 5.26 Mbp.

Emended Description of Hoeflea olei Rahul et al. 2015

The description is as before (Rahul et al., 2015) with the following modification. The genomic G+C content of the type strain is 65.6%. Its approximate genome size is 4.72 Mbp.

Emended Description of Humitalea rosea Margesin and Zhang 2013

The description is as before (Margesin and Zhang, 2013) with the following modification. The genomic G+C content of the type strain is 69.6%. Its approximate genome size is 4.97 Mbp.

Emended Description of Hwanghaeicola aestuarii Kim et al. 2010

The description is as before (Kim J.M. et al., 2010) with the following modification. The genomic G+C content of the type strain is 66.0%. Its approximate genome size is 4.54 Mbp.

Emended Description of Hyphomicrobium zavarzinii Hirsch 1989

The description is as before (Hirsch, 1989) with the following modification. The genomic G+C content of the type strain is 63.7%. Its approximate genome size is 4.65 Mbp.

Emended Description of Hyphomonas adhaerens Weiner et al. 2000

The description is as before (Weiner et al., 2000) with the following modification. The genomic G+C content of the type strain is 61.3%. Its approximate genome size is 3.67 Mbp.

Emended Description of Hyphomonas hirschiana Weiner et al. 1985

The description is as before (Weiner et al., 1985) with the following modification. The genomic G+C content of the type strain is 61.9%. Its approximate genome size is 3.69 Mbp.

Emended Description of Hyphomonas jannaschiana Weiner et al. 1985

The description is as before (Weiner et al., 1985) with the following modification. The genomic G+C content of the type strain is 61.4%. Its approximate genome size is 3.64 Mbp.

Emended Description of Hyphomonas johnsonii Weiner et al. 2000

The description is as before (Weiner et al., 2000) with the following modification. The genomic G+C content of the type strain is 62.5%. Its approximate genome size is 3.62 Mbp.

Emended Description of Hyphomonas oceanitis Weiner et al. 1985

The description is as before (Weiner et al., 1985) with the following modification. The genomic G+C content of the type strain is 60.2%. Its approximate genome size is 4.28 Mbp.

Emended Description of Hyphomonas polymorpha (ex Pongratz 1957) Moore et al. 1984

The description is as before (Moore et al., 1984) with the following modification. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 4.05 Mbp.

Emended Description of Jannaschia faecimaris Jung and Yoon 2014

The description is as before (Jung and Yoon, 2014) with the following modification. The genomic G+C content of the type strain is 62.0%. Its approximate genome size is 3.81 Mbp.

Emended Description of Jannaschia pohangensis Kim et al. 2008

The description is as before (Kim B.Y. et al., 2008) with the following modification. The genomic G+C content of the type strain is 65.5%. Its approximate genome size is 3.73 Mbp.

Emended Description of Jannaschia rubra Macián et al. 2005

The description is as before (Macián et al., 2005) with the following modification. The genomic G+C content of the type strain is 67.9%. Its approximate genome size is 3.55 Mbp.

Emended Description of Jannaschia seosinensis Choi et al. 2006

The description is as before (Choi et al., 2006) with the following modification. The genomic G+C content of the type strain is 65.3%. Its approximate genome size is 3.83 Mbp.

Emended Description of Jhaorihella thermophila Rekha et al. 2011

The description is as before (Rekha et al., 2011) with the following modification. The genomic G+C content of the type strain is 66.0%. Its approximate genome size is 3.77 Mbp.

Emended Description of Kaistia granuli Lee et al. 2007

The description is as before (Lee et al., 2007a) with the following modification. The genomic G+C content of the type strain is 65.9%. Its approximate genome size is 4.77 Mbp.

Emended Description of Kaistia soli Weon et al. 2008

The description is as before (Weon et al., 2008b) with the following modification. The genomic G+C content of the type strain is 64.5%. Its approximate genome size is 5.24 Mbp.

Emended Description of Kiloniella majae Gerpe et al. 2017

The description is as before (Gerpe et al., 2017) with the following addition. The G+C content of the type-strain genome is 45.5%, its approximate size 4.39 Mbp.

Emended Description of Komagataeibacter europaeus (Sievers et al. 1992) Yamada et al. 2013

The description is as before (Yamada et al., 2012) with the following restriction. The genomic G+C content of the type strain is 61.3%. Its approximate genome size is 4.23 Mbp.

Emended Description of Komagataeibacter kombuchae (Dutta and Gachhui 2007) Yamada et al. 2013

The description is as before (Yamada et al., 2012) with the following modification. The genomic G+C content of the type strain is 59.4%. Its approximate genome size is 3.59 Mbp.

Emended Description of Komagataeibacter oboediens (Sokollek et al. 1998) Yamada et al. 2013

The description is as before (Yamada et al., 2012) with the following modification. The G+C content of the type-strain genome is 61.4%, its approximate size 3.75 Mbp.

Emended Description of Komagataeibacter xylinus (Brown 1886) Yamada et al. 2013

The description is as before (Yamada et al., 2012) with the following restriction. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 3.63 Mbp.

Emended Description of Kordiimonas gwangyangensis Kwon et al. 2005 emend. Yang et al. 2013

The description is as before (Yang et al., 2013) with the following restriction. The genomic G+C content of the type strain is 57.5%. Its approximate genome size is 4.08 Mbp.

Emended Description of Kordiimonas lacus Xu et al. 2011 emend. Wu et al. 2016

The description is as before (Wu et al., 2016) with the following modification. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 4.02 Mbp.

Emended Description of Leisingera methylohalidivorans Schaefer et al. 2002 emend. Vandecandelaere et al. 2008

The description is as before (Vandecandelaere et al., 2008b) with the following modification. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 4.65 Mbp.

Emended Description of Lentibacter algarum Li et al. 2012

The description is as before (Li et al., 2012) with the following modification. The genomic G+C content of the type strain is 55.7%. Its approximate genome size is 3.29 Mbp.

Emended Description of Leptonema illini Hovind-Hougen 1983

The description is as before (Hovind-Hougen, 1979) with the following restriction. The genomic G+C content of the type strain is 54.3%. Its approximate genome size is 4.52 Mbp.

Emended Description of Leptospira alexanderi Brenner et al. 1999

The description is as before (Brenner et al., 1999) with the following modification. The genomic G+C content of the type strain is 40.2%. Its approximate genome size is 4.22 Mbp.

Emended Description of Leptospira alstonii Smythe et al. 2013

The description is as before (Smythe et al., 2013) with the following restriction. The genomic G+C content of the type strain is 42.5%. Its approximate genome size is 4.44 Mbp.

Emended Description of Leptospira biflexa (Wolbach and Binger 1914) Noguchi 1918 (Approved Lists 1980) emend. Faine and Stallman 1982

The description is as before (Faine and Stallman, 1982) with the following addition. The genomic G+C content of the type strain is 38.9%. Its approximate genome size is 3.95 Mbp.

Emended Description of Leptospira fainei Perolat et al. 1998

The description is as before (Perolat et al., 1998) with the following addition. The genomic G+C content of the type strain is 43.5%. Its approximate genome size is 4.29 Mbp.

Emended Description of Leptospira inadai Yasuda et al. 1987

The description is as before (Yasuda et al., 1987) with the following restriction. The genomic G+C content of the type strain is 44.6%. Its approximate genome size is 4.46 Mbp.

Emended Description of Leptospira interrogans (Stimson 1907) Wenyon 1926 (Approved Lists 1980) emend. Faine and Stallman 1982

The description is as before (Faine and Stallman, 1982) with the following addition. The genomic G+C content of the type strain is 35.0%. Its approximate genome size is 4.60 Mbp.

Emended Description of Leptospira kirschneri Ramadass et al. 1992

The description is as before (Ramadass et al., 1992) with the following addition. The genomic G+C content of the type strain is 35.9%. Its approximate genome size is 4.41 Mbp.

Emended Description of Leptospira kmetyi Slack et al. 2009

The description is as before (Slack et al., 2009) with the following modification. The genomic G+C content of the type strain is 44.8%. Its approximate genome size is 4.42 Mbp.

Emended Description of Leptospira licerasiae Matthias et al. 2009

The description is as before (Matthias et al., 2008) with the following modification. The genomic G+C content of the type strain is 41.1%. Its approximate genome size is 4.21 Mbp.

Emended Description of Leptospira meyeri Yasuda et al. 1987

The description is as before (Yasuda et al., 1987) with the following modification. The genomic G+C content of the type strain is 38.0%. Its approximate genome size is 4.24 Mbp.

Emended Description of Leptospira noguchii Yasuda et al. 1987

The description is as before (Yasuda et al., 1987) with the following restriction. The genomic G+C content of the type strain is 35.5%. Its approximate genome size is 4.71 Mbp.

