Revision of the paraphyletic genus Koerneria Meyl, 1960 and resurrection of two other genera of Diplogastridae (Nematoda)

Abstract Recent inferences of phylogeny from molecular characters, as well as a reexamination of morphological and biological characters, reject the monophyly of the nematode genus Koerneria Meyl, 1960 (Diplogastridae). Here, Koerneria sensu lato is revised. The genus, which previously consisted of 40 species, is separated into three genera. Almost all of the transferred species are moved to the resurrected genus Allodiplogaster Paramonov & Sobolev in Skrjabin et al. (1954). Koerneria and Allodiplogaster are distinguished from each other by a weakly vs. clearly striated body surface, an undivided vs. divided stomatal cheilostom, and arrangement of the terminal ventral triplet of male genital papillae, namely in that v5 and v6 are paired and separated from v7 vs. v5–v7 being close to each other. Allodiplogaster is further divided into two groups of species, herein called the henrichae and striata groups, based on both morphological and life-history traits. The henrichae group is characterized by papilliform labial sensilla and male genital papillae, a conical tail in both males and females, and an association with terrestrial habitats and insects, whereas the striata group is characterized by setiform labial sensilla and male genital papillae, an elongated conical tail in both sexes, and an association with aquatic habitats. A second genus, Anchidiplogaster Paramonov, 1952, is resurrected to include a single species that is characterized by its miniscule stoma and teeth, unreflexed testis, and a distinct lack of male genital papillae or stomatal apodemes. Lastly, one further species that was previously included in Koerneria sensu lato is transferred to the genus Pristionchus Kreis, 1932. The revision of Koerneria sensu lato is necessitated by the great variability in its subordinate taxa, which occupy a variety of habitats, in addition to the increased attention to Diplogastridae as a model system for comparative mechanistic biology.


Revision of the paraphyletic genus Koerneria
and resurrection of two other genera of Diplogastridae (Nematoda) Introduction Koerneria Meyl, 1960 heretofore consisted of 40 nominal species, following the revision by Sudhaus and Fürst von Lieven (2003) and including species described since then (Suppl. material 1). Several unidentified or undescribed species with molecular vouchers have also been reported (Mayer et al. 2007). Biological characters are variable for the genus, which contains terrestrial species isolated from rich soil environments, associates of several different groups of insects, and aquatic species. Such a diversity of ecologies raises the question of whether distinct subgroups of the genus can be identified and corroborated by independent characters. In their revision of Diplogastridae, Sudhaus and Fürst von Lieven (2003) circumscribed Koerneria by the following characters: (1) presence of stomatal dimorphism, (2) cheilostom often separated into six per-and interradial plates or small stick-like plates (= rugae), (3) vertical striation, (4) stegostom with a dorsal claw-like tooth, a right subventral tooth, and left subventral serrated plates, (5) postdental region of stegostom with two subventrad directed apodemes, (6) female gonad amphidelphic or seldom prodelphic, and (7) no bursa. Later, Fürst von Lieven (2008) transferred K. colobocerca (Andrássy, 1964) Fürst von Lieven, 2008 from Mononchoides Rahm, 1928 and revised the generic definition by adding two characters, (8) intestine sometimes with a prerectum, and (9) tail filiform or conical. Given the breadth of the most recent morphological definition of the genus, the only absolute generic character is the presence of articulated apodemes in the stoma. However, phylogenetic studies of molecular characters strongly indicate that the genus, as currently interpreted, is paraphyletic and thus separable into two or more clades (Fig. 1). Therefore, a taxonomic reorganization of Koerneria sensu lato (= Koerneria sensu Fürst von Lieven 2008) is needed. Such a revision is of particular importance for ongoing studies of the natural history of the genus and its relatives (e.g., Giblin-Davis et al. 2006, Kanzaki et al. 2009) as well as comparative research more generally in Diplogastridae (e.g., Mayer et al. 2009, Ragsdale et al. 2013, as Koerneria sensu lato includes taxa that are sister groups to most other known species in the family. In this article, we revise the genus Koerneria by examining the original and subsequent descriptions of its nominal species. Based on morphological, biological, and molecular evidence, we separate Koerneria sensu lato into three genera. All renamed groups are hypothesized to be monophyletic and follow the precedent of previous classification systems (Andrássy 1984). Besides limiting the scope of Koerneria, we resurrect the genera Anchidiplogaster Paramonov, 1952 andAllodiplogaster Paramonov &Sobolev in Skrjabin et al. (1954). Furthermore, we distinguish two putative clades of Allodiplogaster species, each of which we hypothesize to be monophyletic on available information, although we leave formal revision of this genus to follow molecular studies of unsampled aquatic taxa.

