Heteropsis drepana

C, Lees David, 2016, Heteropsis (Nymphalidae: Satyrinae: Satyrini: Mycalesina): 19 new species from Madagascar and interim revision, Zootaxa 4118 (1), pp. 1-97 : 21-24

Heteropsis drepana group (including subgenus Henotesia Butler, 1879 )

The Ht. drepana group (comprising the ‘ narova -group’ of Torres et al., 2001, otherwise known as subgenus Henotesia sensu Lees et al., 2003 , but referred to here as the ‘ Henotesia clade’, with the addition of H. drepana and Ht. fuliginosa ; see Aduse-Poku et al., 2015) also includes the prior genus Houlbertia. The type species of the last genus is Houlbertia erebennis Oberthür, 1916 , as selected by Viette, 1961, which has priority over Hemming’s, 1964, selection of Erebia passandava Ward, 1871 as the type species; NHM’s Lepindex needs to be updated in this regard. The (subgenus) Henotesia as circumscribed by Lees et al., (2003), who synonymised Houlbertia with it, is based on He. wardii Butler, 1879 , which became on combination a junior subjective homonym of Ht. wardii (Mabille, 1877) . The Ht. drepana group, or at least the Henotesia clade, is supported by both morphological ( Lees, 1997) and molecular data (e.g. Kodandaramaiah et al., 2010). Perhaps the most important morphological feature of the Henotesia clade is the dorso-ventrally flattened aedeagus ( Lees, 1997). However, the entire Ht. drepana group includes moderately to highly sexually dimorphic species ( Lees, 1997). Aduse-Poku et al., (2015, Suppl. Fig. S4) using ten genes found that their 11 sampled members of the Henotesia clade, had 92% bootstrap support as sister to Ht. drepana , and this clade 87% support as sister to the hitherto hard to place Ht. fuliginosa . The sister of Ht.

drepana group is recovered with significant bootstrap support ( Aduse-Poku et al., 2016, in press) as sister to a ‘ Ht. iboina group’ of Malagasy Heteropsis that included Ht. vola , that was previously (and erroneously) placed in (subgenus) Telinga by Lees et al., 2003. The iboina group also comprised Ht. iboina (Ward, 1870) + Ht. parva ( Butler, 1879) and members of the Ht. avelona group (= bicristata group of Torres et al., 2001), including Ht. avelona (Ward, 1870) + Ht. uniformis ( Oberthür, 1916) , Ht. parvidens (Mabille, 1880) , and Ht. bicristata ( Mabille, 1878) . No undescribed species are known in the Ht. iboina group. In general, the question of the number and limits of subgenera in Heteropsis is left open here. If the name Houlbertia Oberthür was restricted to a clade of highly sexually dimorphic species that include Ht. narova , excluding Ht. passandava according to Aduse-Poku et al., (2015, Fig. 1 ), with Henotesia Butler retained as a separate subgenus, this would leave Ht. passandava without a subgeneric name. While more faithful to traditional taxonomy in Madagascar, this also seems complicated and a pragmatic solution might be to synonymise both Henotesia and Houlbertia with the nominotypical subgenus Heteropsis Westwood , perhaps with the addition of the previously enigmatic Ht. fuliginosa , forming together the well-supported clade in Aduse-Poku et al., (2015, Fig. 1 ). However, such an action, that would abandon names in prior use as historic curiosities, awaits a forthcoming study. It should also be noted that without explicit justification (perhaps merely because it seemed different to other African hairy-eyed mycalesines), d’Abrera (1997: 226) included the African species ‘ peitho Plötz, 1880’ in ‘ Houlbertia ’ (but see new genus in Aduse-Poku et al., 2016, in press).

