Hybosorinae

Hybosorinae View in CoL – Clade 4: Kuijtenous + Hybosorus

Two distinct genera comprised this group: Kuijtenous ( Madagascar) and Hybosorus (including Hybosorus alluaudi ) (Afrotropical) ( Fig. 1). The group has an Afrotropical and Oriental distribution, except for Hybosorus illigeri Reiche, 1853 , considered a cosmopolitan invasive species ( Kuijten, 1983; Ocampo, 2002).

Inourresults, Hybosorus isrecoveredasmonophyletic only with the inclusion of Hybosoroides . In a revision of Hybosorus, Kuijten (1983) remarked on the strong similarity between these genera. In that research study, the only differences presented to distinguish those two genera were the density of punctures and the presence of a smooth, median swelling on the pronotum of Hybosoroides . Kuitjen stated that he was inclined to include Hybosoroides in Hybosorus , although as a subgenus. In view of the topologies presented here, the proposal of a subgenus would not be possible, but they support the synonymization of Hybosoroides with Hybosorus .

Hybosorinae – Clade 5: Brenskea +

(( Hypseloderus + Seleucosorus ) +

( Frolovius + Coilodes )) + (( Celaenochrous + Phaeochroops ) + ( Phaeochridius + Phaeochrous ))

This clade is composed of nine genera: two Palaearctic ( Brenskea and Seleucosorus ), two Neotropical ( Coilodes and Frolovius ), four Oriental ( Celaenochrous , Hypseloderus , Phaeochroops and Phaeochridius ) and the most diverse among Hybosorinae genera ( Phaeochrous ), which is widely distributed in the Old World and in the Australian region ( Fig. 1).

Brenskea View in CoL (now included in Hybosorinae View in CoL ) is the sister-group of all the other genera of this branch. Nikolajev (2010) suggested that Brenskea View in CoL belongs to Dynamopodinae View in CoL ( Scarabaeidae View in CoL ), based on the clypeus covering the mandible. However, the present hypothesis disagrees with Nikolajev (2010), since Brenskea View in CoL has not been recovered close to Orubesa View in CoL . In addition, the topology of the clade Brenskea View in CoL + Pachyplectrus View in CoL has not been recovered, as proposed by Ocampo (2006b) and by Bai et al. (2015).

Three of the five genera used in the molecular analysis by Ocampo & Hawks (2006) are grouped in this clade ( Coilodes View in CoL , Phaeochroops View in CoL and Phaeochrous View in CoL ). In the present hypothesis, Phaeochroops View in CoL is closer to Phaeochrous View in CoL than to Coilodes View in CoL . This is different from the hypothesis of Ocampo & Hawks (2006), in which Phaeochroops View in CoL appeared as the sister-group of Coilodes View in CoL and Phaeochrous View in CoL as the sister-group of all the other Hybosorinae View in CoL tested, including Metachaetodus View in CoL and Hybosorus View in CoL .

Among the lineages recovered in this clade, one is formed by Hypseloderus + Seleucosorus ( Fig. 1). Arrow (1909) included Hypseloderus , hitherto monospecific, in Troginae ( Trogidae ), while Schmidt (1913) relocated it into Hybosorinae . In the description of Hypseloderus yupae Masumoto et al., 2013 , the authors stated that the asymmetric genitalia and cupuliform antennal club would be enough to include the species within Hybosorinae ( Masumoto et al., 2013) . The present phylogenetic hypothesis (the first to include this genus) supports this statement.

The monophyly of Coilodes (with two species included) agrees with the hypothesis proposed by Bai et al. (2015). Frolovius , the most recently described Hybosorinae genus is placed as the sister group of Coilodes .

Another branch clusters two Oriental genera, Celaenochrous + Phaeochroops . Kuijten (1984) considered Celaenochrous as close to Phaeochrous . However, the angled shape of the mandible and some other homoplasies positioned this genus closer to Phaeochroops than to Phaeochrous ( Fig. 2).

