Heteropsis (Nymphalidae: Satyrinae: Satyrini: Mycalesina): 19 new species from Madagascar and interim revision Author C, Lees David text Zootaxa 2016 4118 1 1 97 journal article 38930 10.11646/zootaxa.4118.1.1 a4e408c4-2575-4b5c-878b-e1b9010ec619 1175-5326 264597 CFA586DA-10EE-468B-80EE-35351E3845FD Heteropsis tornado Lees, Allaoui & Aduse-Poku , sp. nov. LSID: urn:lsid:zoobank.org:act:22EB2840-C7CE-42D5-B26F-3DDD0DF3D763 Prior references : sp. 54 ( Lees 1997 , Torres et al., 2001 : 462—where considered to belong to an unsampled clade); “Heteopsis_sp.nov_ tornado ” [ sic , nomen nudum ] in Aduse-Poku et al. , 2015. FIGURE 17. Adults in DV (left), VV (right). A : HT ♀ of Heteropsis tornado Lees, Allaoui & Aduse-Poku , sp. nov. (NHMUK010289143; Anjanaharibe Sud). B : PT ♀ of Ht. tornado (Marojejy; KAP-MAD1403). C : ♂ of Ht. tornado (‘Mahassabe’) [from colour transparency]. D : ♀ of Ht. tornado (‘Mahassabe’; BMNH(E) #674881). Scale bar: 30 mm. FIGURE 18. ♂ genitalia in LV (left), DV (right). A : Heteropsis tornado Lees, Allaoui & Aduse-Poku , sp. nov. , ♂ genitalia (‘Mahassabe’; 362DL) (detail of valve inset; drawing modified after Lees, 1997: 106); not to scale. B : ♂ genitalia of Ht. lanyvary . (left: DL-05-761; right: DL-05-845). C : ♂ genitalia of Heteropsis barbarae sp. nov. (Antsamanarana R., Masoala; CK30, 141DL). D : ♂ genitalia of Heteropsis menamenoides Lees , sp. nov. (Mt. Bekolosy; DLBEK94_180; 231DL). Scale bar: 3 mm. Type material., Deposition BMNH : Holotype : ♀ ( Fig. 17 A), Madagascar NE, R.S. Anjanaharibe Sud, slope above R. Marolakana, road between Befingotra and Analamazava, 14.7676o S , 49.4815o E +/- 0.01 km , 1000 +/- 5 m , 20/11/2014 : 09:07, Ahamadi Allaoui. DL 14A-0245, IA518 [isotope voucher], NHMUK 010289143 [ QTR barcode]. Paratype : Deposition ABRI : ♀ ( Fig. 17 B), Madagascar NE, P.N. Marojejy, 670 m , [ 14.4354o S , 49.7647o E ; using GPS point 30B672], ridge above “Vallée de Brookesia” below “Camp 2 Marojejy” and above Manantenina river, primary rainforest, on path, 29/01/2014 , Kwaku Aduse-Poku, MAD 1403 [=KAP-MAD1403; DNA voucher]|KA1015 [=KA-P1015; extract number]|IA313 [isotope voucher]. Deposition summary : BMNH (HT ♀); ABRI (PT ♀). Type locality. Anjanaharibe Sud, slope above R. Marolakana, 14.7676o S , 49.4815o E +/- 0.01 km , 1000 +/- 5 m . Diagnosis. The species is very distinctive especially on the underside (see Fig. 17 ) and not readily confused with any other species. It has a superficial resemblance though on the dorsal surface to Ht. iboina . Ht. tornado is significantly larger than the typical small size of the four known members of (see Lees, 1997 ) of the Ht. strigula group ( Ht. ankova , Ht. lanyvary , Ht. turbata and Ht. pallida ) (three members of which formed a paraphylum in the study of Aduse-Poku et al. , 2015), which like Ht. tornado routinely show expression of HWD ocellus M3. Including referred specimens, the ♂ and ♀ are similar, but the three known ♀♀ are larger than the one known ♂. Description. Wings : upperside uniform mid brown, with a reddish-orange flush in the basal areas of both wings. FWD space-CuA1 ocellus with black iris more than spanning veins CuA1-CuA2. Orng very large, subhexagonal, orange, spanning from upper part of space-M3 down to upper half of 1A. Space-M1 ocellus is slightly elliptic (especially on the HT right FW) and quite large in comparison with Malagasy Heteropsis , with a quite wide black iris and narrow orange ring, which extends slightly into space-M2. In HWD three subelliptic ocelli are expressed, the smallest in space-M2, but whose ring is almost contiguous with the ring of the largest of these