Cyclopocentrus kombuglu Villemant & Rousse, 2016

Villemant, Claire, Jingxian, Liu & Rousse, Pascal, 2016, Deep into the head of a remarkable new genus of Orthocentrinae (Hymenoptera: Ichneumonidae) from the highest peak of Papua New Guinea, Mémoires du Muséum national d'Histoire naturelle 209, pp. 375-392 : 382-390

Cyclopocentrus kombuglu Villemant & Rousse , n. sp.

Figures 1-5

TYPE MATERIAL — Holotype ♂ (MNHN) Papua-New-Guinea. Province Madang, Mt Wilhelm 3700 m (-5.785644, 145.0580), 21-22/10/2012, leg. Gewa, Damag, Novotny, Leponce, understorey, Malaise - MAL-MW3700C-06/16-d06, P3765-09187, DNA extracted, sample n°61, P. Rousse 2014 ”. Paratypes. 1 ♂ (MNHN) same locality, same leg., same trap, 16-17/10/2012, MAL-MW3700C-01/16-d01, P3760-09173, DNA n°62; 3♀ (MNHN), same locality, same leg., same trap:

1♀ 17-18/10/2012, MAL-MW3700C-01/16-d01, P3761-09977, DNA n°59; 1♀ 21-22/10/2012, MAL-MW3700C-06/16-

d06, P3765-09187, DNA n°60; 1♀ 27-28/10/2012, MAL-MW3700C-12/16-d12, P3771-09203, DNA n°58.

DISTRIBUTION — Papua New Guinea, Mt Wilhelm (high altitude primary forest).

ETYMOLOGY — The species name refers to Mt Wilhelm, whose name in Kuman Papuan language is Enduwa Kombuglu.

DESCRIPTION — MALE: Forewing length: 3.1 mm (holotype); 3.3 mm (paratype). Head. Antenna with 22 flagellomeres, all longer than wide and uniformly setose; first, penultimate and apical flagellomeres about 6x, 2x and 3x longer than wide respectively; scape about twice longer than apically wide, face 1.3x wider than high, finely strigose, with long and thin silver setae, facial depression about 0.1 mm at its greatest diameter (1/3 of face width) and about 15 ųm in depth; clypeus short, 5x wider than high; malar space 2x longer than mandibular basal width; frons smooth, 2x wider than high; vertex and gena smooth with sparse fine setae; ocellar triangle equilateral, ocelli oval shaped, interocellar distance 1x and oculo-ocellar distance 2x the greatest diameter of lateral ocellus; gena rounded posteriorly, about 0.4x eye width in dorsal view; occiput entirely smooth, occipital carina lacking, hypostomal carina distinct. Mesosoma. Pronotum and pleura almost smooth, pleura ventrally densely setose; mesoscutum finely punctate and evenly setose, except a pair of small posterior lateral bare patches; cuticle smooth between punctures; scutellum finely setose without lateral carinae; propodeum dorsally smooth and laterally alutaceous, with sparse long and fine silver setae; propodeal carinae weak; area petiolaris shorter than wide, anteriorly rounded; distance between propodeal spiracle and pleural carina shorter than spiracle diameter. Metasoma. First and second tergites parallel sided, 2.5x and 1.5x longer than wide, respectively; thyridium subcircular, 0.2x width of tergite 2 in diameter; third tergite widening toward apex, 1.2x longer than apical width; first and second tergites alutaceous-aciculate, third tergite smooth except basal 1/5 alutaceous, following tergites smooth; all tergites with very sparse, long and fine silver setae; gonosquama subtriangular, elongate. Color. Body dark brown with antennae and legs fading to yellowish; coxae 1-2 and trochanters 1-3 yellow; facial depression bright yellow, sharply contrasting with remainder of face.

FEMALE: Forewing length: 3.2-3.3 mm (three paratypes); propodeal carinae stronger than in male; ovipositor gradually tapering toward apex; ovipositor sheath 0.7x as long as hind tibia, its basal 3/5 stalked; remainder of sexual dimorphism as detailed in generic description.