Materials and methods
Species of Koerneria sensu lato are classified herein into four typological groups (three genera, with one genus further separated into two morphological and ecological groups) based on the following characters or traits, which were selected due to their high availability and reliability, being relatively accurate even in old descriptions: (1) Stomatal morphology, specifically the separation of cheilostom and presence of apodemes (2) Male tail morphology, including shape and arrangement of genital papillae  Kanzaki et al. (2014b), which was inferred from nearly full-length small subunit ribosomal DNA sequences. A subsequent study that included several more species of Koerneria sensu lato likewise showed the well-supported exclusion (99% bootstrap support in likelihood analysis, 100% posterior probability in Bayesian analysis) of a clade of K. luziae + K. ruehmi from all other Diplogastridae, including a monophyletic clade of what is designated herein as Allodiplogaster (Atighi et al. 2013). Asterisks indicate nodes with very strong support as inferred in the former study (>95% bootstrap support, 100% posterior probability).
(3) Female tail morphology (4) Life-history characters, particularly habitat preferences In addition to published literature, several species available in culture were examined for typological characters: Koerneria luziae, isolated from stag beetles from Japan (Kanzaki et al. 2011); Koerneria sp., isolated from Dorcus rectus (Coleoptera: Lucanidae) from Japan; Allodiplogaster spp. RGD227 and RGD228, both isolated from soil-dwelling bees in the United States. Stomatal morphology and tail characters were examined by differential interference contrast (DIC) microscopy according to methods described by Kanzaki (2013). Several schematic illustrations were prepared based on original observations as well as published data. Following the reexamination of informative characters, the original descriptions and revisions were reviewed to determine the generic status of species following the International Code of Zoological Nomenclature (ICZN).

Systematics
Previous inferences of the phylogeny of Koerneria spp. indicate that two species, K. luziae (Körner, 1954) Meyl, 1960and K. ruehmi Atighi, Pourjam, Kanzaki, Giblin-Davis, De Ley, Mundo-Ocampo & Pedram, 2013, form a well-supported clade that is the sister group to most or all other sequenced Diplogastridae (Atighi et al. 2013;Kanzaki et al. 2014b; Fig. 1). Separate from this group is a well-supported clade consisting of K. lucani (Körner, 1954) Meyl, 1960 and some unidentified or undescribed species (Fig. 1). The former two species share two features that clearly distinguish them from other species of Koerneria sensu lato, namely (1) a tube-or ring-like (undivided) cheilostom and (2) male genital papillae arranged such that v5 and v6 (papillae nomenclature follows Sudhaus and Fürst von Lieven 2003) are close to each other and v7 is clearly apart from v6. These two characters diagnose six nominal species of Koerneria sensu lato, including the above two as well as the type species of the genus. We therefore revise the genus as follows:

Generic diagnosis
1) Stomatal dimorphism occasionally present 1 2) Body-wall cuticle with weak vertical striations 3) Cheilostom usually forming short, undivided tube; rugae absent 4) Stegostom with dorsal claw-like tooth, right subventral tooth, and left subventral serrated plates or ridges 1 Dimorphism has been confirmed in at least two species in this group, including one undescribed species (Kanzaki and Ragsdale et al. unpubl (Körner, 1954) Meyl, 1960 K. systenoceri (Körner, 1954) Meyl, 1960 Following this restricted definition of Koerneria, most of the remaining species of Koerneria sensu lato are transferred to the resurrected genera Anchidiplogaster and Allodiplogaster. One species, which was previously combined as K. dubia (Hnatewytsch, 1929) Sudhaus & Fürst von Lieven, 2003 is returned to Anchidiplogaster based on a suite of characters unique to this taxon as well as by the lack of stomatal apodemes, an absolute character of Koerneria as defined both previously and herein: Anchidiplogaster Paramonov, 1952 Generic diagnosis 1) Miniscule, undivided stoma with two small, similarly sized pyramidal teeth (one dorsal and one right subventral) 2) Stomatal apodemes absent 3) Male genital papillae absent 4) Testis without flexure Type and only species =Diplogaster dubia Hnatewytsch, 1929 comb. Anchidiplogaster Paramonov, 1952 All but one of the remaining species of Koerneria sensu lato are transferred to the other resurrected genus, Allodiplogaster. This name has priority (ICZN 23.1) over other names that are available for this taxonomic grouping, which consist of Diplenteron Andrássy, 1964, Glauxinemella Gagarin, 1998, and Gobindonema Khera, 1970. Furthermore, Al-lodiplogaster is separated into two putatively monophyletic groups of species, which we designate as the "henrichae group" and "striata group" based on morphological and biological evidence. Fig. 2 =Diplenteron Andrássy, 1964: D. colobocercus Andrássy, 1964=Gobindonema Khera, 1970: G. filicaudata Khera, 1970nec Gobindonema Sood & Prashad, 1974 =Glauxinemella Gagarin, 1998: G. striata Gagarin, 1998 Generic diagnosis 1) Stomatal dimorphism occasionally present 2) Body-wall cuticle with clear vertical striations 3) Cheilostom separated into six per-and interradial plates or rugae 4) Stegostom with dorsal claw-like tooth, right subventral tooth, and left subventral serrated plates or ridges 5) Postdental region of stegostom with left and right apodemes directed subventrad 6) Female gonad amphidelphic; rarely prodelphic 7) Distal triplet papillae of males (v5-7) close to each other 8) Tail of male and females highly variable in shape 9) Known from variable habitats including terrestrial insect associates and aquatic species

