Second contribution to the knowledge of the Lithosiini of Gabon : the genus Nanna Birket-Smith ( Lepidoptera , Erebidae , Arctiinae )

The presence of the genus Nanna in Gabon is treated. Four species are recorded, two of which are new to science ( N. molouba sp. nov. and N. semigrisea sp. nov.), and the hitherto unknown female of Nanna ceratopygia is described. A checklist and a key to the species of the genus Nanna are provided.


Introduction
This second contribution on the Lithosiini of Gabon (the fi rst one is Durante 2012) deals with the presence of the genus Nanna Birket-Smith, 1965 in the Ivindo National Park (North-eastern Gabon).This genus was proposed by its author mainly based on West African material: Kiriakoff (1963) described Lithosia colonoides (Kiriakoff, 1963) from Democratic Republic of the Congo and recently Kühne (2005Kühne ( , 2007) ) extended its range to East Africa (see list below).Hitherto it has not been recorded from Gabon.
The genus was diagnosed by its author by means of the irregular weak sclerotizations of the aedeagus and the absence of the valvellae; he also added (Birket-Smith 1965: 35) a small plate-like cornutus in the vesica, the vinculum not forming any saccus, and the distally rounded, membranous supravalva.
With the present contribution, the number of Nanna species rises to 13, two of which are newly recorded from Gabon and two more are here described as new.
A key to the species of the genus Nanna is provided below.

Material and methods
All the Gabonese material used for this paper comes from the moths collected by the fi rst and second authors in the National Park of Ivindo, during the short rainy season from February to April.
The Park, in the North-East of Gabon at about 500 m a.s.l., has an extension of 300 000 ha, most of which is covered by primary forest and is drained by a lot of watercourses, the most important being Ivindo, the major, Djidji and Langoué (Vande Weghe 2006).The Ipassa Research Station, the place of the captures, is in the far north of the park, and the forest around is somewhat degraded with residual primary forest, old and young secondary forest, and bush and crop areas.The Ivindo National Park presents a forest with transitional characteristics between the Western Congolese forest with Atlantic infl uences and the real Western Congolese forest, without a clear line of demarcation (Vande Weghe 2010).The climate is equatorial with two wet seasons (a short one from February to April/May, and a longer one from August to November), a short dry season in December-January, and a long dry season from May/June to July/August.The moths were attracted at night using three blended light lamps of 500 watt placed at about 200 m from each other, in front of a white cotton sheet.The collection started actively from 7 p.m. to about midnight.Much material was also collected in the morning on and around the white sheets.
One sheet was positioned in front of the secondary forest around the station, the second one in front of the Ivindo River on the top of a slope high over the primary forest canopy (Fig. 1), and the third one facing the research station inwards (Fig. 2).The Lithosiini moths were killed immediately after capture, using ammonia vapours, and pinned the following morning.After a few months, the moths were rehydrated and spread with standard methods (e.g., Winter 2000).
Pictures were taken in the fi eld (landscapes) and in the laboratory (adults) with a digital camera; genitalia were photographed and mounted under a trinocular stereomicroscope.The pictures were improved using the software Photoshop CS4.
Pictures of genitalia are not in due proportion.
The material is stored at the MSNS and the fi rst author's collections.

Remarks
In the original paper, Birket-Smith does not provide any information about the etymology and grammatical gender.It is very probably derived from the name "Nanna", the god of the moon in Sumerian mythology.That name should be masculine, but the author treats it as feminine, making it agree with the specifi c names, such as magna.According to the art.30.2.3 of ICZN, the genus name Nanna has to be considered feminine.

Key to the males of the species of Nanna
The following key is based mainly on the genitalia because the habitus of the species included in the genus Nanna is very similar.Furthermore the females are not included due to our fragmented knowledge about them.

Description Habitus
Wingspan: ♀ 39-44 mm.General aspect similar to the male, larger and with a bigger abdomen.