Emended Description of Leptospira santarosai Yasuda et al. 1987

The description is as before (Yasuda et al., 1987) with the following restriction. The genomic G+C content of the type strain is 41.8%. Its approximate genome size is 3.98 Mbp.

Emended Description of Leptospira terpstrae Smythe et al. 2013

The description is as before (Smythe et al., 2013) with the following restriction. The genomic G+C content of the type strain is 38.2%. Its approximate genome size is 4.09 Mbp.

Emended Description of Leptospira vanthielii Smythe et al. 2013

The description is as before (Smythe et al., 2013) with the following restriction. The genomic G+C content of the type strain is 38.9%. Its approximate genome size is 4.23 Mbp.

Emended Description of Leptospira wolbachii Yasuda et al. 1987

The description is as before (Yasuda et al., 1987) with the following modification. The genomic G+C content of the type strain is 39.2%. Its approximate genome size is 4.08 Mbp.

Emended Description of Leptospira wolffii Slack et al. 2008

The description is as before (Slack et al., 2008) with the following modification. The genomic G+C content of the type strain is 45.6%. Its approximate genome size is 4.40 Mbp.

Emended Description of Leptospira yanagawae Smythe et al. 2013

The description is as before (Smythe et al., 2013) with the following restriction. The genomic G+C content of the type strain is 38.2%. Its approximate genome size is 4.06 Mbp.

Emended Description of Limimaricola cinnabarinus (Tsubuchi et al. 2013) Wirth and Whitman 2018

The description is as before (Wirth and Whitman, 2018) with the following modification. The G+C content of the type-strain genome is 66.7%, its approximate size 3.90 Mbp.

Emended Description of Limimaricola hongkongensis (Lau et al. 2004) Wirth and Whitman 2018

The description is as before (Wirth and Whitman, 2018) with the following modification. The G+C content of the type-strain genome is 68.3%, its approximate size 3.19 Mbp.

Emended Description of Limimaricola pyoseonensis (Moon et al. 2010) Wirth and Whitman 2018

The description is as before (Wirth and Whitman, 2018) with the following modification. The G+C content of the type-strain genome is 70.3%, its approximate size 3.91 Mbp.

Emended Description of Limimonas halophila Amoozegar et al. 2013

The description is as before (Amoozegar et al., 2013) with the following modification. The genomic G+C content of the type strain is 69.5%. Its approximate genome size is 3.04 Mbp.

Emended Description of Litoreibacter albidus Romanenko et al. 2011

The description is as before (Romanenko et al., 2011a) with the following modification. The genomic G+C content of the type strain is 59.2%. Its approximate genome size is 3.58 Mbp.

Emended Description of Litoreibacter arenae (Kim et al. 2009) Kim et al. 2012)

The description is as before (Kim et al., 2012a) with the following modification. The genomic G+C content of the type strain is 60.2%. Its approximate genome size is 3.69 Mbp.

Emended Description of Litoreibacter janthinus Romanenko et al. 2011

The description is as before (Romanenko et al., 2011a) with the following modification. The genomic G+C content of the type strain is 57.5%. Its approximate genome size is 3.75 Mbp.

Emended Description of Litorimonas taeanensis Jung et al. 2011 emend. Nedashkovskaya et al. 2013

The description is as before (Nedashkovskaya et al., 2013) with the following modification. The G+C content of the type-strain genome is 46.9%, its approximate size 2.78 Mbp.

Emended Description of Loktanella atrilutea Hosoya and Yokota 2007

The description is as before (Hosoya and Yokota, 2007) with the following modification. The genomic G+C content of the type strain is 64.9%. Its approximate genome size is 4.21 Mbp.

Emended Description of Loktanella cinnabarina Tsubouchi et al. 2013

The description is as before (Tsubouchi et al., 2013) with the following modification. The genomic G+C content of the type strain is 66.7%. Its approximate genome size is 3.90 Mbp.

Emended Description of Loktanella hongkongensis Lau et al. 2004

The description is as before (Lau et al., 2004) with the following modification. The genomic G+C content of the type strain is 68.3%. Its approximate genome size is 3.19 Mbp.

Emended Description of Loktanella koreensis Weon et al. 2006

The description is as before (Weon et al., 2006) with the following modification. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 3.65 Mbp.

Emended Description of Loktanella maritima Tanaka et al. 2014

The description is as before (Tanaka et al., 2014) with the following addition. The genomic G+C content of the type strain is 53.4%. Its approximate genome size is 3.68 Mbp.

Emended Description of Loktanella pyoseonensis Moon et al. 2010

The description is as before (Moon et al., 2010) with the following modification. The genomic G+C content of the type strain is 70.3%. Its approximate genome size is 3.91 Mbp.

Emended Description of Loktanella rosea Ivanova et al. 2005

The description is as before (Ivanova et al., 2005) with the following restriction. The genomic G+C content of the type strain is 57.7%. Its approximate genome size is 3.51 Mbp.

Emended Description of Loktanella sediminum Liang et al. 2015

The description is as before (Liang et al., 2015) with the following modification. The genomic G+C content of the type strain is 54.4%. Its approximate genome size is 3.26 Mbp.

Emended Description of Loktanella tamlensis Lee 2012

The description is as before (Lee, 2012) with the following modification. The genomic G+C content of the type strain is 56.9%. Its approximate genome size is 3.19 Mbp.

Emended Description of Loktanella vestfoldensis Van Trappen et al. 2004

The description is as before (Van Trappen et al., 2004) with the following modification. The genomic G+C content of the type strain is 61.8%. Its approximate genome size is 3.72 Mbp.

Emended Description of Lutibaculum baratangense Anil Kumar et al. 2012

The description is as before (Anil Kumar et al., 2012) with the following modification. The genomic G+C content of the type strain is 68.6%. Its approximate genome size is 4.30 Mbp.

Emended Description of Magnetospirillum gryphiswaldense Schleifer et al. 1992

The description is as before (Schleifer et al., 1991) with the following modification. The genomic G+C content of the type strain is 63.3%. Its approximate genome size is 4.37 Mbp.

Emended Description of Magnetospirillum marisnigri Dziuba et al. 2016

The description is as before (Dziuba et al., 2016) with the following modification. The genomic G+C content of the type strain is 64.7%. Its approximate genome size is 4.62 Mbp.

Emended Description of Magnetovibrio blakemorei Bazylinski et al. 2013

The description is as before (Bazylinski et al., 2013b) with the following modification. The genomic G+C content of the type strain is 54.3%.

Emended Description of Maliponia aquimaris Jung et al. 2016

The description is as before (Jung et al., 2016) with the following modification. The G+C content of the type-strain genome is 67.2%, its approximate size 5.31 Mbp.

Emended Description of Mameliella alba Zheng et al. 2010

The description is as before (Zheng et al., 2010) with the following modification. The genomic G+C content of the type strain is 65.2%. Its approximate genome size is 5.26 Mbp.

Emended Description of Mameliella atlantica Xu et al. 2015

The description is as before (Xu et al., 2015) with the following modification. The genomic G+C content of the type strain is 65.0%. Its approximate genome size is 5.90 Mbp.

Emended Description of Maribius salinus Choi et al. 2007

The description is as before (Choi et al., 2007) with the following modification. The genomic G+C content of the type strain is 67.7%. Its approximate genome size is 3.57 Mbp.

Emended Description of Marinovum algicola (Lafay et al. 1995) Martens et al. 2006

The description is as before (Martens et al., 2006) with the following addition. The genomic G+C content of the type strain is 65.0%. Its approximate genome size is 5.39 Mbp.

Emended Description of Maritimibacter alkaliphilus Lee et al. 2007

The description is as before (Lee et al., 2007b) with the following modification. The genomic G+C content of the type strain is 64.1%. Its approximate genome size is 4.53 Mbp.

Emended Description of Marivita hallyeonensis Yoon et al. 2012

The description is as before (Yoon et al., 2012) with the following modification. The genomic G+C content of the type strain is 60.4%. Its approximate genome size is 4.19 Mbp.

Emended Description of Martelella mediterranea Rivas et al. 2005

The description is as before (Rivas et al., 2005) with the following modification. The genomic G+C content of the type strain is 62.4%. Its approximate genome size is 5.64 Mbp.

Emended Description of Mesorhizobium albiziae Wang et al. 2007

The description is as before (Wang et al., 2007) with the following modification. The genomic G+C content of the type strain is 62.1%. Its approximate genome size is 6.27 Mbp.

Emended Description of Mesorhizobium alhagi Chen et al. 2010

The description is as before (Chen W.-M. et al., 2010) with the following restriction. The genomic G+C content of the type strain is 62.7%. Its approximate genome size is 6.97 Mbp.

Emended Description of Mesorhizobium erdmanii Martínez-Hidalgo et al. 2015

The description is as before (Martínez-Hidalgo et al., 2015) with the following modification. The genomic G+C content of the type strain is 62.7%. Its approximate genome size is 7.02 Mbp.