Species of the Ht. drepana group with the most vivid blue upperside ♂♂ such as Ht. erebennis occur at lower elevations than those with dark ♂♂, further discussed below. The last group tend to be found in more montane areas, up to over 2300 m, in the case of Ht. pauliani , which is as high as any mycalesines are found in Madagascar. ♂♂ of most species are quite rarely attracted to ripe fruit and Ht. andravahana stat. rev. (see also below) has been observed to feed on flowers of Dichaetanthera Endl. ( Melastomataceae ; Lees, 1997), whereas the opposite sex is regularly attracted to fruit. Adults sometimes feed on in situ fruits. Species fly from understory to lower canopy, males following lightgaps, sometimes resting on top of clumps of bamboo, whereas Ht. drepana males hilltop and rest on top of ridge bushes in sunny weather. Many if not all species seems to be closely associated with different types of forest bamboo, but early stage data is very sparse. Stands of bamboo (such as of the giant Cathariostachys S.Dransf. in the lowlands) can sometimes support large sympatrically flying populations of several closely related species. ♀♀ of the more sexually dimorphic members of the Ht. drepana group tend to be very sedentary resting low in the undergrowth on leaf litter (where they are cryptic with fallen leaves including those of bamboo), and may not be noticed at all when not venturing off paths. The ♀♀ of Ht. passandava itself closely resemble those of the distantly related Ht. iboina (Ward, 1870) and Oberthür (1916) not only described its ♀ as a form (fitensis) of ‘ iboina ’, but in a different genus than Culapa Moore, 1878 , as Henotesia andravahana var. macrophthalma Oberthür, 1916 , in which he also included ♀♀ of Ht. passandava along with Ht. obscura ( Oberthür, 1916; see Lees, 1997). The more montane members of the Ht. drepana group have ♀♀ that tend to be vertical trunk-resting (as do often the ♂♂, likely explaining their less sexually dimorphic wing patterns).

Within the Ht. drepana group are two or more groups of phenetically similar, moderately sexually dimorphic species, which inhabit montane forest ridges between 1000–2000 m. Two sampled members in the phylogeny of Aduse-Poku et al., (2015, Fig. 1 ) are Ht. viettei Lees, 2003 and Ht. ‘ andravahana ’ (= Ht. westwoodi sp. nov., see below), which were sister to Ht. obscura + Ht. difficilis (Mabille, 1880) (= Ht. andravahana stat. rev., see also below). Unlike the last two species, the ♂♂ of the first pair have dark, non-reflective uppersides, and correspond to sp. 13 (‘ Henotesia sp. 13’ Kremen, 1994: 417; now including sp. 74), and sp. 13B, respectively, of Lees (1997: 64). This species pair exhibit varying degrees of whitish or cream highlighting distal of the HWV Mb. The Ht. harveyi sub-group of the Ht. drepana group represents a tightly knit species complex with dark-uppersides and includes Ht. harveyi sp. nov. and Ht. vanewrighti sp. nov. (sp. 12, and sp. 63, respectively), as well as at least one undescribed species (sp. 20, which is the same as sp. 77; see Lees, 1997; Torres et al., 2001). The Ht. harveyi sub-group exhibits the presence or absence of a russet or dark patch on the underside of the FW along vein 1A+2A (in Madagascar, only the Ht. avelona group of mycalesines also exhibit a FWV patch). Another dark non-reflective upperside pair, more similar to Ht. viettei + Ht. westwoodi , although not represented in the phylogeny of Aduse-Poku et al., (2015), and so with relationship unresolved, is Ht. imerina sp. nov. (sp. 22; same as sp. 76 of Lees, 1997; Torres et al., 2001) and Ht. pauliani (sp. 37 of Lees, 1997). In order to describe these five new species, it is necessary to go into some detail first to clarify the identity of Mycalesis andravahana Mabille, 1878 , which potentially blurs the identity of some of these species, and then to lectotypify it.

Mycalesis andravahana problem. Although Mabille’s original (1878) short unillustrated description of Mycalesis andravahana, in Latin , does not mention the number of specimens, citing only Madagascar as type locality, the first and main part of the short description matches at least the (apparently lost, see below) ♂ specimen illustrated in Mabille ([1885]: Pl. 5, Fig. 6.7). However, this first part does not match the ♂ of the taxon represented by the “var.” in the latter work and illustrated in Mabille [1885]: Pl. 5, Fig. 8 ). The additional details in Mabille ([1885 -7]) suggests that the original description had been based on at least three phenotypically divergent specimens, two ♂ and one ♀ (that illustrated dorsally by Mabille, [1885]: Pl. 5, Fig. 9 ).