Discussion
The separation of Koerneria as circumscribed herein from Allodiplogaster, Anchidiplogaster, and Pristionchus is strongly supported by the structure of the cheilostom, arrangement of male genital papillae, and phylogeny as inferred independently from molecular sequence characters (Atighi et al. 2013;Kanzaki et al. 2011Kanzaki et al. , 2013Kanzaki et al. , 2014b. Although for the previous, wider definition of Koerneria the name Koerneria was itself a junior synonym of Allodiplogaster, the name Koerneria is retained for a group of six described species, all of which are unambiguously unified by morphological characters and which are represented by a clade not nested within any other valid genus. Furthermore, species of Koerneria in the revised sense are apparently the sister group to most or all species of Diplogastridae, with the possible one exception of Leptojacobus dorci (see Kanzaki et al. 2014b), and therefore the revision of this genus will be useful for ongoing research on the family as a comparative model system. The distinct morphology of Anchidiplogaster dubia, especially its lack of genital papillae in several observed male specimens separates this species from all other Diplogastridae and supports its reestablishment in that genus. Further indicating the distinctness of this species from most other known Diplogastridae is its unreflexed testis flexure, as a flexure was previously considered a plesiomorphic character of the entire family Fürst von Lieven 2003, Andrássy 2005), although one other diplogastrid genus, Leptojacobus Kanzaki, Ragsdale, Susoy & Sommer, 2014, has since been described to have no flexure. In the original description of A. dubia, the above missing features were explicitly given as diagnostic characters (Hnatewytsch 1929) and thus were unlikely to be simply missed in all specimens examined. An undivided stoma is also unusual among most diplogastrids with subventral teeth, although this type of stoma is present in Koerneria in the revised sense and is thus insufficient in itself to diagnose Anchidiplogaster. The presence of asymmetrical teeth and a glandular postcorpus clearly support its identity as a diplogastrid, the otherwise divergent stomatal morphology in this species, namely its diminutive, undivided stomatal cavity armed with triangular teeth, obscure its relationships to other taxa within the family. Because molecular characters are not available for this species, its position in Diplogastridae cannot be predicted with any certainty.
The split of Allodiplogaster into the henrichae and striata groups is also supported by morphology, principally by the tail of both sexes, which is usually much longer in striata group than in henrichae group, and by the male genital papillae and labial sensilla, which are distinctly setiform in striata group. The separation of the two proposed groups of Allodiplogaster is only confounded by overlapping morphological characters, namely the fish-bone-like swellings along the pharyngeal lumen that are present in henrichae group species, Allodiplogaster spp. RGD227 and RGD228, as well as in one species of the striata group, A. carinata, although not in another striata group species such as A. pararmata (Fürst von Lieven 2001, Fürst von Lieven andSudhaus 2000, Kanzaki et al. unpubl. obs.). However, no species of the striata group have been molecularly characterized thus far, so the phylogenetic position of that group has yet to be tested by gene sequence data. Reisolation of species, particularly of the striata group, will be necessary for fully testing the system of intrageneric grouping presented here. Complete taxonomic revision of Allodiplogaster is therefore not performed here due to a lack of molecular evidence, although further studies may confirm the separation of the two species groups designated here into different genera.
Other morphological characters may further support to separate henrichae group from as a monophyletic clade distinct from the striata group and from Koerneria. Kühne (1995) examined three henrichae group species by scanning electron microscopy and reported important morphological features of the male tail: the modification in v5 and v6 subventral distal papillae, whereby v5 is rooted in a socket-like base and has split tip, and v6 having two anterior and posterior appendages and being likewise rooted in a socket-like base (Fig. 2). This morphology, which is similar to the trifurcate genital papillae of Diplogasteroides spp. (Kiontke et al. 2001, is otherwise unique among Diplogastridae and Rhabditidae thus far characterized (e.g., Sudhaus 2000, Kanzaki et al. 2012a), and has additionally been found in two undescribed henrichae group species (Giblin-Davis et al. unpubl. obs.). The diagnostic utility of modified genital papillae has already been shown for another clade of Diplogastridae: in Pristionchus and the closely related genus Parapristionchus Kanzaki, Ragsdale, Herrmann, Mayer, Tanaka & Sommer, 2012, the v5 and v6 papillae are shrouded at the base by a socket-like structure, where the tip of v6 is split into two papilla-like projections (e.g., Kanzaki et al. 2012bKanzaki et al. , 2014a. Therefore, further examination of male tail morphology in other species of Koerneria sensu lato may confirm the importance of v5 and v6 as diagnostic of the henrichae group of Allodiplogaster. Yet another putative character for the henrichae group is the presence of a prerectum, as suggested to distinguish the "Diplenteron group" of Koerneria sensu lato (Fürst von Lieven 2008) and which was reported for A. colobocerca and A. sudhausi. The prerectum, which is a shallow but distinct constriction separating the posterior part of the intestine from the anterior part, can be found in three undescribed Allodiplogaster species (Kanzaki and Giblin-Davis unpubl. obs.), lending preliminary support to this idea. This character is not present in Koerneria nor in any species of striata group as interpreted from original descriptions, and therefore it may additionally support monophyly of the henrichae group.