Genitalia
Seventh segment slightly sclerotized, one and a half times longer than the sixth segment; the sternite with a deep concavity in the middle of its caudal margin to accommodate the sterigma; pleurae with one triangular membranous thickness, cephalad with a shallow pouch consisting of membranous thick tissue.Eighth segment slightly sclerotized, as long as the sixth one; sternite with two membranous lateral shallow concavities and the central portion forming a well sclerotized squarish lamella postvaginalis.Anterior apophyses very short, about one third as long as the eight segment; posterior apophyses two times longer than the anterior ones.Dorsal pheromone glands a little longer than the posterior apophyses.
Ostium bursae membranous, no lamella antevaginalis.Ductus bursae almost as long as large broad, internally covered by spiculate thorns; its caudal portion naked, with a small dorsal dome-shaped appendix holding the ductus seminalis.Corpus bursae lenticular, internally covered by spiculate thorns except around the signum.Signum concave, roundish, with small thorns.

Remarks
This species, hitherto only known from the type locality and Nigeria, is now also reported from Gabon.The female is recorded and hereafter described for the fi rst time (an allotype is designed in the sense of ICZN art.72.1.3,Recommendation 72A).Nanna eningae (Plötz, 1880) Figs 3C-F (Strand 1912;Birket-Smith 1965); DRC (RMCA coll.);Gabon (new locality).
Remarks Kühne (2005) synonymized Nanna montana and N. eningae var.pia with N. eningae based on the variability of the wing colour and the colour pattern of N. eningae (see Kühne 2007: 354), and on comparisons with the type of N. eningae (Kühne 2007: 354), disregarding the male genitalic differences described by Birket-Smith between N. montana and N. pia.They were probably considered to represent only a minor intraspecifi c variability.According to Birket-Smith (1965: 43, 47), the types of N. eningae and N. pia were untraceable, but Kühne (2005) explained that he was able to compare his specimens of N. eningae and N. pia with the types in the Plötz collection.However, Kühne (2007) admitted to the possibility that the N. eningae-group is composed of several taxa, but he did not go further into this matter.
Durante & Panzera ( 2001) did not provide clear distinctive characters for lack of material and classifi ed a single specimen from Nigeria as N. montana, based on male genitalic morphology.In this respect, it should be remarked that one specimen from Ghana (RMCA collection) shows all the characteristics described by Birket-Smith (1965) about N. pia, leading to the conclusion that the variability between N. pia (eningae) and N. montana could be concrete.
At present, although the examined series from Gabon does not show any genitalic variability within itself (in contrast with the above supposed "minor intraspecifi c variability"), the conclusions of Kühne (2005) are here cautiously accepted, since the synonymy can be supported by the similarity of the Gabonese female genitalia with the ones of N. pia fi gured in Birket-Smith (1965) and the similarity of the Gabonese male genitalia with the ones of N. montana, both fi gured in Birket-Smith (1965) and in Durante & Panzera (2001), in addition to the existing wing colour variability (see Fig. 3E).

Diagnosis
Species quite distinguishable in the genus thanks to the light grey forewings and the usually paler hindwings; it is, however, very similar to Nanna griseata.Male genitalia similar to those of Nanna collinsii and N. griseata, but the processus basalis plicae is longer and more falcate in N. semigrisea sp.nov.; moreover, N. collinsi has shorter valvae and a larger aedeagus, N. griseata a more slender aedeagus.

Etymology
The name originates from the combination of the Latin prefi x semi-(meaning "half") and the adjective griseus (meaning "grey").