Despite several searches over two decades and no doubt by previous potential authors interested in resolving the problem (including P. Viette, who had planned to revise the Malagasy satyrines; all of his ms lectotypes of this group are unpublished), no possible ♂ syntypes of Mycalesis andravahana had been traced in the three main feasible repositories for Mabille material, BMNH and MNHN (DCL, pers. obs.), or ZHMU (W. Mey, pers. comm.). Although the ♂ ( Fig. 6A , ‘6, 7’) might have been a good choice for the lectotype of Mycalesis andravahana Mabille, 1878 , unfortunately both it and the “var.” appear to have been lost or destroyed. However, Lees (1997: 33, 62) and Lees et al., (2003: [789], note 45) already suggested a possible solution to the vexing problem of typification of this nominal taxon by fixing as LT a ♀ specimen that came from the Henley Grose-Smith collection (via the Joicey bequest to BMNH). This specimen bears a historical rectangular “ TYPE ” label, without labelling of its type identity and is now numbered as BMNH(E) #674841. Because of its wide and distinctive HWV white postmedial band, which is strongly indented around the HWV space-CuA1 ocellus, this specimen had been thought (Lees et al., (2003: [789], note 45; where “ H. 13” was a lapsus for ‘ H. 13B’), to correspond to ‘sp. 13B’ (described below as Ht. westwoodi , a species apparently endemic to the southeastern mountains of Madagascar, from Andohahela to Ranomafana and Andringitra National Parks).

A close re-examination of this specimen (BMNH(E) #674841) now reveals diagnostic features (a highly crenate HW, and HWV ocelli, apart from that in space-CuA1, expressed as strong white spots, as well as a strongly indented near-right-angle of the Mb around CuA1), that together identify it as a potentially dry season form of a species that was also described by Mabille as Mycalesis difficilis Mabille, 1880 . This last taxon is illustrated in Mabille, ([1885], Pl. 7B, ‘2’, ‘2a’) and reproduced here ( Fig. 6A , ‘9’). M. difficilis is based on a single individual, for which a matching ♀ HT (DCL-DB-2872), previously unlabelled as a type, was already located in MNHN ( Lees 1997: 389; Lees et al., 2003). Its (metallic blue upperside) ♂ was described by Oberthür 1916 as Henotesia undulosa (together with a non-reflective form¸ var. luctuosa ), and its ♀ described independently in the same work as He. andravahana var. macrophthalma (see Lees, 1997). Although it is has not previously been clear what species is represented by the ♀ whose dorsal surface only is illustrated in Mabille ([1885]: Pl. 5, Fig. 9 ) as an example of Mycalesis andravahana , indeed it cannot be ruled out that this illustrates (considering its small ocelli), a potentially dry seasonal form of Mabille’s nominal species M. difficilis . In fact the resemblance to a recent photograph of a living ♀ from Anjanaharibe Sud is striking ( Fig. 6 C). About the original ♀ specimen, Mabille ([1887]: 58) stated “La femelle est d'une couleur plus claire que le mâle, surtout sur le disque. L'ocelle des ailes supérieures est très grand, d'un fauve clair; l'ocelle des inférieures est très petit. En dessous, les premières ailes ont les zébrures blanchâtres jusque dans la cellule; l'espace terminal est d'un gris cendré. Aux inférieures, il y a les mêmes dessins, mais plus délayés; la bandelette blanche est d'un gris sale. Les autres parties de l'insecte sont comme dans le type mâle que nous avons décrit plus haut”. Just below this he stated “L'envergure est de 36 millimetres chez le ♂, de 44 chez la ♀. Mycalesis Andravahana à été envoyé a le Henley Grose Smith des parties boisées du nord-est de l'île”. The HT of Mycalesis difficilis had the same provenance (Mabille, 1880), but as implied above, would represent a likely wet seasonal form of the same species. The potential ♀ ST of M. andravahana coming from the Grose-Smith collection (BMNH(E) #674841; Fig. 6 D) measures 50.2 mm absolute wingspan and between 23–23.3 mm FWL. However, all four wings are fixed to the body with the FW tergal edges angled in a non-orthogonal orientation, with blobs of red sealing wax. My digital reconstruction of the wings shows that the apex to apex dimension, when reconstructed as Mabille ([1885]: Pl. 5, Fig. 9 ) actually would measure 44– 45 mm apex to apex. It should be acknowledged that in general, Heteropsis are in general hard to identify from uppersides, more so for a painting of a ♀. Oberthür (1916: 205) was convinced that it represented the ♀ of the taxon/taxa in his collection that he referred to as “ wardii - andravahana ”. This usage might reflect his insight into the likely compound nature of Mabille’s description (and redescription) of M. andravahana ; indeed, Oberthür labelled one specimen (now numbered BMNH(E) 647848) with the above compound taxon name. Perhaps (or not), Oberthür was aware of the underside pattern of the specimen in the Grose-Smith collection (whether or not he ever visited the Grose-Smith or the Joicey collection, housed in Surrey, where the specimen would have gone in 1910).