RESULTS OF PHYLOGENETIC ANALYSIS

The final tree of the combined 28S + CO1 bayesian analysis is displayed on Figure 7. The monophyly of the subfamily Orthocentrinae (A) and Orthocentrus genus - group (B) are strongly supported, with 100% pp in both cases. The Orthocentrus genus - group is only partially resolved, though a robust clade (C, 88% pp) encompassing all genera but Orthocentrus and affiliated species is clearly differentiated. This C clade is further subdivided into two robust D and E clades (92 and 96% pp, respectively), the former clustering the reference Stenomacrus and Neurateles species whereas all PNG species are clustered apart in the latter E clade. This E clade is in turn subdivided into three major clades: one (F, 98% pp) with all the Plectiscus species (including the two references from Canada), a second (G, 84% pp) with some of our “nr Plectiscus ” species (undetermined genera, morphologically close to Plectiscus ), and the third (H, 89% pp) with the remaining “nr Plectiscus ” species and Cyclopocentrus n. gen.

RESULTS OF IMAGE ANALYSES

We chiefly investigated the facial depression of males. Optical pictures show a yellow cuticular surface, most probably membranous, overlaid by the darker sclerotized external cuticle of face ( Figure 2a). The STEM images depict more precisely the facial depression and the invagination of its lateral walls ( Figure 1c). They also show the crown of setae surrounding the depression ( Figure 1b). In female, the evenly setose face shows only a shallow glabrous central concavity at its center ( Figure 4b).

The CT scan images provide details about the internal structure of the male depression ( Figures 2 a-d). The mediosagittal slice and the parasagittal slices made at lower margin, center and upper margin of the concavity show that the inner side of the depression is coated with a thick cell layer (CC) whose density is lower than that of the remainder of the integument. Such a layer is not visible anywhere else on the inner side of the head exoskeleton. Moreover, the depression appears to be in contact with the digestive track (DT) which is first dilated and oriented forward closely to the face, before being constricted and abruptly bent posteriorly from the upper margin of depression. The dilatation of the digestive track is not due to any food absorption because it appears in black on the images and is therefore empty.

DISCUSSION

WHY DESCRIBING A NEW ORTHOCENTRINE GENUS?

Orthocentrine wasps are noticeably morphologically variable, leading previous authors to split them into different subfamilies (e.g. Microleptinae sensu Townes (1971) or Helictinae Gupta (1987)) . Referring to larval characters, Wahl (1990) proposed later to include most of Microleptinae sensu Townes (1971) in the Orthocentrinae . This was then confirmed by molecular analysis ( Quicke et al. 2009). The Orthocentrinae sensu Townes (1971) are now referred to as the Orthocentrus genus-group (Wahl 1990, Wahl & Gauld 1998). Morphologically, the division between the Orthocentrus genus-group and the remainder of the subfamily matches respectively the fusion or not of the clypeus and the face. It is clearly apparent in our tree where the Orthocentrus genus-group appears as a derived group within Orthocentrinae ( Figure 7).

As pointed out by Broad (2010), apart from Plectiscus

the most speciose genera of the Orthocentrus genus-group (i.e. Orthocentrus , Stenomacrus , Neurateles ) are clearly not monophyletic. Moreover, their definitions were mostly based on north temperate species. These genera failed to represent the extremely high tropical diversity of the group ( Veijalainen et al. 2012), most of its tropical species being thus unsatisfactorily linked to one of the existing genera. This also fully agrees with our molecular results where only Plectiscus is recovered as a monophyletic clade while many others are provisionally stated as “nr Plectiscus ” or “nr Orthocentrus ”.