Emended Description of Mesorhizobium loti (Jarvis et al. 1982) Jarvis et al. 1997 emend. Hameed et al. 2015

The description is as before (Hameed et al., 2015) with the following restriction. The G+C content of the type-strain genome is 62.4%, its approximate size 7.45 Mbp.

Emended Description of Mesorhizobium mediterraneum (Nour et al. 1995) Jarvis et al. 1997

The description is as before (Jarvis et al., 1997) with the following addition. The genomic G+C content of the type strain is 62.0%. Its approximate genome size is 7.20 Mbp.

Emended Description of Mesorhizobium metallidurans Vidal et al. 2009

The description is as before (Vidal et al., 2009) with the following addition. The genomic G+C content of the type strain is 62.5%. Its approximate genome size is 6.23 Mbp.

Emended Description of Mesorhizobium muleiense Zhang et al. 2012

The description is as before (Zhang J.J. et al., 2012) with the following modification. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 6.81 Mbp.

Emended Description of Mesorhizobium qingshengii Zheng et al. 2013

The description is as before (Zheng et al., 2013) with the following modification. The genomic G+C content of the type strain is 62.7%. Its approximate genome size is 7.06 Mbp.

Emended Description of Mesorhizobium soli Nguyen et al. 2015

The description is as before (Nguyen et al., 2015) with the following modification. The G+C content of the type-strain genome is 62.6%, its approximate size 6.27 Mbp.

Emended Description of Mesorhizobium temperatum Gao et al. 2004

The description is as before (Gao et al., 2004) with the following modification. The genomic G+C content of the type strain is 61.9%. Its approximate genome size is 7.17 Mbp.

Emended Description of Methylarcula marina Doronina et al. 2000

The description is as before (Doronina et al., 2000) with the following modification. The G+C content of the type-strain genome is 63.0%, its approximate size 4.23 Mbp.

Emended Description of Methylobacterium chloromethanicum McDonald et al. 2001

The description is as before (McDonald et al., 2001) with the following modification. The G+C content of the type-strain genome is 68.1%, its approximate size 6.18 Mbp.

Emended Description of Methylobacterium gossipiicola Madhaiyan et al. 2012

The description is as before (Madhaiyan et al., 2012) with the following modification. The genomic G+C content of the type strain is 68.7%. Its approximate genome size is 4.52 Mbp.

Emended Description of Methylobacterium komagatae Kato et al. 2008

The description is as before (Kato et al., 2008) with the following modification. The genomic G+C content of the type strain is 67.5%. Its approximate genome size is 5.50 Mbp.

Emended Description of Methylobacterium nodulans Jourand et al. 2004

The description is as before (Jourand et al., 2004) with the following addition. The genomic G+C content of the type strain is 68.4%. Its approximate genome size is 8.84 Mbp.

Emended Description of Methylobacterium organophilum Patt et al. 1976 (Approved Lists 1980)

The description is as before (Patt et al., 1976) with the following modification. The G+C content of the type-strain genome is 71.4%, its approximate size 6.75 Mbp.

Emended Description of Methylobacterium oryzae Madhaiyan et al. 2007

The description is as before (Madhaiyan et al., 2007) with the following modification. The genomic G+C content of the type strain is 69.5%. Its approximate genome size is 6.52 Mbp.

Emended Description of Methylobacterium phyllosphaerae Madhaiyan et al. 2009

The description is as before (Madhaiyan et al., 2009) with the following modification. The genomic G+C content of the type strain is 69.6%. Its approximate genome size is 6.20 Mbp.

Emended Description of Methylobacterium phyllostachyos Madhaiyan and Poonguzhali 2014

The description is as before (Madhaiyan and Poonguzhali, 2014) with the following modification. The genomic G+C content of the type strain is 68.7%. Its approximate genome size is 6.01 Mbp.

Emended Description of Methylobacterium platani Kang et al. 2007

The description is as before (Kang et al., 2007) with the following modification. The genomic G+C content of the type strain is 71.2%. Its approximate genome size is 7.01 Mbp.

Emended Description of Methylobacterium pseudosasicola Madhaiyan and Poonguzhali 2014

The description is as before (Madhaiyan and Poonguzhali, 2014) with the following modification. The genomic G+C content of the type strain is 68.4%. Its approximate genome size is 6.85 Mbp.

Emended Description of Methylobacterium radiotolerans (Ito and Iizuka 1971) Green and Bousfield 1983

The description is as before (Green and Bousfield, 1983) with the following addition. The genomic G+C content of the type strain is 71.0%. Its approximate genome size is 6.90 Mbp.

Emended Description of Methylobacterium tarhaniae Veyisoglu et al. 2013

The description is as before (Veyisoglu et al., 2013) with the following modification. The genomic G+C content of the type strain is 70.4%. Its approximate genome size is 6.74 Mbp.

Emended Description of Methylobrevis pamukkalensis Poroshina et al. 2015

The description is as before (Poroshina et al., 2015) with the following modification. The genomic G+C content of the type strain is 68.9%. Its approximate genome size is 4.38 Mbp.

Emended Description of Methylocapsa acidiphila Dedysh et al. 2002

The description is as before (Dedysh et al., 2002) with the following modification. The genomic G+C content of the type strain is 61.8%. Its approximate genome size is 4.10 Mbp.

Emended Description of Methyloceanibacter caenitepidi Takeuchi et al. 2014

The description is as before (Takeuchi et al., 2014) with the following modification. The genomic G+C content of the type strain is 62.8%. Its approximate genome size is 3.42 Mbp.

Emended Description of Methylocella silvestris Dunfield et al. 2003

The description is as before (Dunfield et al., 2003) with the following modification. The genomic G+C content of the type strain is 63.1%. Its approximate genome size is 4.31 Mbp.

Emended Description of Methylocystis parvus (ex Whittenbury et al. 1970) Bowman et al. 1993

The description is as before (Bowman et al., 1993) with the following restriction. The genomic G+C content of the type strain is 63.4%. Its approximate genome size is 4.48 Mbp.

Emended Description of Methyloferula stellata Vorobev et al. 2011

The description is as before (Vorobev et al., 2011) with the following addition. The genomic G+C content of the type strain is 59.5%. Its approximate genome size is 4.24 Mbp.

Emended Description of Methyloligella halotolerans Doronina et al. 2014

The description is as before (Doronina et al., 2013) with the following modification. The genomic G+C content of the type strain is 63.6%. Its approximate genome size is 3.19 Mbp.

Emended Description of Methylorubrum extorquens (Urakami and Komagata 1984) Green and Ardley 2018

The description is as before (Green and Ardley, 2018) with the following restriction. The G+C content of the type-strain genome is 68.3%, its approximate size 5.72 Mbp.

Emended Description of Methylorubrum populi (Van Aken et al. 2004) Green and Ardley 2018

The description is as before (Green and Ardley, 2018) with the following modification. The genomic G+C content of the type strain is 69.4%. Its approximate genome size is 5.85 Mbp.

Emended Description of Methylorubrum salsuginis (Wang et al. 2007) Green and Ardley 2018

The description is as before (Green and Ardley, 2018) with the following addition. The genomic G+C content of the type strain is 69.6%. Its approximate genome size is 5.32 Mbp.

Emended Description of Methylosinus sporium (ex Whittenbury et al. 1970) Bowman et al. 1993

The description is as before (Bowman et al., 1993) with the following restriction. The G+C content of the type-strain genome is 65.2%, its approximate size 3.79 Mbp.

Emended Description of Methylosinus trichosporium (ex Whittenbury et al. 1970) Bowman et al. 1993

The description is as before (Bowman et al., 1993) with the following modification. The genomic G+C content of the type strain is 65.9%. Its approximate genome size is 4.96 Mbp.

Emended Description of Microvirga guangxiensis Zhang et al. 2009

The description is as before (Zhang et al., 2009) with the following modification. The genomic G+C content of the type strain is 61.4%. Its approximate genome size is 4.72 Mbp.

Emended Description of Microvirga subterranea Kanso and Patel 2003

The description is as before (Kanso and Patel, 2003) with the following modification. The G+C content of the type-strain genome is 65.1%, its approximate size 5.15 Mbp.

Emended Description of Microvirga vignae Radl et al. 2014

The description is as before (Radl et al., 2014) with the following modification. The genomic G+C content of the type strain is 61.1%. Its approximate genome size is 6.47 Mbp.

Emended Description of Monaibacterium marinum Chernikova et al. 2017

The description is as before (Chernikova et al., 2017) with the following modification. The G+C content of the type-strain genome is 58.9%, its approximate size 3.73 Mbp.

Emended Description of Nautella italica Vandecandelaere et al. 2009

The description is as before (Vandecandelaere et al., 2009) with the following restriction. The genomic G+C content of the type strain is 60.0%. Its approximate genome size is 4.06 Mbp.