Description Habitus
Wingspan: ♂ 36-41 mm; ♀ 31 mm.Forewing upper side ground colour light grey, slightly darkening distally; costa with the proximal two fi fths greyish; pattern with two black dots on the median area, one on the costal margin, and another one between the CuA2 and the anal vein.Hindwing upper side ground colour white, sometimes shading to grey toward the apex and termen.Forewing underside ground colour darker grey than the upper side, gradually shading to creamy white toward the base and the inner margin.Hindwing underside as in the upper side.Fringes of the same colour as that wing's ground colour.
Male legs light yellow with the distal end of the femur dark grey; forelegs with femur and tibia shading into light grey; in the female homogeneous light yellow.Praetarsi blackish.Spurs formula 0-2-4.
Abdomen greyish white, shading to ochreous yellow starting from about the sixth segment to the terminalia.Ala valvae sclerotized, distally tapering.The distal process strongly sclerotized, covered by thorns dorsally for all its length and completed only on its last fourth, ending in a point.Supravalva more membranous, fi nger shaped, just a little shorter than ala valvae.
Aedeagus quite narrow, partially membranous (pseudaedoeagus of Birket-Smith 1965), with a distal narrow sclerotized bar terminating in a curl.Vesica small, membranous, with two proximal (numbered 1 and 2 in Fig. 4C) and two distal (3 and 4) lobes covered by very small spines; opposite to the curl (arrow c), a moderate swelling (s).FEMALE.Seventh segment slightly sclerotized, one and a half times longer than the sixth segment.Eighth segment about half as long as the sixth one, membranous.Anterior apophyses as short as the eight segment; posterior apophyses three times longer than the anterior ones.Dorsal pheromone glands a little longer than posterior apophyses.
Lamella postvaginalis present as a strong shield inserting cephalad in the seventh sternum, which surrounds the lamella with two lateral well evident folds.
Ostium membranous, with two small anterior swellings, which cannot be interpreted as lamellae antevaginalis.
Corpus bursae ovoid, internally covered by small slender thorns except around the signum, with a small dome-shaped posterior appendix; ductus seminalis arising from the appendix.Signum concave, oblong, covered by very tiny spines.

Description Habitus
Wingspan: ♂ 36-42 mm.Forewing upper side ground colour whitish; basal third of the costa sometimes blackish; pattern with two black dots on the median area, one on the costal margin, and another one between the CuA2 and the anal vein.Hindwing upper side ground colour white.Forewing underside ground colour slightly darker than the upper side, with a creamy fascia along the costal margin.Hindwing underside as in the upper side.Fringes of the same colour as that wing's ground colour.
Head yellow, labial palpi yellow, porrect or slightly downward, male antennae bipectinate, yellow along the fi rst fourth, shading to brown toward the apex.
Forelegs yellow with femur and tibia externally grayish brown, tarsus lighter; mid-and hindlegs yellow with articulation between femur and tibia grayish brown.All praetarsi brown.Spurs formula 0-2-4.Abdomen yellow.
Basis valvae sclerotized at the junction with the juxta, with which it forms a pseudo-articulation; membranous at supravalvar level.
Distal process of the ala valvae strongly sclerotized, with a triangular process (Fig. 4D, arrow) at the distal third of the valvar length, and a robust hook-like termination ending in a big thorn and dorsally covered by slim spines.Distal process of the supravalva less sclerotized than the previous one, large and spatulate.
Aedeagus quite narrow, partially membranous (pseudaedoeagus of Birket-Smith 1965), without coecum; vesica small, with sparse tiny thorns and two lobes, one at the base and one apically on the main lobe (respectively numbered 1 and 2 in Fig. 4D).

Discussion
The genus Nanna is an entirely African genus, with mainly Western and Central distribution, and it seems to be chiefl y associated with a tropical habitat with a high humidity level.
From a strictly systematic point of view, the genus is quite well distinguishable from others by its large wingspan, generally white wing colour and peculiarities in its genital apparatus (see Introduction).
In Gabon, at the moment, it consists of four species; however, given the high biodiversity in other groups (even insects other than moths; Durante, personal observation), it is very probable that the number will increase following future research.Most species, based on present knowledge, occur in Cameroon (six), followed by Gabon (four) and Nigeria (three), due probably to the extension of the recording survey in each country.

Fig. 1 .
Fig. 1.Primary forest on the Ivindo River at the Ipassa research station.Courtesy of Piero Carlino.
Fig. 2. The Ipassa research station in Ivindo National Park.
DiagnosisSpecies similar to N. collinsi, N. colonoides, N. eningae and N. melanosticta, but easily distinguishable thanks to the bipectinate antennae of N. molouba sp.nov.Nanna magna also has bipectinate antennae, but its genitalia are clearly different.