Very small moths, wingspan between 3 and 5 mm, usually with a pattern of metallic-silvery spots, but in some species not metallic, typically comprising a dorsal and costal spot at 1/4 sometimes united into a fascia and a postmedial fascia, which also may be broken into two spots. In some species part of this pattern is absent. Males never with androconial scales or hair-pencils. Separated from externally somewhat similar Antispila species by the reduced venation (Figs 29–34); in Antispila the discoidal cell is present and more veins are retained (Fig. 35); also most Antispila species are larger and have more antennal segments. Separated from most Heliozela species by more extensive colour pattern, the few Heliozela that do have more spots can be separated also by the venation with discoidal cell (Fig. 36), also Heliozela species have more antennal segments. Species of Antispilina and some in the “Antispila” ampelopsifoliella group have a very similar venation and are more difficult to separate; most Holocacista differ in the male genitalia by the usually long appendix on the phallus, moreover they have a small epiphysis, which is absent in the other genera with reduced venation.

Adults. Very small moths, forewing length ca. 1.5–2.5 mm (wingspan ca. 3–5 mm).

Head (Figs 21–24). Almost oval in outline. Eyes in latero-ventral position, ventral margin not reaching lower margin of head. Eye phragma narrow, weakly melanised. No sutures present. Anterior tentorial arms very slender, prominently curved laterally before converging towards frons. Vestiture comprising lamellar scales, firmly appressed on head, in dry specimens scales on vertex sometimes raised, probably an artefact as a result of drying. Mouthparts: labrum narrow, pilifers absent. Mandibles small, as long as broad, relatively well sclerotised (Fig. 23). Maxilla with galea well developed and longer than head; maxillary palp reduced to a single segment. Labial palp well developed, 3-segmented, drooping, slightly shorter than head capsule; distal segment from slightly longer to 1.5× second segment; depression for Organ von Rath not seen. Antenna (Fig. 24) ca. half length of forewing with 14–20 segments [best counted in denuded specimens on slides], no sexual dimorphism. Scape and pedicel of equal length, slightly shorter than flagellomeres. Flagellomeres cylindrical, longer than wide, each with two annuli of scales, often differently coloured, resulting in visible dark and pale rings from above. Pecten present, but not easily visible; Nielsen (1980) counted 4–6 hairs in H. rivillei.

Thorax. Vestiture of appressed lamellar scales, usually concolorous with ground colour of forewings. Foreleg with small but distinct epiphysis of about 36–48 μm in H. rivillei (Fig. 26) to 50–82 µm in H. capensis (Fig. 25), distinctly smaller than in Heliozela (Fig. 27) (150–165 μm in two measured European specimens) and without the microtrichia along the inner side, which probably serve as antennal cleaning apparatus. In Antispila, Antispilina (Fig. 28) and Coptodisca epiphysis completely lacking.

Wings. Male retinaculum a series of 7–12 hook-shaped bristles, arising from a thickened serrate portion of Sc. Frenulum in male a strong curved bristle (e.g., Fig. 30), in female two bristles present (Figs 29, 34); no pseudofrenular bristles in male. Humeral field with scattered microtrichia, otherwise restricted on wing membrane to area just posterior of retinaculum, arranged in longitudinal rows. Scale sockets regularly spaced, not in distinct rows.

Venation in forewing (Figs 29–34) with Sc to middle of costa. R unbranched, a separate vein, to costa. Rs with M and CuA, ending in 3–4 branches, interpreted as Rs1+2 to costa, Rs3+4 to termen, M and CuA to dorsum; in one species, H. sp. _Terminalia_SA, tentatively placed here, Rs+M+Cu even more reduced, with only 2 branches. Hindwing with Sc+R to costa, Rs+ M with 2–3 branches, Rs to costa, 1 or 2 branches of M to termen and dorsum; CuA a separate vein to dorsum.

Wing pattern (Figs 1–16). On forewing typically comprising a pair of opposite pale, often metallic spots at 1/3 and a similarly coloured fascia or opposite spots at 2/3 on a dark background, brown to black, or brassy and shining. Variation exists in relative size, whether spots are joined to a fascia, or fascia is broken, or spots may be absent (e.g., Figs 11, 14, 16). Many species show sexual dimorphy in pattern, with females having more or larger pale elements than male. Only a single species from Arabian Peninsula has a different pattern with much yellow, probably as an adaptation to the desert habitat (Fig. 13). A fringe line often available, with fringe scales pale. Hindwings uniform grey. Androconial scales absent in all species examined.

Pregenital abdomen. Abdominal sclerites weakly sclerotised. Anterior sternum II subtriangular, free.

Male genitalia. Vinculum (S IX) very long, anteriorly often reaching beyond anterior margin of segment VI, almost cylindrical; tegumen (TIX) narrow, usually with a medial posterior process, probably a composite structure with uncus. Gnathos absent. Valva rather narrow, with stalked pectinifer halfway to inner margin, pecten comprising 6–12 blunt sensilla; transtilla typically with medial anterior projection, sublateral processes long. Phallocrypt (manica) with some to many strongly-sclerotised conical spines, often arranged in an asymmetric fashion, or with many smaller spines. Phallus outer tube often with remarkable ventrally-curved appendix on phallus, or appendices of different sizes and shapes. Juxta present and often bilobed or reduced to narrow ventral process.

Female genitalia. SVIII pointed, T VIII deeply indented. Oviscapt with few lateral cusps. Anterior and posterior apophyses subequal in length. Spermathecal papilla usually with circular sclerotisation. Ductus spermathecae with many coils.