Emended Description of Neoasaia chiangmaiensis Yukphan et al. 2006

The description is as before (Yukphan et al., 2005) with the following addition. The genomic G+C content of the type strain is 61.5%. Its approximate genome size is 3.41 Mbp.

Emended Description of Neokomagataea thailandica Yukphan et al. 2011

The description is as before (Yukphan et al., 2011) with the following modification. The genomic G+C content of the type strain is 52.4%. Its approximate genome size is 2.49 Mbp.

Emended Description of Neomegalonema perideroedes (Thomsen et al. 2006) Oren 2017

The description is as before (Oren, 2017b) with the following modification. The G+C content of the type-strain genome is 67.2%, its approximate size 3.41 Mbp.

Emended Description of Neorhizobium alkalisoli (Lu et al. 2009) Mousavi et al. 2015

The description is as before (Mousavi et al., 2014) with the following modification. The G+C content of the type-strain genome is 60.3%, its approximate size 5.87 Mbp.

Emended Description of Neorhizobium galegae (Lindström 1989) Mousavi et al. 2015

The description is as before (Mousavi et al., 2014) with the following restriction. The genomic G+C content of the type strain is 61.2%. Its approximate genome size is 6.46 Mbp.

Emended Description of Neorhizobium huautlense (Wang et al. 1998) Mousavi et al. 2015

The description is as before (Mousavi et al., 2014) with the following restriction. The G+C content of the type-strain genome is 60.0%, its approximate size 5.73 Mbp.

Emended Description of Neorickettsia risticii (Holland et al. 1985) Dumler et al. 2001

The description is as before (Dumler et al., 2001) with the following addition. The genomic G+C content of the type strain is 41.3%.

Emended Description of Neorickettsia sennetsu (Misao and Kobayashi 1956) Dumler et al. 2001

The description is as before (Dumler et al., 2001) with the following addition. The genomic G+C content of the type strain is 41.1%.

Emended Description of Neptunicoccus sediminis Zhang et al. 2018

The description is as before (Zhang Y.-J. et al., 2018) with the following addition. The G+C content of the type-strain genome is 57.5%, its approximate size 3.86 Mbp.

Emended Description of Nereida ignava Pujalte et al. 2005

The description is as before (Pujalte et al., 2005) with the following modification. The genomic G+C content of the type strain is 54.1%. Its approximate genome size is 2.84 Mbp.

Emended Description of Nitratireductor aquibiodomus Labbé et al. 2004

The description is as before (Labbé et al., 2004) with the following modification. The genomic G+C content of the type strain is 61.3%. Its approximate genome size is 4.52 Mbp.

Emended Description of Nitratireductor indicus Lai et al. 2011

The description is as before (Lai et al., 2011c) with the following modification. The genomic G+C content of the type strain is 60.9%. Its approximate genome size is 4.99 Mbp.

Emended Description of Nitratireductor pacificus Lai et al. 2011

The description is as before (Lai et al., 2011b) with the following modification. The genomic G+C content of the type strain is 65.5%. Its approximate genome size is 4.47 Mbp.

Emended Description of Nitrobacter winogradskyi Winslow et al. 1917 (Approved Lists 1980)

The description is as before (Winslow et al., 1917) with the following addition. The genomic G+C content of the type strain is 62.0%. Its approximate genome size is 3.40 Mbp.

Emended Description of Novosphingobium acidiphilum Glaeser et al. 2009

The description is as before (Glaeser et al., 2009) with the following addition. The genomic G+C content of the type strain is 64.3%. Its approximate genome size is 3.71 Mbp.

Emended Description of Novosphingobium aromaticivorans (Balkwill et al. 1997) Takeuchi et al. 2001

The description is as before (Takeuchi et al., 2001) with the following restriction. The genomic G+C content of the type strain is 65.1%. Its approximate genome size is 4.23 Mbp.

Emended Description of Novosphingobium barchaimii Niharika et al. 2013

The description is as before (Niharika et al., 2013b) with the following modification. The genomic G+C content of the type strain is 64.0%. Its approximate genome size is 5.31 Mbp.

Emended Description of Novosphingobium capsulatum (Leifson 1962) Takeuchi et al. 2001

The description is as before (Takeuchi et al., 2001) with the following modification. The genomic G+C content of the type strain is 65.7%. Its approximate genome size is 4.84 Mbp.

Emended Description of Novosphingobium fuchskuhlense Glaeser et al. 2013

The description is as before (Glaeser et al., 2013) with the following addition. The genomic G+C content of the type strain is 65.4%. Its approximate genome size is 3.96 Mbp.

Emended Description of Novosphingobium guangzhouense Sha et al. 2017

The description is as before (Sha et al., 2017) with the following modification. The G+C content of the type-strain genome is 63.5%, its approximate size 5.97 Mbp.

Emended Description of Novosphingobium lindaniclasticum Saxena et al. 2013

The description is as before (Saxena et al., 2013) with the following modification. The genomic G+C content of the type strain is 64.6%. Its approximate genome size is 4.86 Mbp.

Emended Description of Novosphingobium mathurense Gupta et al. 2009

The description is as before (Gupta et al., 2009) with the following addition. The genomic G+C content of the type strain is 63.3%. Its approximate genome size is 4.84 Mbp.

Emended Description of Novosphingobium nitrogenifigens Addison et al. 2007

The description is as before (Addison et al., 2007) with the following addition. The genomic G+C content of the type strain is 64.0%. Its approximate genome size is 4.15 Mbp.

Emended Description of Novosphingobium pentaromativorans Sohn et al. 2004

The description is as before (Sohn et al., 2004) with the following modification. The genomic G+C content of the type strain is 63.0%. Its approximate genome size is 5.46 Mbp.

Emended Description of Novosphingobium subarcticum (Nohynek et al. 1996) Takeuchi et al. 2001

The description is as before (Takeuchi et al., 2001) with the following restriction. The genomic G+C content of the type strain is 65.1%. Its approximate genome size is 6.30 Mbp.

Emended Description of Novosphingobium subterraneum (Balkwill et al. 1997) Takeuchi et al. 2001

The description is as before (Takeuchi et al., 2001) with the following modification. The genomic G+C content of the type strain is 63.3%. Its approximate genome size is 4.70 Mbp.

Emended Description of Oceanibacterium hippocampi Balcázar et al. 2013

The description is as before (Balcázar et al., 2012) with the following modification. The genomic G+C content of the type strain is 66.7%. Its approximate genome size is 4.83 Mbp.

Emended Description of Oceanibaculum pacificum Dong et al. 2010

The description is as before (Dong et al., 2010) with the following modification. The genomic G+C content of the type strain is 65.7%. Its approximate genome size is 3.89 Mbp.

Emended Description of Oceanicella actignis Albuquerque et al. 2012

The description is as before (Albuquerque et al., 2012) with the following modification. The genomic G+C content of the type strain is 72.4%. Its approximate genome size is 3.28 Mbp.

Emended Description of Oceanicola granulosus Cho and Giovannoni 2004

The description is as before (Cho and Giovannoni, 2004) with the following modification. The genomic G+C content of the type strain is 70.4%. Its approximate genome size is 4.04 Mbp.

Emended Description of Octadecabacter antarcticus Gosink et al. 1998 emend. Billerbeck et al. 2015

The description is as before (Billerbeck et al., 2015) with the following modification. The genomic G+C content of the type strain is 54.6%. Its approximate genome size is 4.88 Mbp.

Emended Description of Octadecabacter arcticus Gosink et al. 1998 emend. Billerbeck et al. 2015

The description is as before (Billerbeck et al., 2015) with the following modification. The genomic G+C content of the type strain is 55.1%. Its approximate genome size is 5.48 Mbp.

Emended Description of Octadecabacter ascidiaceicola Kim et al. 2016

The description is as before (Kim et al., 2016a) with the following modification. The G+C content of the type-strain genome is 54.9%, its approximate size 3.23 Mbp.

Emended Description of Oharaeibacter diazotrophicus Lv et al. 2017

The description is as before (Lv et al., 2017) with the following modification. The G+C content of the type-strain genome is 71.6%, its approximate size 4.99 Mbp.

Emended Description of Orientia tsutsugamushi (Hayashi 1920) Tamura et al. 1995

The description is as before (Tamura et al., 1995) with the following restriction. The genomic G+C content of the type strain is 29.9%. Its approximate genome size is 1.45 Mbp.

Emended Description of Pacificibacter marinus (Jung et al. 2011) Park et al. 2015

The description is as before (Park et al., 2014c) with the following modification. The genomic G+C content of the type strain is 52.0%. Its approximate genome size is 3.87 Mbp.

Emended Description of Palleronia marisminoris Martínez-Checa et al. 2005 emend. Albuquerque et al. 2015

The description is as before (Albuquerque et al., 2015) with the following modification. The genomic G+C content of the type strain is 66.2%. Its approximate genome size is 3.94 Mbp.