ocelli in space-M3 whose ring nearly spans veins M3-CuA1, the second largest and most elliptic below it in space-CuA1 spanning about ¾ of inter-vein distance CuA1-CuA2. This is an eyespot expression configuration seen in the Ht. ankova clade, which is reminiscent also of ‘ Telinga ’ oculus ( Marshall , 1880) in the Western Ghats. Proximad of the smallest HWD ocellus, whose ring is about half the diameter of the largest, an orange Mf (median fleck) or crescent mark hugs vein M3 as far as its intersection with Mb (faintly visible on HWD). This Mf is a characteristic of quite a few species of the Ht. strigula group, most markedly so in Ht. strigula . HW distinctly crenate (as strongly so as in Ht. bicristata ), outwardly reflexed to each darker part of the fringe (forming in effect very convex tails) at the end of veins Rs-CuA1. Underside has an ochreous yellow cast, irrorated with rufous brown especially proximad of the Mb. FWV space-CuA1 Orng varying from light orange to darker orange towards the tergum, where occupying the width of space-1A, its Ore (ocellus ring extension) forming a ‘tornado’ pattern comprising the pale light orange-yellow outer arm of, in effect, a spiral., The paler yellowish colour of this outer arm of the spiral contrasts with the eccentric, elliptic to sub-hexagonal shape of the FWV Orng; such a pattern is evident, with the proximad yellow arm not usually as well separated costad, in a few other Malagasy Heteropsis which would be placed in different clades, such as Ht. iboina , Ht. kremenae , and Ht. undulans ( Oberthür, 1916 ) . This yellow arm follows a strongly concave russet-brown Mb before it bends back sharply towards near-mid costa. FWV space-M1 ocellus smaller than on upperside, with narrow yellowish ring. HWV space-Rs-M2 ocelli not expressed and HWD orange crescent in space-M2 at base of space near the Mb appears as a prominent light Mf. Space-M3 and CuA1 ocelli about equal sized occupying slightly less than half of respective intervein-spans and with narrow yellowish rings faintly bordered by a rufous brown ring, subelliptic. Space-CuA2 ocellus not obviously expressed. HWV Mb russet, irregular, most angularly convex in space-M2, fairly smoothly concave in space-M3-CuA1, and convexly angled returning towards near mid-anal margin, but terminating just below vein CuA2. Ochreous background more evident distad of HWV Mb, while strong brown irroration is slightly more intense in basal areas of both wings than in distad areas. The russet PMb in HWV crudely mirrors the Mb, but is slightly straighter. A diffuse wavy brown line shadows margin in both wings, but is stronger on FWV, and a narrower brown Sml closely follows it, then closely delineating the crenate margin; before chequered brown and ochreous fringe runs a still darker brown line along the margin in both wings. Four transverse brown Cbs in the FWV cell delineate a more ochreous and weakly irrorated band in the middle of the cell, while the costa itself is quite evenly strigulated. Variation. The ♀ paratype very similar in pattern to the HT ♀. In this specimen, the HWD space-M2 orange Mf is even more obvious. It has a somewhat rounder FWD Orng and a small black spot at the anterior part of this ring; this ‘satellite’ eyespot not seen in either the HT nor referred specimens. On the FWV of PT, ‘tornado’ effect (for Ore) is more extreme, and Mb angularly inflexed rather than smoothly concave proximad of the yellow Ore. The HT was collected effectively in a late dry season, whereas the PT collected well into the wet season, so there is no evidence as yet for seasonal variation. The ♂ (not included in type