Oberthür nevertheless named his form ‘ macrophthalma ’ as a “var.” of “ Henotesia andravahana ”. In any case, it now transpires that the only surviving potential syntype of Mycalesis andravahana (BMNH(E) #674841) that was previously located ( Lees, 1997, Lees et al., 2003) does not disagree with Mabille’s ([1885-7]) redescription in any significant aspect. The dorsal illustration would indeed represent a quite accurate portrayal of the original specimen. There is indication, especially on the non hand-coloured copy of Mabille’s work available on Biodiversity Heritage Library ( Fig. 6 B), of the pale HWV postmedian area showing through to the dorsal surface, just as it does in BMNH(E) #674841, and as clearly indicated in the illustration of Mycalesis difficilis in Mabille ([1885 : Pl. 7B) ( Fig. 6A ). The relative size of dorsal FW and HW ocelli and their rings in these two versions of Mabille’s illustration (Pl. 5, ‘9’) is also convincingly accurate as corresponding.

The specimen labelled “BMNH(E) #674841” is therefore now designated as the lectotype ♀ of Mycalesis andravahana Mabille 1878 .

As mentioned above, Oberthür (1916) also included ♀♀ of what has been known (Lees et al., 2003) as ‘ Ht. difficilis ’, as well as of Ht. obscura and Ht. passandava , in his series of Henotesia andravahana var. macrophthalma . As it was originally curated (pers. obs.), Oberthür’s series of “ wardii - andravahana ”, however, came mainly from Fianarantsoa, with a few specimens from Antsianaka, and essentially comprised ♂♂ and a few ♀♀ of just one species, previously known as morphospecies 12 (‘ Henotesia ? obscura (sp. 12)’ of Kremen, 1994: 417) from Ranomafana; actually, the unrelated Ht. obscura , which like the nominal species Henotesia undulosa Oberthür, 1916 , has dull metallic blue upperside ♂♂, is not known from the Southeast of Madagascar). ‘Sp. 12’, as known from Ranomafana, is now described below as Ht. harveyi (a manuscript name of Lees, 1997). Ht. harveyi not only has ♂♂ with very dark brown uppersides, but also lacks the distinctly white postmedial band which as mentioned above is present in Ht. viettei . It is not certain if the species without HWV whitish postmedian highlighting represented as “var.” of Mycalesis andravahana in Mabille’s description and redescription (1878; [1885]: Pl. 5, Fig. 8 ; [1887], text) and reproduced here in Fig. 6A is referable to Ht. harveyi , but it might be. However, Aurivillius (1911) described a similar ‘var.’ lacking the white postmedian HWV band as ‘ Henotesia andravahana ab. marmorata ’ (specimen represented in Fig. 7 C). Its HT ♂ (‘TYPUS’, specimen NHRS- TOBI000000050) was examined by DCL in SNHM. This is an unavailable name (clearly infrasubspecific, Art 45.6.2), but the specimen matches well Ht. harveyi , in fact much more clearly than does that illustrated in Mabille ([1885]: Pl. 5, 8). As judged by Mabille’s [1885] plate, Ht. harveyi may have been one of the species under Mabille’s eyes when he described M. andravahana , even though not the first mentioned, or focal, form of the description. In Ht. harveyi , although the ventral irrorated pattern and general colouration resemble the ventral illustration in Mabille ([1885]: Pl. 5, Fig. 8 ), generally the median and pre-median HWV bands are much more clearly delineated. This is the case in the HT of Aurivillius’ aberration, and in most specimens of Ht. harveyi , but not so for Mabille’s ♂ ‘var.’ ( Fig. 6A , ‘8’).

To resolve the Mycalesis andravahana problem, the following synonymy of Heteropsis andravahana is thus proposed:

Heteropsis andravahana ( Mabille, 1878) . LT ♀, here designated ( Fig. 6 D), lacking abdomen (the only known surviving potential ST, from the Grose-Smith collection): BMNH(E) #674841, bearing labels “LT (round purplebordered label)/ Type HT (round red-bordered label)/ TYPE. [white red printed label]/Ex Grose-Smith, 1910/Joicey Bequest Brit. Mus. 1934-120./[Colour photocopy of Fig. 9 from plate 5 of Mabille, 1887]”;

Mycalesis difficilis Mabille, 1880 , syn. nov. HT ♀ (Lees et al 2003: note 42, [p. 789]): MNHN (DCL-DB- 2872);

Henotesia undulosa Oberthür, 1916 , syn. nov. LT ♂, here designated: BMNH(E) #674825, LT (round purplebordered label)/ Type (round red-bordered label)/ Madagascar, Antsianaka, Perrot Freres, 2e Semestre 1893/Ex Oberthür Coll. Brit. Mus. 1927-3/[copy of f. 2830 in Oberthür, 1916];