Therefore, a conservative approach is favored when defining new taxa within that group. However, we chose to describe Cyclopocentrus n. gen. as a distinct genus firstly because it shows an original set of morphological apomorphies. Apart from the significant exception of the facial depression, however, none of these apomorphies is unique in Cyclopocentrus n. gen. within the Orthocentrus group: Neurateles and Plectiscus also lack the epicnemial carina ( Wahl & Gauld 1998; Quicke et al. 2009), some Stenomacrus also exhibit apical bristles on hypopygium (G. Broad, pers.com.), the flexible basal stalk on ovipositor sheath occurs in several species of Orthocentrus , Neurateles and Stenomacrus ( Broad 2010) , and the absence of subapical notch FIGURE 6 of ovipositor may either be considered as plesiomorphic Plectiscus sp. Sensillar areas present on ventral face of median flagellar

segments of a female from Mt Wilhelm. Scale bar 30 µm. or as a secondary reversal because of the variability of this character within Orthocentrinae ( Wahl & Gauld 1998; Quicke et al. 2009).

We are fully aware that pure morphological arguments are of weak relevance to define a new genus in such a phylogenetically obscure group. Our second argument is thus based on the molecular analysis. Cyclopocentrus n. gen. is depicted there as fully distinct from all the currently described genera. It is included in a separate clade (H) with PNG species of undetermined genera. The closest described genus is Plectiscus whose clade (F) is distinct, strongly supported and validated by non-PNG references. Moreover Cyclopocentrus n. gen. may definitively not be confounded with Plectiscus because of the ovipositor length and structure, the wing venation and the sculpture of metasoma (see comments in the genus description).

INTERNAL STRUCTURE OF THE FACIAL CONCAVITY

The noticeable sexual dimorphism observed for this character strongly suggests a role in sexual communication. The invaginate shape of the male facial depression, the distinct nature and coloration of its ground, and the surrounding setae made us think to a possible sensory organ. The contrasting yellow color might be a mere visual stimulus but this hypothesis was rejected after examination of the CT scan images. They indeed show an expanded anterior enlargement of the digestive track located closely to the inner face of the depression and apparently related to it by a differentiated cell layer. This enlargement is supposed to be the hollow pharynx referring to its position relative to the tentorium and the buccal cavity.

References are rare to interpret these images. Some authors investigated the internal anatomy of the head in Hymenoptera , notably for honeybees (e.g. Snodgrass 1910) and Formicidae (e.g. Whelden 1960; Hansen et al. 1999). Available data on other families are still scarcer but we could find punctual investigations for Xyelidae ( Beutel & Vilhelmsen 2007) , Braconidae (Vinson 1967) and even Ichneumonidae ( Benham 1972) . No one shows such an enlargement of digestive tracks close to the face.

As a result, we hypothesize that this facial depression could be involved in sound reception. External sounds would made the ground (tympanal organ?) of the male depression vibrate and the vibrations amplified by the dilated pharynx acting as a soundbox. This organ would intra or interspecific sound communication in the closed environment of the high-altitude rainforest of PNG. Hearing organs in insects are defined by the presence of a tympanal membrane backed by an air-filled space or cavity and innervated by a chordotonal sensory organ, a cluster of sensilla connected to moveable parts of skeletal cuticle ( Hoy & Roberts 1996; Field & Matheson 1998). Eardrums in insects typically arose through a thinning of the cuticle, but the CT-scan images do not show here any thinning of the internal cuticule of the depression. However in some insects the sound-receiving cuticular region is not conspicuously thinned, although the cuticle might nonetheless vibrate in response to sound and function as an eardrum ( Hoy & Roberts 1996; Göpfert & Hennig 2016).

Our hypothesis remains of course highly speculative because such a facial tympanal organ is to now totally unknown in Hymenoptera . More researches are needed to test this hypothesis. We could not so far examine the internal structure of head in female for comparison, confirm the membranous nature of the putative eardrum, nor investigate the histological nature of the cell coating layer to confirm or not if it actually connects the putative eardrum to the pharynx. But this is at minimum another major argument to support the taxonomical originality of Cyclopocentrus kumbuglu n. gen. and n. sp.