Emended Description of Pannonibacter phragmitetus Borsodi et al. 2003

The description is as before (Borsodi et al., 2003) with the following modification. The genomic G+C content of the type strain is 63.1%. Its approximate genome size is 4.78 Mbp.

Emended Description of Paracoccus aestuarii Roh et al. 2009

The description is as before (Roh et al., 2009) with the following modification. The G+C content of the type-strain genome is 67.7%, its approximate size 3.74 Mbp.

Emended Description of Paracoccus alcaliphilus Urakami et al. 1989

The description is as before (Urakami et al., 1989) with the following restriction. The genomic G+C content of the type strain is 64.3%. Its approximate genome size is 4.61 Mbp.

Emended Description of Paracoccus alkenifer Lipski et al. 1998

The description is as before (Lipski et al., 1998) with the following addition. The genomic G+C content of the type strain is 67.3%. Its approximate genome size is 3.19 Mbp.

Emended Description of Paracoccus bengalensis Ghosh et al. 2006

The description is as before (Ghosh et al., 2006) with the following modification. The G+C content of the type-strain genome is 67.4%, its approximate size 4.99 Mbp.

Emended Description of Paracoccus contaminans Kämpfer et al. 2016

The description is as before (Kämpfer et al., 2016) with the following modification. The genomic G+C content of the type strain is 68.7%. Its approximate genome size is 3.03 Mbp.

Emended Description of Paracoccus denitrificans (Beijerinck and Minkman 1910) Davis 1969 (Approved Lists 1980) emend. Rainey et al., 1999

The description is as before (Rainey et al., 1999) with the following restriction. The genomic G+C content of the type strain is 66.8%. Its approximate genome size is 5.19 Mbp.

Emended Description of Paracoccus halophilus Liu et al. 2008

The description is as before (Liu et al., 2008) with the following modification. The genomic G+C content of the type strain is 65.2%. Its approximate genome size is 4.01 Mbp.

Emended Description of Paracoccus isoporae Chen et al. 2011

The description is as before (Chen et al., 2011a) with the following modification. The genomic G+C content of the type strain is 65.8%. Its approximate genome size is 3.52 Mbp.

Emended Description of Paracoccus pantotrophus (Robertson and Kuenen 1984) Rainey et al. 1999

The description is as before (Rainey et al., 1999) with the following modification. The G+C content of the type-strain genome is 67.6%, its approximate size 4.41 Mbp.

Emended Description of Paracoccus sediminis Pan et al. 2014

The description is as before (Pan et al., 2014) with the following modification. The genomic G+C content of the type strain is 66.0%. Its approximate genome size is 3.65 Mbp.

Emended Description of Paracoccus seriniphilus Pukall et al. 2003

The description is as before (Pukall et al., 2003) with the following modification. The genomic G+C content of the type strain is 61.6%. Its approximate genome size is 4.20 Mbp.

Emended Description of Paracoccus solventivorans Siller et al. 1996 emend. Lipski et al. 1998

The description is as before (Lipski et al., 1998) with the following restriction. The genomic G+C content of the type strain is 68.7%. Its approximate genome size is 3.38 Mbp.

Emended Description of Paracoccus sphaerophysae Deng et al. 2011

The description is as before (Deng et al., 2011) with the following modification. The genomic G+C content of the type strain is 69.1%. Its approximate genome size is 3.36 Mbp.

Emended Description of Paracoccus yeei Daneshvar et al. 2003

The description is as before (Daneshvar et al., 2003) with the following modification. The genomic G+C content of the type strain is 67.5%. Its approximate genome size is 4.43 Mbp.

Emended Description of Pararhizobium giardinii (Amarger et al. 1997) Mousavi et al. 2016

The description is as before (Mousavi et al., 2015) with the following addition. The genomic G+C content of the type strain is 60.7%. Its approximate genome size is 6.81 Mbp.

Emended Description of Pararhodospirillum photometricum (Molisch 1907) Lakshmi et al. 2014

The description is as before (Lakshmi et al., 2014) with the following addition. The genomic G+C content of the type strain is 64.7%. Its approximate genome size is 3.88 Mbp.

Emended Description of Parvibaculum lavamentivorans Schleheck et al. 2004

The description is as before (Schleheck et al., 2004) with the following modification. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 3.91 Mbp.

Emended Description of Pelagibaca bermudensis Cho and Giovannoni 2006

The description is as before (Cho and Giovannoni, 2006) with the following modification. The genomic G+C content of the type strain is 66.4%. Its approximate genome size is 5.43 Mbp.

Emended Description of Pelagibacterium halotolerans Xu et al. 2011

The description is as before (Xu et al., 2011) with the following modification. The genomic G+C content of the type strain is 61.4%. Its approximate genome size is 3.95 Mbp.

Emended Description of Pelagibacterium luteolum Xu et al. 2011

The description is as before (Xu et al., 2011) with the following modification. The genomic G+C content of the type strain is 60.4%. Its approximate genome size is 4.30 Mbp.

Emended Description of Pelagicola litoralis Kim et al. 2008

The description is as before (Kim Y.-G. et al., 2008) with the following modification. The genomic G+C content of the type strain is 54.8%. Its approximate genome size is 3.64 Mbp.

Emended Description of Phaeobacter gallaeciensis (Ruiz-Ponte et al. 1998) Martens et al. 2006

The description is as before (Martens et al., 2006) with the following modification. The genomic G+C content of the type strain is 59.4%. Its approximate genome size is 4.54 Mbp.

Emended Description of Phaeobacter italicus (Vandecandelaere et al. 2009) Wirth and Whitman 2018

The description is as before (Wirth and Whitman, 2018) with the following restriction. The G+C content of the type-strain genome is 60.0%, its approximate size 4.06 Mbp.

Emended Description of Phenylobacterium composti Weon et al. 2008

The description is as before (Weon et al., 2008a) with the following modification. The genomic G+C content of the type strain is 70.0%. Its approximate genome size is 3.30 Mbp.

Emended Description of Phenylobacterium deserti Khan et al. 2017

The description is as before (Khan et al., 2017) with the following modification. The G+C content of the type-strain genome is 68.2%, its approximate size 3.87 Mbp.

Emended Description of Phenylobacterium immobile Lingens et al. 1985

The description is as before (Lingens et al., 1985) with the following modification. The genomic G+C content of the type strain is 66.7%. Its approximate genome size is 3.33 Mbp.

Emended Description of Phenylobacterium kunshanense Chu et al. 2015

The description is as before (Chu et al., 2015) with the following modification. The G+C content of the type-strain genome is 69.3%, its approximate size 4.18 Mbp.

Emended Description of Phyllobacterium brassicacearum Mantelin et al. 2006

The description is as before (Mantelin et al., 2006) with the following modification. The G+C content of the type-strain genome is 57.3%, its approximate size 5.79 Mbp.

Emended Description of Phyllobacterium endophyticum Flores-Félix et al. 2013

The description is as before (Flores-Félix et al., 2013) with the following modification. The G+C content of the type-strain genome is 57.0%, its approximate size 5.51 Mbp.

Emended Description of Phyllobacterium leguminum Mantelin et al. 2006

The description is as before (Mantelin et al., 2006) with the following modification. The G+C content of the type-strain genome is 59.8%, its approximate size 3.72 Mbp.

Emended Description of Phyllobacterium rubiacearum (ex von Faber 1912) Knösel 1984

The description is as before (Knösel, 1984) with the following modification. The G+C content of the type-strain genome is 59.2%, its approximate size 5.58 Mbp.

Emended Description of Phyllobacterium salinisoli León-Barrios et al. 2018

The description is as before (León-Barrios et al., 2018) with the following modification. The G+C content of the type-strain genome is 59.9%, its approximate size 5.04 Mbp.

Emended Description of Phyllobacterium sophorae Jiao et al. 2015

The description is as before (Jiao et al., 2015) with the following modification. The G+C content of the type-strain genome is 57.0%, its approximate size 6.36 Mbp.

Emended Description of Pleomorphomonas diazotrophica Madhaiyan et al. 2013

The description is as before (Madhaiyan et al., 2013) with the following modification. The G+C content of the type-strain genome is 65.1%, its approximate size 4.53 Mbp.

Emended Description of Pleomorphomonas koreensis Im et al. 2006

The description is as before (Im et al., 2006) with the following modification. The genomic G+C content of the type strain is 67.2%. Its approximate genome size is 4.62 Mbp.

Emended Description of Pontibaca methylaminivorans Kim et al. 2010

The description is as before (Kim K.K. et al., 2010) with the following modification. The genomic G+C content of the type strain is 66.5%. Its approximate genome size is 2.65 Mbp.

Emended Description of Ponticaulis koreensis Kang and Lee 2009

The description is as before (Kang and Lee, 2009) with the following modification. The genomic G+C content of the type strain is 54.8%. Its approximate genome size is 3.40 Mbp.