Henotesia undulosa var. luctuosa Oberthür, 1916 , syn. nov. LT ♂, here designated (of 4 potential STs), lacking abdomen: BMNH(E) #674827, bearing labels “LT (round purple-bordered label)/ Type (round red-bordered label)/ Madagascar, Antsianaka, Perrot Freres, 2e Semestre 1893/Ex Oberthür Coll. Brit. Mus. 1927-3/ Gallienia undulosa var. luctuosa Obthr. [label in Oberthür’s handwriting]”;

Henotesia andravahana var. macrophthalma Oberthür, 1916 , syn. nov. LT ♀, here designated, lacking abdomen (of 34 original STs, mixed series): BMNH(E) #674826, bearing labels: “LT (round purple-bordered label)/ Type (round red-bordered label)/ Madagascar, Antsianaka, Perrot Freres, 2e Semestre 1893/Ex Oberthür Coll. Brit. Mus. 1927-3/[copy of f. 3042 in Oberthür, 1916]”;

The above lectotypification releases the four montane species below to be described anew.

Figures

FIGURE 1. Venation and wing patterning diagram of a fairly typical species of Heteropsis from Madagascar (Heteropis antahala group; ♂), DV. Sdbp = Supra-discocellular brush and patch. Cub = Cubital brush. Ventral banding: Cb = Costal bars. Mf = Median fleck (pale highlight). Mb = Median band. PMb = Premedian band. Ore = Ocellus ring extension. Pml = Premarginal line. Sml = Submarginal line.

FIGURE 7. Adults in DV (left), VV (right) or in natural postures. A: HT ♂ of Heteropsis harveyi Lees & Kremen, sp. nov. (NHMUK 010289119). B: PT ♀ of Ht. harveyi (NHMUK 010289121). C: HT ♂ of Henotesia andravahana ab. marmorata Aurivillius, 1911 (NHRS-TOBI 000000050). D: pair of Ht. harveyi in copula (Maromizaha, photo courtesy Lance Featherstone). E: ♂ of Ht. imerina (Ankazomivady, 23 / 03 / 2003, photo: D. C. Lees), for comparison, showing relatively produced ‘ tail’ on HW vein M 3.

FIGURE 8. ♂ genitalia in LV (left), DV (right). A: Heteropsis cf. harveyi (see text) (Anjozorobe; 98 DL). B: Ht. vanewrighti Lees, sp. nov. (Mt. Bekolosy; 233 DL). C: Ht. imerina Lees, sp. nov. (Ankazomivady; 154 DL). Scale bar: 3 mm.

FIGURE 9. Adults in DV (left), VV (right) A: HT ♂ of Heteropsis vanewrighti Lees, sp. nov. (Mt. Bekolosy; NHMUK 010289124). B: PT ♀ of Ht. vanewrighti (Mt. Bekolosy; NHMUK 010289125). C: HT ♂ of Ht. imerina Lees, sp. nov. (Mantadia; NHMUK 010289132). D: PT ♀ of Ht. imerina (Ankasoka, route Lakato; DCL-DB- 2906, MNHN). Scale bar: 30 mm.

FIGURE 6. A: pastiche of illustrations from Mabille ([1885]) in Grandidier’s work, Figs 6, 7, 8, 9 from Pl. 5, Figs 2, 2 a from Pl. 7 B. The ♂ “ var. ” in (Pl. 5: Fig. 8) of that work might possibly represent Ht. harveyi Lees & Kremen, sp. nov., while the ♂ in (Pl. 5, Figs 7, 8) probably represents Ht. viettei (Lees, 2003) and the ♀ upperside in (Pl. 5: Fig. 9) probably represents the same species, described as Mycalesis difficilis Mabille, 1880 (probable HT illustrated: Pl. 5, Figs 2, 2 a). B: alternative illustration prior to hand colouring (from copy of Mabille ([1885]) in BHL online library) of Mabille’s Pl. 5: Fig. 9; C: Ht. andravahana (Mabille, 1878), stat. rev. ♀ (Anjanaharibe Sud 25 / 11 / 2014, photo: D. C. Lees), representing a dry seasonal form of this species, for comparison with D: chosen LT ♀ of Mycalesis andravahana, DV, VV (BMNH (E) # 674841; considered to represent the ♀ specimen depicted by Mabille ([1885]: Pl. 5. Fig. 9.