The superfamily Tenebrionoidea is one of the most challenging clades in the beetle tree-of-life owing to its vast species richness and complex taxonomic history. Within this group, the family Melandryidae has long been overlooked and its systematics remains poorly known. Using available sequence data, we infer the most comprehensive phylogeny of Melandryidae to date. Our results support the polyphyly of Melandryidae with three independent clades spread across Tenebrionoidea. To accommodate these results, we restrict the status of Melandryidae and resurrect the family Osphyidae stat. rev. The third clade corresponding to the tribe Serropalpini pro parte is placed as incertae sedis within Tenebrionoidea pending further investigation and additional taxon sampling to resolve its phylogenetic placement.

Beetle evolution, false darkling beetles, Melandryinae, Osphyinae, Serropalpini, Tenebrionoidea systematics

The superfamily Tenebrionoidea has a worldwide distribution, encompasses 28 families and comprises over 33,000 species in more than 3,000 genera, approximately two thirds of which belong to the family Tenebrionidae (Barclay and Bouchard 2023). Tenebrionoidea exhibits a wide array of ecologies with saproxylic, saprophagous, mycetophagous, predatory and ectoparasitic taxa. Members of Tenebrionoidea are characterized by their tarsomere formula: 5-5-4, rarely 3-3-3 or 3-4-4 in males but never 5-5-5. In molecular analyses, the monophyly of Tenebrionoidea is not supported by all studies. They are sometimes inferred as monophyletic (Hunt et al. 2007; Gunter et al. 2014; Cai et al. 2022), but in most recent phylogenomic treatments are recovered as paraphyletic due to the placement of Lymexyloidea as a derived lineage in this group (McKenna et al. 2015, 2019; Batelka et al. 2016). Despite recent developments in the phylogenomics of beetles, higher-level phylogenetic relationships in the Tenebrionoidea remain uncertain (Gunter et al. 2014). Studies focusing on intra-Tenebrionoidea relationships are scarce and have often resulted in poorly supported backbones preventing a better understanding of the group evolution. The most comprehensive molecular phylogeny of Tenebrionoidea was inferred by Kergoat et al. (2014). Of the 18 families of Tenebrionoidea included in this study, five were found to be paraphyletic or polyphyletic. Some cases such as the Rhipiphoridae are well-documented, these beetles being known to form a paraphyletic family due to the placement of the monophyletic Mordellidae (Kergoat et al. 2014; Batelka et al. 2016).

Another interesting case within Tenebrionoidea is the family Melandryidae. Previously named Serropalpidae, it is represented by more than 420 species classified in ca. 60 genera (Nikitsky and Pollock 2010). Most of them are placed in Melandryinae, one of the two recognized subfamilies of Melandryidae, itself divided into eight tribes: Anisoxiellini, Dircaeini, Hypulini, Melandryini, Orchesiini, Serropalpini, Xylitini, and Zilorini. Melandryidae have a great variety of morphologies (Fig. 1) and a wide range of sizes, from 1.1 mm to 22 mm (Nikitsky and Pollock 2010). They are vaguely characterized by the last maxillary palpomere which is usually triangular and securiform or hook-shaped, markedly large. Their antennae have 11 antennomeres (excepted in Conopalpus, which has 10 antennomeres), filiform or serrate, sometimes dilated apically. The abdomen consists of five visible ventrites (exceptionally six ventrites) where at least the first two are fused. The tarsal formula is 5-5-4 in both sexes. Basal metatarsomeres are elongate (Lawrence 1982; Pollock 2002). Because of their cryptic habits, the ecology of many species remains unknown, especially outside of the Holarctic region. They are mostly referred to as mycophagous or saproxylic beetles linked with dead or decaying wood or associated with fungi on wood or in the soil litter (Hammond and Lawrence 1989; Nikitsky and Pollock 2010; Konvička 2016a; Evans 2021).

Figure 1. 

Morphological diversity of Melandryidae. A Dircaea australis B Dolotarsus lividus C Marolia alicantina D Melandrya caraboides E Serropalpus barbatus F Osphya aeneipennis. Scale bars: 5 mm. (Pictures: A, F Vít Kabourek B–E Aleš Sedláček)

The English common name of the family ‘false darkling beetles’ illustrates well the definition of the family as tenebrionid-like but not tenebrionid beetles. Indeed, it has always included various genera difficult to place and has always ‘lost’ some genera or subfamilies due to revisions pointing to the incoherence of their placement. The delimitation and definition of the family have markedly changed through time with the distinction of multiple families (Stenotrachelidae, Synchroidae, Tetratomidae) and with the placement of some species/genera in other families (e.g. Pseudeucinetus in Limnichidae, Sphalma in Pythidae) (Nikitsky and Pollock 2010).

Crowson (1966) followed by Lawrence (1982, 1991) considered the Melandryidae to be divided into three subfamilies: Eustrophinae, Melandryinae and Osphyinae, based on the combination of larval and imaginal features. Later, Nikitsky (1998) transferred the Eustrophinae to Tetratomidae. The subfamilial division within Melandryidae has not changed since then. Currently two subfamilies of Melandryidae are recognized: Melandryinae and Osphyinae (Nikitsky and Pollock 2010). These two taxa have fundamentally different imaginal morphologies: Melandryinae have simple claws and a head not notably narrowed behind the eyes, while Osphyinae have claws with distinct basal teeth and a distinctly narrowed head behind the eyes (Pollock 2002; Nikitsky and Pollock 2010; Crowson 1966). Crowson (1966) himself questioned the reliability to keep Osphyinae within the Melandryidae. Recently, Melandryidae have repeatedly been shown to be polyphyletic (Kergoat et al. 2014; Liu et al. 2023), highlighting the fact that some taxa are still erroneously included in this family. A renewed interest in the taxonomy and systematics of Melandryidae prompts a better understanding of high-level phylogenetic relationships and an updated classification of the different constituent groups (Nikitsky and Saitȏ 2014; Konvička 2016b; Recalde Irurzun et al. 2017; Konvička and Brustel 2021; Choi et al. 2020; Cosandey 2023a, b, c, 2024). Based on an exhaustive compilation of available Melandryidae sequence data and of a selection of neighboring clades, we investigated the phylogeny of Melandryidae with a special attention on the relatedness of Melandryinae and Osphyinae.