Emended Description of Ponticoccus lacteus Yang et al. 2015

The description is as before (Yang Y. et al., 2015) with the following modification. The genomic G+C content of the type strain is 65.1%. Its approximate genome size is 5.42 Mbp.

Emended Description of Porphyrobacter cryptus da Costa et al. 2003

The description is as before (Rainey et al., 2003) with the following modification. The genomic G+C content of the type strain is 67.9%. Its approximate genome size is 2.95 Mbp.

Emended Description of Porphyrobacter neustonensis Fuerst et al. 1993

The description is as before (Fuerst et al., 1993) with the following modification. The genomic G+C content of the type strain is 65.3%. Its approximate genome size is 3.09 Mbp.

Emended Description of Primorskyibacter sedentarius Romanenko et al. 2011

The description is as before (Romanenko et al., 2011b) with the following restriction. The G+C content of the type-strain genome is 60.8%, its approximate size 5.06 Mbp.

Emended Description of Prosthecomicrobium hirschii Staley 1984

The description is as before (Staley, 1984) with the following restriction. The genomic G+C content of the type strain is 68.9%. Its approximate genome size is 6.46 Mbp.

Emended Description of Pseudaminobacter salicylatoxidans Kämpfer et al. 1999

The description is as before (Kämpfer et al., 1999) with the following modification. The G+C content of the type-strain genome is 62.7%, its approximate size 4.84 Mbp.

Emended Description of Pseudodonghicola xiamenensis (Tan et al. 2009) Hameed et al. 2014

The description is as before (Hameed et al., 2014) with the following modification. The genomic G+C content of the type strain is 63.6%. Its approximate genome size is 4.73 Mbp.

Emended Description of Pseudolabrys taiwanensis Kämpfer et al. 2006

The description is as before (Kämpfer et al., 2006) with the following modification. The G+C content of the type-strain genome is 64.4%, its approximate size 5.59 Mbp.

Emended Description of Pseudooceanicola antarcticus (Huo et al. 2014) Lai et al. 2015

The description is as before (Lai et al., 2015) with the following modification. The G+C content of the type-strain genome is 66.3%, its approximate size 4.23 Mbp.

Emended Description of Pseudooceanicola batsensis (Cho and Giovannoni 2004) Lai et al. 2015

The description is as before (Lai et al., 2015) with the following modification. The genomic G+C content of the type strain is 66.1%. Its approximate genome size is 4.44 Mbp.

Emended Description of Pseudooceanicola marinus (Lin et al. 2007) Lai et al. 2015

The description is as before (Lai et al., 2015) with the following modification. The genomic G+C content of the type strain is 66.8%. Its approximate genome size is 4.53 Mbp.

Emended Description of Pseudooceanicola nanhaiensis (Gu et al. 2007) Lai et al. 2015

The description is as before (Lai et al., 2015) with the following modification. The genomic G+C content of the type strain is 67.9%. Its approximate genome size is 4.66 Mbp.

Emended Description of Pseudooceanicola nitratireducens (Zheng et al. 2010) Lai et al. 2015

The description is as before (Lai et al., 2015) with the following modification. The genomic G+C content of the type strain is 64.2%. Its approximate genome size is 4.07 Mbp.

Emended Description of Pseudorhodobacter antarcticus Chen et al. 2013

The description is as before (Chen C.-X. et al., 2013) with the following modification. The genomic G+C content of the type strain is 59.1%. Its approximate genome size is 3.88 Mbp.

Emended Description of Pseudorhodobacter aquimaris Jung et al. 2012

The description is as before (Jung et al., 2012a) with the following modification. The genomic G+C content of the type strain is 58.7%. Its approximate genome size is 3.82 Mbp.

Emended Description of Pseudorhodobacter wandonensis Lee et al. 2013

The description is as before (Lee et al., 2013) with the following modification. The genomic G+C content of the type strain is 57.8%. Its approximate genome size is 3.89 Mbp.

Emended Description of Pseudoroseovarius halocynthiae (Kim et al. 2012) Sun et al. 2015

The description is as before (Sun et al., 2015) with the following modification. The genomic G+C content of the type strain is 57.1%. Its approximate genome size is 3.40 Mbp.

Emended Description of Pseudoroseovarius sediminilitoris (Park and Yoon 2013) Sun et al. 2015

The description is as before (Sun et al., 2015) with the following modification. The genomic G+C content of the type strain is 58.7%. Its approximate genome size is 3.41 Mbp.

Emended Description of Pseudoruegeria haliotis Hyun et al. 2013

The description is as before (Hyun et al., 2013b) with the following modification. The genomic G+C content of the type strain is 63.0%. Its approximate genome size is 5.04 Mbp.

Emended Description of Pseudoruegeria lutimaris Jung et al. 2010

The description is as before (Jung et al., 2010) with the following modification. The genomic G+C content of the type strain is 62.3%. Its approximate genome size is 5.81 Mbp.

Emended Description of Pseudoruegeria sabulilitoris Park et al. 2014

The description is as before (Park et al., 2014a) with the following modification. The genomic G+C content of the type strain is 62.4%. Its approximate genome size is 5.32 Mbp.

Emended Description of Puniceibacterium sediminis Zhang et al. 2015

The description is as before (Zhang D.-C. et al., 2015) with the following modification. The genomic G+C content of the type strain is 60.8%. Its approximate genome size is 4.64 Mbp.

Emended Description of Reyranella massiliensis Pagnier et al. 2011

The description is as before (Pagnier et al., 2011) with the following addition. The genomic G+C content of the type strain is 64.9%. Its approximate genome size is 5.77 Mbp.

Emended Description of Rhizobium altiplani Baraúna et al. 2016

The description is as before (Baraúna et al., 2016) with the following modification. The G+C content of the type-strain genome is 59.5%, its approximate size 8.10 Mbp.

Emended Description of Rhizobium esperanzae Cordeiro et al. 2017

The description is as before (Cordeiro et al., 2017) with the following modification. The G+C content of the type-strain genome is 61.1%, its approximate size 6.29 Mbp.

Emended Description of Rhizobium etli Segovia et al. 1993

The description is as before (Segovia et al., 1993) with the following addition. The genomic G+C content of the type strain is 61.0%. Its approximate genome size is 6.53 Mbp.

Emended Description of Rhizobium flavum Gu et al. 2014

The description is as before (Gu et al., 2014) with the following modification. The genomic G+C content of the type strain is 61.6%. Its approximate genome size is 4.66 Mbp.

Emended Description of Rhizobium hainanense Chen et al. 1997

The description is as before (Chen et al., 1997) with the following modification. The genomic G+C content of the type strain is 59.6%. Its approximate genome size is 7.25 Mbp.

Emended Description of Rhizobium jaguaris Rincón-Rosales et al. 2013

The description is as before (Rincón-Rosales et al., 2013) with the following addition. The G+C content of the type-strain genome is 59.4%, its approximate size 8.03 Mbp.

Emended Description of Rhizobium leguminosarum (Frank 1879) Frank 1889 (Approved Lists 1980) emend. Ramírez-Bahena et al. 2008

The description is as before (Ramírez-Bahena et al., 2008) with the following modification. The genomic G+C content of the type strain is 60.6%. Its approximate genome size is 7.81 Mbp.

Emended Description of Rhizobium leucaenae Ribeiro et al. 2012

The description is as before (Ribeiro et al., 2012) with the following modification. The genomic G+C content of the type strain is 59.4%. Its approximate genome size is 6.68 Mbp.

Emended Description of Rhizobium lusitanum Valverde et al. 2006

The description is as before (Valverde et al., 2006) with the following modification. The genomic G+C content of the type strain is 59.6%. Its approximate genome size is 7.92 Mbp.

Emended Description of Rhizobium miluonense Gu et al. 2008

The description is as before (Gu et al., 2008) with the following modification. The genomic G+C content of the type strain is 59.7%. Its approximate genome size is 6.81 Mbp.

Emended Description of Rhizobium mongolense van Berkum et al. 1998

The description is as before (van Berkum et al., 1998) with the following addition. The genomic G+C content of the type strain is 59.5%. Its approximate genome size is 7.17 Mbp.

Emended Description of Rhizobium multihospitium Han et al. 2008

The description is as before (Han et al., 2008) with the following modification. The genomic G+C content of the type strain is 59.8%. Its approximate genome size is 7.32 Mbp.

Emended Description of Rhizobium rhizogenes (Riker et al. 1930) Young et al. 2001

The description is as before (Young et al., 2001) with the following restriction. The genomic G+C content of the type strain is 59.9%. Its approximate genome size is 7.04 Mbp.

Emended Description of Rhizobium selenitireducens Hunter et al. 2008

The description is as before (Hunter et al., 2007) with the following modification. The genomic G+C content of the type strain is 63.5%. Its approximate genome size is 4.97 Mbp.