The relationship of Phaeochridius + Phaeochrous in the same clade is retrieved for the first time, since Phaeochridius has not been tested in any of the previous analyses proposed for Hybosoridae or its subfamilies ( Grebennikov et al., 2004; Ocampo, 2006b; Ocampo & Hawks, 2006; Ballerio & Grebennikov, 2016).

OUTGROUPS: ORUBESA

The position of Orubesa in Scarabaeoidea is controversial because this genus is usually placed in Dynamopodinae ( Scarabaeidae ). Some authors have included it in Hybosoridae ( Allsopp, 1984; Nikolajev, 1999), but none of them tested this hypothesis in a phylogenetic analysis. Only Ballerio & Grebennikov (2016) included Orubesa athleta (Semenov, 1895) in the analysis of Ceratocanthinae . Similar to our results ( Figs 1, 2), Ballerio & Grebennikov (2016) recovered Orubesa as more related to Ochodaeidae than to Hybosoridae or Scarabaeidae (this family is represented in both studies by Aphodiini ). The difference in Ballerio & Grebennikov’s (2016) hypothesis is in the position of Orubesa as the sister-group of the clade formed by Glaphyridae + Ochodaeidae . Thus, Orubesa does not belong to Hybosoridae , but its position in Scarabaeoidea remains questionable.

The relationships of Thinorycter (another Dynamopodinae genus) are also inconsistent with its position within Scarabaeoidea . Nikolajev (1999) suggested that Thinorycter belongs to Hybosoridae , similarly to Orubesa . We obtained only one specimen of Thinorycter ; its inclusion in the final analysis was not possible, due to several missing morphological structures necessary to code in the character matrix. None of the preliminary analyses placed Thinorycter in Hybosoridae . Recently, Li et al. (2019) has proposed a hypothesis based on morphological data, in which Thinorycter appears as the sister-group of Glaphyridae (represented by Eulasia , Glaphyrus and Pygopleurus ) and distant from Orubesa , suggesting that this genus does not belong to Scarabaeidae , as it has been classified, or even to Hybosoridae , as suggested by Nikolajev (1999).

HYBOSORIDAE SISTER-GROUP

Glaphyridae View in CoL + ( Ochodaeidae View in CoL + Orubesa View in CoL ) were recovered as the sister-group of Hybosoridae View in CoL ( Figs 1, 2), agreeing with the hypothesis based on the DNA sequence of eight nuclear genes proposed by McKenna et al. (2015). Despite this result, the relationships among Hybosoridae View in CoL , Glaphyridae View in CoL and Ochodaeidae View in CoL are still uncertain. Several phylogenetic analyses, all of them with morphological and molecular characters, have yielded divergent results ( Browne & Scholtz, 1999; Ocampo & Hawks, 2006; Smith et al., 2006; Lawrence et al., 2011; Zhang et al., 2018; McKenna, 2019).

In the phylogenetic hypothesis based on 134 morphological characters from adults and larvae proposed by Browne & Scholtz (1999), Hybosoridae View in CoL appear as the sister-group of Ochodaeidae View in CoL . The hypothesis proposed by Browne & Scholtz (1999) was supported in the molecular analysis proposed for Hybosoridae View in CoL by Ocampo & Hawks (2006). According to the hypothesis proposed by Smith et al. (2006), based on molecular data, these three Scarabaeoidea families are polytomic.

In more comprehensive analyses, the sister-group of Hybosoridae is even more uncertain, as shown in the morphological hypothesis proposed by Lawrence et al. (2011) for Coleoptera , based on 516 characters from adults and larvae, in which Hybosoridae are the sister-group of the clade Glaresidae + ( Glaphyridae + Melolonthidae ). In addition, in the analysis based on molecular data proposed by Zhang et al. (2018), Hybosoridae appear as the sister-group of Scarabaeidae , and both form a clade with Glaphyridae .