The best maximum likelihood tree inferred in IQ-TREE is shown in Figure 2 (Available as a supplementary file in the Zenodo repository). Branch support across the phylogeny is moderate in particular for the deeper nodes of the tree; however, the monophyly of most families is inferred with varying levels of branch support. The only three families that are not recovered as monophyletic are the Melandryidae, Ripiphoridae and Tetratomidae. The inferred phylogeny recovers Melandryidae as polyphyletic and divided in three clades: 1) Osphyinae (clade I), 2) Mikadonius and Enchodes (clade II), and 3) Melandryinae without Mikadonius and Enchodes (clade III), Clade I is recovered as sister to Scraptiidae with moderate branch support. Despite low branch support for the relationships between families in this part of the tree, the placement of Clade I in a larger clade comprising Mycteridae, Salpingidae, Scraptiidae and Trictenomidae is recovered with moderate branch support (SH-aLRT = 92, UFBoot = 30). Within the robustly supported Clade I (SH-aLRT = 100, UFBoot = 100), we recover Osphya as paraphyletic due to the placement of Conopalpus testaceus as sister to a clade comprising O. orientalis and O. sinensis with moderate branch support (SH-aLRT = 80, UFBoot = 64). Clade II is inferred as sister to Aderidae, Synchroidae and most of Tetratomidae with moderate branch support (SH-aLRT = 83, UFBoot = 34). We infer that clade III is sister to Tetratoma, the placement of which results in Tetratomidae being polyphyletic with low branch support. Clade II and III are placed with moderate branch support (SH-aLRT = 80, UFBoot = 44) in a larger clade comprising Aderidae and Tetratomidae. Within Clade III, the relationships between major clades are poorly supported. The genera Melandrya and Phloiotrya are recovered as paraphyletic with moderate to strong branch support due to the respective placements of Phryganophilus ruficollis as sister to Melandrya dubia and M. pictipennis, and of the genus Abdera as sister to Phloiotrya planiuscula. We infer a moderately to strongly supported sister relationship between Dircaea and a monophyletic Microtonus, between a monophyletic Hypulus and (Phloiotrya + Abdera + Anisoxya), and between Microscapha and a monophyletic Orchesia.

Figure 2. 

Best maximum likelihood tree inferred in IQ-TREE for Melandryidae including closely related families. The polyphyletic Melandryidae s.l. are shown in green. Habitus of selected species are indicated with colored triangles. (Pictures: Udo Schmidt)

This study presents the first phylogenetic work specifically focusing on Melandryidae. Our results are largely in adequation with those of studies focusing on beetle phylogenomics (McKenna et al. 2015; Cai et al. 2022) or Tenebrionoidea in particular (Gunter et al. 2014; Kergoat et al. 2014); all of them inferred Melandryidae s.l. as a polyphyletic family with Osphyidae and Melandryidae s.str. forming distinct clades. Crowson (1966) already doubted the placement of Osphyidae inside of the Melandryidae. Our results demonstrate clearly that Osphyidae forms a family on its own. A third clade of Melandryidae s.l. was present. It contained Mikadonius and Enchodes, two Serropalpini genera already pointed out to display unusual morphological features (Nikitsky and Pollock 2010). Moreover, Enchodes crepusculus was previously placed in the genus Synchroa (now placed in Synchroidae) before being transferred to Mikadonius by Crowson (1966), illustrating once more the ambiguity of the relatedness of this species (and genus). The placement of these two genera is still unclear and they should be considered as Tenebrionoidea incertae sedis since they apparently do not belong to Melandryidae but they cannot be placed in another family. One possibility is that the tribe Serropalpini forms a distinct family Serropalpidae that needs to be resurrected. However, further analyses including more Serropalpini taxa are needed to confirm or reject this hypothesis. Taxa previously considered as part of Melandryidae – Synchroidae and Tetratomidae – were found as sister clades to Melandryidae with Tetratomidae being polyphyletic. Stenotrachelidae, also previously placed in Melandryidae was recovered as a basal lineage of Tenebrionoidea as in McKenna et al. (2015). This family grouped with Lymexylidae, a taxon considered as part of the Tenebrionoidea or forming its own superfamily depending on the authors. Further studies focusing on Tenebrionoidea should try to elucidate the placement of Stenotrachelidae and its relationships with Lymexylidae. The present work is a further step in the comprehension of the systematics of Tenebrionoidea, a clade known for its complex taxonomy and also in the global knowledge of evolutionary relatedness of beetles. This refinement of the Melandryidae classification will be useful in the framework of ongoing taxonomic and phylogenetic efforts (Nikitsky and Saitȏ 2014; Konvička 2016b; Recalde Irurzun et al. 2017; Konvička and Brustel 2021; Choi et al. 2020; Cosandey 2023a, b, c, 2024).

We would like to thank Gaël Kergoat for his advices in an early stage in this work. We acknowledge Vít Kabourek, Aleš Sedláček and Udo Schimdt for letting us use their pictures. We thank the reviewers and the editorial board for their comments that helped improve this article. Finally, we thank all the people contributing to genetic databases, thus permitting studies like ours.