Emended Description of Rhizobium tropici Martínez-Romero et al. 1991

The description is as before (Martínez-Romero et al., 1991) with the following addition. The genomic G+C content of the type strain is 59.5%. Its approximate genome size is 6.69 Mbp.

Emended Description of Rhodobacter aestuarii Venkata Ramana et al. 2009

The description is as before (Venkata Ramana et al., 2009) with the following modification. The genomic G+C content of the type strain is 61.1%. Its approximate genome size is 3.84 Mbp.

Emended Description of Rhodobacter blasticus (Eckersley and Dow 1981) Kawasaki et al. 1994

The description is as before (Kawasaki et al., 1993) with the following modification. The G+C content of the type-strain genome is 66.5%, its approximate size 3.59 Mbp.

Emended Description of Rhodobacter capsulatus (Molisch 1907) Imhoff et al. 1984

The description is as before (Imhoff et al., 1984) with the following addition. The genomic G+C content of the type strain is 66.5%. Its approximate genome size is 3.67 Mbp.

Emended Description of Rhodobacter maris Venkata Ramana et al. 2008

The description is as before (Venkata Ramana et al., 2008) with the following modification. The G+C content of the type-strain genome is 65.1%, its approximate size 3.83 Mbp.

Emended Description of Rhodobacter megalophilus Arunasri et al. 2008

The description is as before (Arunasri et al., 2008) with the following modification. The genomic G+C content of the type strain is 68.8%. Its approximate genome size is 4.86 Mbp.

Emended Description of Rhodobacter viridis Shalem Raj et al. 2013

The description is as before (Shalem Raj et al., 2013) with the following modification. The G+C content of the type-strain genome is 65.4%, its approximate size 3.86 Mbp.

Emended Description of Rhodoblastus acidophilus (Pfennig 1969) Imhoff 2001

The description is as before (Imhoff, 2001) with the following restriction. The genomic G+C content of the type strain is 65.2%. Its approximate genome size is 4.71 Mbp.

Emended Description of Rhodoblastus sphagnicola Kulichevskaya et al. 2006

The description is as before (Kulichevskaya et al., 2006) with the following modification. The G+C content of the type-strain genome is 63.9%, its approximate size 5.09 Mbp.

Emended Description of Rhodomicrobium udaipurense Ramana et al. 2013

The description is as before (Ramana V.V. et al., 2013) with the following modification. The genomic G+C content of the type strain is 62.5%. Its approximate genome size is 3.63 Mbp.

Emended Description of Rhodomicrobium vannielii Duchow and Douglas 1949 (Approved Lists 1980)

The description is as before (Duchow and Douglas, 1949) with the following addition. The genomic G+C content of the type strain is 62.2%. Its approximate genome size is 4.01 Mbp.

Emended Description of Rhodopseudomonas pentothenatexigens Kumar et al. 2013

The description is as before (Kumar et al., 2013) with the following modification. The G+C content of the type-strain genome is 66.0%, its approximate size 5.38 Mbp.

Emended Description of Rhodopseudomonas pseudopalustris Venkata Ramana et al. 2012

The description is as before (Venkata Ramana et al., 2012) with the following modification. The genomic G+C content of the type strain is 64.6%. Its approximate genome size is 5.27 Mbp.

Emended Description of Rhodospira trueperi Pfennig et al. 1998

The description is as before (Pfennig et al., 1997) with the following modification. The genomic G+C content of the type strain is 67.0%. Its approximate genome size is 4.21 Mbp.

Emended Description of Rhodospirillum rubrum (Esmarch 1887) Molisch 1907 (Approved Lists 1980)

The description is as before (Molisch, 1907) with the following addition. The genomic G+C content of the type strain is 65.4%. Its approximate genome size is 4.41 Mbp.

Emended Description of Rhodothalassium salexigens (Drews 1982) Imhoff et al. 1998

The description is as before (Imhoff et al., 1998) with the following restriction. The G+C content of the type-strain genome is 68.6%, its approximate size 3.35 Mbp.

Emended Description of Rhodovibrio salinarum (Nissen and Dundas 1985) Imhoff et al. 1998

The description is as before (Imhoff et al., 1998) with the following modification. The genomic G+C content of the type strain is 66.0%. Its approximate genome size is 4.18 Mbp.

Emended Description of Rhodovulum euryhalinum (Kompantseva 1989) Hiraishi and Ueda 1994

The description is as before (Hiraishi and Ueda, 1994a) with the following modification. The G+C content of the type-strain genome is 67.9%, its approximate size 3.78 Mbp.

Emended Description of Rhodovulum imhoffii Srinivas et al. 2007

The description is as before (Srinivas et al., 2007a) with the following modification. The genomic G+C content of the type strain is 63.1%. Its approximate genome size is 2.91 Mbp.

Emended Description of Rhodovulum kholense Anil Kumar et al. 2008

The description is as before (Anil Kumar et al., 2008) with the following modification. The genomic G+C content of the type strain is 67.8%. Its approximate genome size is 4.46 Mbp.

Emended Description of Rhodovulum marinum Srinivas et al. 2006

The description is as before (Srinivas et al., 2006) with the following modification. The G+C content of the type-strain genome is 67.2%, its approximate size 3.94 Mbp.

Emended Description of Rhodovulum viride Srinivas et al. 2014

The description is as before (Srinivas et al., 2014) with the following modification. The G+C content of the type-strain genome is 67.7%, its approximate size 4.52 Mbp.

Emended Description of Rickettsia australis Philip 1950 (Approved Lists 1980)

The description is as before (Philip, 1950) with the following addition. The genomic G+C content of the type strain is 32.3%. Its approximate genome size is 1.32 Mbp.

Emended Description of Rickettsia bellii Philip et al. 1983

The description is as before (Philip et al., 1983) with the following modification. The genomic G+C content of the type strain is 31.6%. Its approximate genome size is 1.52 Mbp.

Emended Description of Rickettsia conorii Brumpt 1932 (Approved Lists 1980)

The description is as before (Brumpt, 1932) with the following addition. The genomic G+C content of the type strain is 32.4%. Its approximate genome size is 1.27 Mbp.

Emended Description of Rickettsia heilongjiangensis Fournier et al. 2006

The description is as before (Fournier et al., 2003) with the following addition. The G+C content of the type-strain genome is 32.3%, its approximate size 1.28 Mbp.

Emended Description of Rickettsia honei Stenos et al. 1998

The description is as before (Stenos et al., 1998) with the following addition. The genomic G+C content of the type strain is 32.4%. Its approximate genome size is 1.27 Mbp.

Emended Description of Rickettsia hoogstraalii Duh et al. 2010

The description is as before (Duh et al., 2010) with the following addition. The genomic G+C content of the type strain is 32.4%. Its approximate genome size is 1.48 Mbp.

Emended Description of Rickettsia japonica Uchida et al. 1992

The description is as before (Uchida et al., 1992) with the following restriction. The genomic G+C content of the type strain is 32.4%. Its approximate genome size is 1.28 Mbp.

Emended Description of Rickettsia prowazekii Rocha-Lima 1916 (Approved Lists 1980)

The description is as before (da Rocha-Lima, 1916) with the following addition. The genomic G+C content of the type strain is 29.0%. Its approximate genome size is 1.11 Mbp.

Emended Description of Rickettsia raoultii Mediannikov et al. 2008

The description is as before (Mediannikov et al., 2008) with the following modification. The G+C content of the type-strain genome is 32.6%, its approximate size 1.48 Mbp.

Emended Description of Rickettsia sibirica Zdrodovskii 1948 (Approved Lists 1980)

The description is as before (Zdrodovskii, 1948) with the following addition. The genomic G+C content of the type strain is 32.5%. Its approximate genome size is 1.25 Mbp.

Emended Description of Rickettsia slovaca Sekeyová et al. 1998

The description is as before (Sekeyová et al., 1998) with the following addition. The genomic G+C content of the type strain is 32.5%. Its approximate genome size is 1.28 Mbp.

Emended Description of Rickettsia tamurae Fournier et al. 2006

The description is as before (Fournier et al., 2006) with the following addition. The genomic G+C content of the type strain is 32.5%. Its approximate genome size is 1.45 Mbp.

Emended Description of Rickettsia typhi (Wolbach and Todd 1920) Philip 1943 (Approved Lists 1980)

The description is as before (Philip, 1943) with the following addition. The genomic G+C content of the type strain is 28.9%. Its approximate genome size is 1.11 Mbp.

Emended Description of Robiginitomaculum antarcticum Lee et al. 2007

The description is as before (Lee et al., 2007d) with the following modification. The genomic G+C content of the type strain is 52.5%. Its approximate genome size is 2.77 Mbp.

Emended Description of Roseiarcus fermentans Kulichevskaya et al. 2014

The description is as before (Kulichevskaya et al., 2014) with the following modification. The G+C content of the type-strain genome is 68.2%, its approximate size 6.79 Mbp.