series) is similar to the HT ♀ but slightly smaller. Like the referred ♀, the referred ♂ is very consistent in wing shape and pattern, although slightly smaller ( 35 mm wsp/ 18.5 mm fwl). However on its HWV, ocelli spaces Rs-M2 are expressed as white points, so perhaps they were collected in a drier season. In the referred specimens also, no orange Mf is obvious in HWD space-M2. Androconia : no ♂♂ in type series/not examined. Wingspan/fwl : 37.15/ 20.07 mm (HT ♀); 34.9 mm / 19.5 mm (PT ♀). Palps : penultimate segment with whitish stripe close to compound eye, fringed with darker brown hairscales above where in potential contact. A light ochreous stripe away from eye is fringed by brown hair-scales. Mesad face of labial palps whitish. Etymology. Refers in particular to the tornado-like markings, the sporadic rarity of this species in collections, and also alludes to the chance survival of an apparently deeply branching species in cyclone-prone Madagascar . The impression is of the discovery in Madagascar of a relictual member of a greatly diversified lineage (Aduse- Poku et al. , 2015, Fig. 1 ). In different endemic radiating mycalesine clades in Madagascar , there is a trend towards reduction of the number of ocelli expressed in the upperside HW, so its configuration might also be correlated with a typically more open habitat, such as the marshy areas typical for Ht. turbata . Discussion. This species was first discovered by myself in an uncurated drawer of the Oberthür collection (as it was arranged at BMNH in May 1994 ), with a unique ♂ and ♀ labelled as from ‘Mahassabe’ (see Referred specimens), collection dating pre-1923 ( Lees, 1997: 65 ). Because of the distance of the locality of these specimens from recent material and the uncertainty of the labels (see also above under Ht. sogai ) they are not actually included in the type series. The exceptional rarity of this distinctive species, despite intensive prior sampling from the 1890s to the present in the Fianarantsoa (?) and Marojejy-Anjanaharibe Sud areas, is noteworthy. Perhaps this species has naturally small populations, or normally occupies a particular niche that has not been properly sampled and difficult of access, such as open cliff slopes, like the spectacular habitat facing ‘Camp II Marojejya’. Mahasoabe (‘Mahassabe’; Fig. 30 D, lower brown dot) is a village located at around 1100 m , and its surroundings are largely deforested today, but the nearest primary forest is over 20 km to the east, and assumed plantations or secondary forest can be viewed on Google Earth from 4 km to the west. The corresponding printed label is perhaps anyway as mentioned before unreliable, considering that that the same label is placed on series of characteristically northern and lowland species like Ht. erebina . The species is distinctive enough to be unmistakable, but all available types in the Ht. subsimilis and Ht. strigula groups were examined. In particular I checked the STs of Culapa ornata Oberthür, 1916 (which was synonymised by d’Abrera, 1980 : 182 with Ht. turbata (Butler, 1880) ; HT ♀: ‘Madagascar’, Cowan, BMNH(E) 673769). The LT ♂ of Culapa ornata ( Fig. 19 E, representing Ht. turbata ), here designated: BMNH(E) 673678, Madagascar ; PLT ♂, Antananarivo, R. Toy: BMNH(E) 673670)) and syntypes of Culapa pallida Oberthür, 1916 (LT ♂, missing abdomen, Fig. 19 D, here designated, bearing label: “ Madagascar , Antsianaka, Perrot Freres, 2e Semestre 1890”|BMNH(E) 673775”; PLT ♂, BMNH(E) 673779 and ♀♀ BMNH(E) 673776 (illustrated in Oberthür, 1916 , Pl. 356: f. 3054), BMNH(E) 673778 and BMNH(E) 673780, all from Madagascar , Antsianaka, Perrot