Termitotrox venus sp. n. is described from Cambodia and represents the second discovery of Termitotrox Reichensperger, 1915 from the Indo-Chinese subregion of the Indomalayan region. Most of the type series was collected from refuse dumps in fungus garden cells of Macrotermes cf. gilvus (Hagen, 1858). Macrotermes Holmgren, 1910 was previously an unknown host of Termitotrox species. The new species is easily distinguished from all known congeners by having wing-shaped trichomes on the elytra and the larger body size, at 2.5 mm in length. We also describe the mouthparts to complement the description of the genus Termitotrox.

Termitophily, Termitotroginae, Termitotrogini, Isoptera, Termitidae, Macrotermitinae, Macrotermes, new species, Indo-Chinese subregion, mouthparts

Members of the genus Termitotrox Reichensperger, 1915 are blind, flightless termitophilous scarabs associated with the fungus-growing termite genera Protermes Holmgren, 1910, Odontotermes Holmgren, 1912 or Hypotermes Holmgren, 1917 (Isoptera, Termitidae, Macrotermitinae). The genus was previously known from the Ethiopian region (eight species) and the Indian subregion (two species) of the Indomalayan region (Krikken 2008), until the discovery of Termitotrox cupido Maruyama, 2012 from Cambodia, representing the first species of Termitotrox from the Indo-Chinese subregion of the Indomalayan region (Maruyama 2012a). Recently, we collected another undescribed species of Termitotrox in Cambodia from fungus garden cells of Macrotermes Holmgren, 1910 (also Macrotermitinae) – a previously unknown host of Termitotrox. This is the second discovery of the genus in the Indo-Chinese subregion of the Indomalayan region. In this paper the new species is described and biological information about it is provided.

In August 2014, we examined fungus gardens of the termite genera Macrotermes, Microtermes Wasmann, 1902, Odontotermes and Hypotermes in Siem Reap, Cambodia. After examining more than 300 fungus gardens, we found 8 Termitotrox beetles from fungus garden cells of Macrotermes cf. gilvus (Hagen, 1858) and Hypotermes makhamensis Ahmad, 1965, seven specimens in seven cells of three colonies of M. cf. gilvus and one specimen from one cell of H. makhamensis. The beetles were put in a killing tube (35 ml) with tissue paper and ethyl acetate; a day later they were removed from the tube and kept in 80% ethanol. All specimens were dried and mounted for morphological observation. Dissected genitalia and mouthparts were mounted in Euparal on a small glass plate (10×5 mm), and subsequently glued onto a paper card (6×5 mm) and pinned under the respective specimen (Maruyama 2004). A permanent mount of mouthparts was also made. Specimen photographs were taken using a Canon EOS 60D with a Canon MP-E 65 mm 1–5× macro lens and Kenko extension tubes and stacked using CombineZP software. Images of living beetle were taken using a Canon EOS 7D with a EF 100mm F2.8L Macro lens and Kenko extension tubes. Terminology of the species description follows Krikken (2008). All measurements in the paper are given in millimeters as follows: minimum length – maximum length (mean ± SD). The type series is deposited in Maruyama collection in the Kyushu University Museum, Fukuoka, Japan.

Termite association. Of the eight Termitotrox venus beetles recovered from fungus garden cells of Macrotermes cf. gilvus (Figs 11–13), seven were found on or inside the refuse dumps. The refuse dumps do not contain any fungal carpophores. The beetles appear camouflaged inside the refuse and move slowly, so they are difficult to collect. Only one specimen was found on the wall of fungus garden cell of Hypotermes makhamensis (Figs 11–13) despite more than 200 fungus garden cells of this termite were examined; this is probably accidental (it may be caused by the underground connection of the colonies of the two termite species ). Therefore, we think that the true host of T. venus is M. cf. gilvus.

All other known termitotrogines are associated with either Protermes, Odontotermes or Hypotermes (Krikken 2008; Maruyama 2012a). A phylogenetic analysis of fungus-growing termites revealed that these three genera form a monophyletic group, with Odontotermes being paraphyletic with respect to Hypotermes; however, Macrotermes did not group with this clade, and is instead only distantly related (Aanen and Eggleton 2005). In contrast, Termitotrox venus and T. cupido have a clear synapomorphy in the wing-shaped trichomes on the elytra, so these species are apparently closely related to each other. Therefore, the host relationship between species of Termitotrox and genera of Macrotermitinae is unlikely to have arisen via co-cladogenesis. This type of relationship between termite hosts and termitotrogine scarabs is similar to that observed in Corythoderini. Corythoderines are also known to be associated with Odontotermes and Macrotermes (Tangelder and Krikken 1982; Bordat and Moretto 2010; Maruyama 2012b). This capacity to utilize phylogenetically unrelated hosts suggests that perhaps both Macrotermes and the group formed by Protermes+Odontotermes+Hypotermes produce similar nest odors, which are targeted by termitophilous scarabs in search of host colonies.

The pronotal basomedian section and the elytral median projection of Termitotrox venus form a structure (Figs 2, 4) similar to that seen in Eocorythoderus incredibilis Maruyama, 2012 (and, to a lesser extent in Termitotrox cupido), which was also found in a Macrotermes cf. gilvus nest in Siem Reap. Maruyama (2012b) revealed that this structure functions as a handle that allows the termite to grip the beetle and carry it. We did not observe Termitotrox venus being carried by worker termites during our survey, but this structure is probably used for the same behavior. In addition, the number of damaged specimens (broken legs, tibia or tarsi) of this species was lower than that of E. incredibilis (damaged/undamaged: T. venus 2/8, E. incredibilis 5/10, based on the type series from Maruyama (2012b)), hence, perhaps, T. venus could be mostly a synoekete (ignored by the hosts) except during certain periods, such as the movement of the host colony.

Using Wasmannian terminology (Wasmann 1894), Vårdal and Forshage (2010) suggested Termitotrogini may be synechthrans (treated with hostility by the hosts) because of the defensive morphology of the species known at that time. However, at least T. venus and T. cupido seem to be mainly synoeketes because both species appeared to be ignored by termites in the field. Based on both morphology and field observations, Corythoderini were proposed to be symphilic (Vårdal and Forshage 2010). Although symphilic behavior was recorded for E. incredibilis (Maruyama 2012b), the rate of specimen damage is nevertheless high, and this species has, in overall, a more defensive morphology compared to the other species of Corythoderini. Hence, the biology of E. incredibilis may vary from persecuted (synechthran) to integrated (symphile). On the other hand, T. venus may be a largely ignored (synoekete) but based on the morphology similar to that of E. incredibilis (trichomes, carrying “handle”), may at times exhibit symphilic behavior. The discoveries of these new combinations of lifestyles in termitophilous beetles require a more flexible framework than that proposed by Wasmann (1894).

Size difference. Termitotrox venus is larger than T. cupido, and the beetle size seems to be correlated with the body size of the primary host of each of these species. Therefore, inquiline size may be affected by host size — a relationship paralleling that seen between termitophilous Staphylinidae and their hosts (Maruyama, personal observation).

We thank Dr. Joseph Parker (Columbia University), Dr. Andrey Frolov (Universidade Federal de Mato Grosso), Mr. Alberto Ballerio (Brescia) and anonymous reviewers for critically reading the manuscript. SK thanks Dr. Taisuke Kanao and Mr. Shûhei Yamamoto (Kyushu University) for information about their field survey in Cambodia. This study was supported in part by a financial incentive of Kyushu University Foundation Yamakawa Prize for SK.