Emended Description of Roseibacterium elongatum Suzuki et al. 2006

The description is as before (Suzuki et al., 2006) with the following modification. The genomic G+C content of the type strain is 65.7%. Its approximate genome size is 3.56 Mbp.

Emended Description of Roseibium denhamense Suzuki et al. 2000 emend. Biebl et al. 2007

The description is as before (Biebl et al., 2007) with the following restriction. The G+C content of the type-strain genome is 57.3%, its approximate size 4.85 Mbp.

Emended Description of Roseibium hamelinense Suzuki et al. 2000 emend. Biebl et al. 2007

The description is as before (Biebl et al., 2007) with the following restriction. The G+C content of the type-strain genome is 56.4%, its approximate size 4.72 Mbp.

Emended Description of Roseinatronobacter thiooxidans Sorokin et al. 2000

The description is as before (Sorokin et al., 2000) with the following modification. The genomic G+C content of the type strain is 60.1%. Its approximate genome size is 3.68 Mbp.

Emended Description of Roseivivax halodurans Suzuki et al. 1999

The description is as before (Suzuki et al., 1999a) with the following modification. The genomic G+C content of the type strain is 66.3%. Its approximate genome size is 4.49 Mbp.

Emended Description of Roseivivax halotolerans Suzuki et al. 1999

The description is as before (Suzuki et al., 1999a) with the following modification. The genomic G+C content of the type strain is 64.1%. Its approximate genome size is 3.77 Mbp.

Emended Description of Roseivivax isoporae Chen et al. 2012

The description is as before (Chen M.-H. et al., 2012) with the following restriction. The genomic G+C content of the type strain is 69.8%. Its approximate genome size is 4.90 Mbp.

Emended Description of Roseivivax jejudonensis Jung et al. 2015

The description is as before (Jung et al., 2014a) with the following modification. The genomic G+C content of the type strain is 68.0%. Its approximate genome size is 4.31 Mbp.

Emended Description of Roseivivax lentus Park et al. 2010

The description is as before (Park et al., 2010) with the following modification. The genomic G+C content of the type strain is 66.1%. Its approximate genome size is 4.44 Mbp.

Emended Description of Roseobacter denitrificans Shiba 1991

The description is as before (Shiba, 1991) with the following modification. The genomic G+C content of the type strain is 58.9%. Its approximate genome size is 4.32 Mbp.

Emended Description of Roseobacter litoralis Shiba 1991

The description is as before (Shiba, 1991) with the following restriction. The genomic G+C content of the type strain is 57.2%. Its approximate genome size is 4.75 Mbp.

Emended Description of Roseomonas aerilata Yoo et al. 2008

The description is as before (Yoo et al., 2008) with the following modification. The genomic G+C content of the type strain is 69.7%. Its approximate genome size is 6.43 Mbp.

Emended Description of Roseomonas cervicalis Rihs et al. 1998 emend. Sánchez-Porro et al. 2009

The description is as before (Sánchez-Porro et al., 2009) with the following modification. The genomic G+C content of the type strain is 72.0%. Its approximate genome size is 4.44 Mbp.

Emended Description of Roseomonas lacus Jiang et al. 2006 emend. Sánchez-Porro et al. 2009

The description is as before (Sánchez-Porro et al., 2009) with the following modification. The genomic G+C content of the type strain is 68.7%. Its approximate genome size is 6.39 Mbp.

Emended Description of Roseomonas mucosa Han et al. 2003 emend. Sánchez-Porro et al. 2009

The description is as before (Sánchez-Porro et al., 2009) with the following addition. The genomic G+C content of the type strain is 70.4%. Its approximate genome size is 4.86 Mbp.

Emended Description of Roseomonas rosea (Kämpfer et al. 2003) Sánchez-Porro et al. 2009

The description is as before (Sánchez-Porro et al., 2009) with the following modification. The genomic G+C content of the type strain is 70.8%. Its approximate genome size is 5.34 Mbp.

Emended Description of Roseomonas stagni Furuhata et al. 2008

The description is as before (Furuhata et al., 2008) with the following modification. The genomic G+C content of the type strain is 70.7%. Its approximate genome size is 6.38 Mbp.

Emended Description of Roseomonas vinacea Zhang et al. 2008

The description is as before (Zhang Y.-Q. et al., 2008) with the following modification. The genomic G+C content of the type strain is 70.3%. Its approximate genome size is 6.35 Mbp.

Emended Description of Roseospirillum parvum Glaeser and Overmann 2001

The description is as before (Glaeser and Overmann, 1999) with the following modification. The genomic G+C content of the type strain is 69.9%. Its approximate genome size is 3.50 Mbp.

Emended Description of Roseovarius azorensis Rajasabapathy et al. 2014

The description is as before (Rajasabapathy et al., 2014) with the following modification. The genomic G+C content of the type strain is 63.1%. Its approximate genome size is 3.86 Mbp.

Emended Description of Roseovarius confluentis Jia et al. 2017

The description is as before (Jia et al., 2017) with the following modification. The G+C content of the type-strain genome is 63.2%, its approximate size 4.46 Mbp.

Emended Description of Roseovarius halotolerans Oh et al. 2009

The description is as before (Oh et al., 2009) with the following modification. The genomic G+C content of the type strain is 63.8%. Its approximate genome size is 3.73 Mbp.

Emended Description of Roseovarius lutimaris Choi et al. 2013

The description is as before (Choi et al., 2013) with the following modification. The genomic G+C content of the type strain is 59.9%. Its approximate genome size is 4.28 Mbp.

Emended Description of Roseovarius marisflavi Li et al. 2014

The description is as before (Li Z. et al., 2013) with the following modification. The genomic G+C content of the type strain is 60.3%. Its approximate genome size is 4.08 Mbp.

Emended Description of Roseovarius nanhaiticus Wang et al. 2010

The description is as before (Wang et al., 2010) with the following modification. The genomic G+C content of the type strain is 63.3%. Its approximate genome size is 3.70 Mbp.

Emended Description of Roseovarius nubinhibens González et al. 2003

The description is as before (González et al., 2003) with the following modification. The genomic G+C content of the type strain is 63.9%. Its approximate genome size is 3.67 Mbp.

Emended Description of Roseovarius tolerans Labrenz et al. 1999

The description is as before (Labrenz et al., 1999) with the following restriction. The genomic G+C content of the type strain is 63.1%. Its approximate genome size is 3.77 Mbp.

Emended Description of Rubrimonas cliftonensis Suzuki et al. 1999

The description is as before (Suzuki et al., 1999b) with the following modification. The genomic G+C content of the type strain is 71.6%. Its approximate genome size is 4.86 Mbp.

Emended Description of Rubritepida flocculans Alarico et al. 2002

The description is as before (Alarico et al., 2002) with the following modification. The genomic G+C content of the type strain is 73.4%. Its approximate genome size is 3.84 Mbp.

Emended Description of Ruegeria atlantica (Rüger and Höfle 1992) Uchino et al. 1999 emend. Vandecandelaere et al. 2008

The description is as before (Vandecandelaere et al., 2008a) with the following restriction. The genomic G+C content of the type strain is 56.4%. Its approximate genome size is 4.82 Mbp.

Emended Description of Ruegeria faecimaris Oh et al. 2011

The description is as before (Oh et al., 2011) with the following modification. The genomic G+C content of the type strain is 56.7%. Its approximate genome size is 4.12 Mbp.

Emended Description of Ruegeria halocynthiae Kim et al. 2012

The description is as before (Kim et al., 2012b) with the following modification. The genomic G+C content of the type strain is 56.5%. Its approximate genome size is 4.24 Mbp.

Emended Description of Ruegeria intermedia Kämpfer et al. 2013

The description is as before (Kämpfer et al., 2013a) with the following addition. The genomic G+C content of the type strain is 64.0%. Its approximate genome size is 3.86 Mbp.

Emended Description of Ruegeria lacuscaerulensis (Petursdottir and Kristjansson 1999) Yi et al. 2007 emend. Vandecandelaere et al. 2008

The description is as before (Vandecandelaere et al., 2008a) with the following modification. The genomic G+C content of the type strain is 63.0%. Its approximate genome size is 3.52 Mbp.

Emended Description of Ruegeria pomeroyi (González et al. 2003) Yi et al. 2007 emend. Vandecandelaere et al. 2008

The description is as before (Vandecandelaere et al., 2008a) with the following modification. The genomic G+C content of the type strain is 64.1%. Its approximate genome size is 4.60 Mbp.

Emended Description of Saccharibacter floricola Jojima et al. 2004

The description is as before (Jojima et al., 2004) with the following modification. The genomic G+C content of the type strain is 51.2%. Its approximate genome size is 2.38 Mbp.

Emended Description of Salipiger aestuarii (Park et al. 2011) Wirth and Whitman 2018