Freres, 2e Semestre 1893; there are also two potential ST specimens of C. pallida not labelled ‘Antsianaka’), were also examined. Also, considering a superficial resemblance, the syntypes of Mycalesis iboina Ward, 1870 in BMNH (Coll. Ward) (LT ♂, here designated: BMNH(E) #674853, PLTs, BMNH(E) #674854-674856) and MNHN (Coll. Galichon) (PLT ♂♂) were examined; also the “HT” ♂ of Culapa parva Butler, 1879 (BMNH(E) #674859, ex Grose-Smith) and the ♂ and ♀ STs of its synonym ( Lees, 1997 ; Lees et al. , 2003), Mycalesis irrorata Mabille, 1880 (LT ♂, here designated: BMNH(E) #674860; ex Coll. Grose-Smith; PLT ♀, BMNH(E) #674861, “Madag” ex Coll. P. Mabille). Additional information. ♂ genitalia: 362DL (referred specimen, Fig. 18 A): about 2 mm long; from LV, tegumen with bent down to fairly straight uncus leading to downward hook at tip without a distinct ‘head’. Gnathos from square base narrow and tapered, slightly uprecurved to tip. Hinge with vinculum wider than half maximum dimension of tegumen. Saccus rather short, stubby and bulbous. Aedeagus distinctly longer than valve, fairly stout, uprecurved distad of ostium. Valves fairly short, arms uprecurved and leading to club, which has spinoid setae at the end and a distinct ‘beak’ pointing dorsad/mesad. The overall impression is a similarity to genitalia of some of the Ht. strigula group (valves not symmetrically ‘skittle’-shaped’ from LV as in Ht. subsimilis group and rather uncus is more straight from LV; Lees 1997 : 106, Fig. 7 f). The genitalia of Ht. turbata are actually quite similar to those of Ht. tornado . DNA divergences : COI-5P cluster number BOLD:ACW4939 (exemplar BMAD192-15, DL14A-0245), closest to members of Ht. strigula group, e.g. 6.7% divergent to Ht. barbarae sp. nov and 8.3% to Ht. turbata and a similar distance to Ht. ankova . Phylogeny/sister species : Lees (1997: 156) had over 96% jackknife support for a sister relation of Ht. tornado (“FFTFR”) with the Ht. ankova ‘clade’ of (then three) species, but this resolution appeared only in the total evidence parsimony analysis. The species would therefore be tentatively placed in the Ht. strigula group based on morphological features and support. Molecular sequencing of multiple genes provides evidence Ht. tornado is sister to all others of the Ht. strigula group, including the ‘ ‘ Ht. ankova ’ sub-group (Aduse-Poku et al. , 2015). Ecology and distribution. Behaviour : largely unknown. The HT was found resting on the ground in the morning at the edge of a road close to primary forest; the paratype ♀ was sitting on a path above a steep slope in primary forest, presumably attracted by fallen fruit. Hostplant : unknown. Early stages : unknown. Distribution : Marojejy, Anjanaharibe Sud; (?) Mahasoabe, SE of Fianarantsoa ( Fig. 30 D, dark brown dots). Elevational range : 675 (PT ♀)– 1000 m . (HT ♀) [possibly to ca. 1100 m . at “Mahassabe”]? Habitat : one of the PT ♀♀ was caught in tall canopy primary forest on a path above a steep slope and small stream, about 100 m above the Manantenina river at Marojejy. The HT ♀ was caught on a slope at the edge of a currently largely impassable road, close to an island in the road and next to some primary vegetation and ginger, above a steep stream draining into marshland bordering the Marolakana River which was only around 0.15 km away. No other individuals could be found despite repeated searches in these surrounding habitats. Referred specimens. ♂ [ BMNH ], Mahassabe [Mahasoabe], forêt au S.-E de Fianarantsoa Madagascar , ex Lamberton, 1922|Ex Oberthür Coll. Brit. Mus. 1927-3|362 DL ; ♀ [ BMNH ], same data but BMNH (E) #674881.