Larva. Larvae yellow or whitish, usually with darker head capsule. Larva of H. rivillei described in detail by Grandi (1931) and Marchi (1956). Head prognathous, legs and prolegs absent, but paired ambulatory calli on T2 and 3 (ventral and dorsal) and fused ventro-medial – calli on A3–6. Larvae with four feeding instars and a fifth non-feeding instar that constructs the case in which it pupates.

Hostplants. Several species feed on Vitaceae and Rubiaceae, a few species on Anacardiaceae, and single species each on Balsaminaceae, Dilleniaceae, Geraniaceae and Plumbaginaceae. A species feeding on Combretaceae is tentatively added, but this requires confirmation.

Life history. Eggs are inserted in leaf tissue, often near a vein or leaf margin. All species construct leafmines (Figs 70–75, 79, 83–93), usually starting as a narrow linear mine, later usually widening into a blotch, or sometimes remaining an irregularly wide gallery, and cut out an oval shield, comprising the epidermal layers, during the penultimate instar. Frass is deposited in a central line in the mine or filling the mine, later often scattered in the blotch or pushed by the larva to one side. The shields (Figs 76–78), later forming the cocoons, are more or less flat, without the raised ridge that is characteristic for Antispila. They attach this cocoon to any surface (trunks, leaves, leaf litter, etc.) where the non-feeding final instar larva pupates. Adults are usually day flying (Figs 80–82), and rarely come to light.

Mainly Old World tropics and subtropics: Afrotropical, Oriental and Australian regions, north to Taiwan and southern Europe (type species). Some DNA barcodes suggest that the genus also occurs in South and Central America, but no adults have yet been studied from this area.

The species listed in the checklist below, both named ones and unnamed ones, share the external and venation characters described above, and those dissected also the male genitalia characters. Those species that we have been able to sequence form a well-supported clade in a phylogenetic analysis (both Bayesian and Maximum Likelihood) of the Heliozelidae based on four genes (unpublished study in progress), as part of a larger clade of genera with reduced venation (including also Antispilina Hering, 1941, Coptodisca Walsingham, 1895 and the “Antispila” ampelopsifoliella group). While checking several Indian species described by Meyrick, we could also change the following generic assignment: Heliozela anna (Fletcher, 1920), comb. n. (from Antispila, feeding on Myrtaceae) (Figs 20, 36), whereas the following remain in their original genus: Antispila argostoma Meyrick, 1916 (Figs 18, 35) and A. aristarcha Meyrick, 1916 (Fig. 19), both feeding on Vitaceae. While we assign an unnamed species feeding on Impatiens from Vietnam to Holocacista here (Figs 16, 34), the Indonesian Microplitica metadesmia (Meyrick, 1934) that likewise feeds on Impatiens, has a completely different venation, more similar to Heliozela Herrich-Schäffer, 1853. Pending further study of this species and the type species of Microplitica Meyrick, 1935 (Microplitis desmophanes Meyrick, 1922), we leave it in Microplitica for now.

In the checklist below we provide the original genus in brackets, type locality, and the hostplant of the types. The species are listed geographically, first the named ones, then the unnamed ones.

Palearctic species

H. rivillei (Stainton, 1855): p. 89 (Elachista)

Malta, Vitis vinifera L. [type species]

African species

H. capensis van Nieukerken & Geertsema, sp. n.

South Africa, Western Cape, Paarl, Vitis vinifera L.

H. salutans (Meyrick, 1921): p. 108, comb. n. (Antispila)

South Africa, [Kwazulu Natal], Durban, [Rhoicissus sp.]

H. varii (Mey, 2011): p. 156, comb. n. (Antispilina)

South Africa, Western Cape, Cape Town, Pelargonium cucullatum (L.) L’Hérit.

Asian species

H. micrarcha (Meyrick, 1926): p. 261, comb. n. (Antispila)

India, [Karnataka], Karwar, Lannea coromandelica (Houtt.) Merr. (=Odina wodier Roxb., Anacardiaceae)

H. pariodelta (Meyrick, 1929): p. 541, comb. n. (Antispila)

India, Bihar, Pusa, Lannea coromandelica (Houtt.) Merr. (=Odina wodier Roxb., Anacardiaceae)

H. selastis (Meyrick, 1926): p. 261, comb. n. (Antispila)

India, [Karnataka], Karwar, Psychotria dalzellii Hook.f. (Rubiaceae)

Unnamed species:

Palearctic species

sp. Dyerophytum_UAE

United Arab Emirates, Fujairah, Dyerophytum indicum (Gibbs ex Wight) Kuntze (Plumbaginaceae)

African species [see also Appendix A]

sp. Rhoicissus_tridentata

South Africa, Rhoicissus tridentata (L. f.) Wild & R. B. Drumm. subsp. cuneifolia (Eckl. & Zeyh.) Urton, R. tomentosa (Lam.) Wild & R.B.Drumm. (Vitaceae)

sp. Rhoicissus_tomentosa

South Africa, Rhoicissus tomentosa (Lam.) Wild & R.B.Drumm. (Vitaceae)

sp. Rhoicissus_PundaMilia

South Africa, Rhoicissus digitata (L.f.) Gilg. & M.Brandt (Vitaceae)

sp. Cissus_integrifolia

South Africa, Cissus integrifolia (Baker) Planch. (Vitaceae)

sp. Lannea_SA

South Africa, Lannea discolor (Sond.) Engl. (Anacardiaceae)

sp. Terminalia_SA [placement tentative]

South Africa, Terminalia prunioides M.A. Lawson (Combretaceae)

Asian species

sp. Leea_Borneo

Indonesia, Kalimantan Timur, Leea indica (Burm.f.) Merr. (Vitaceae)

sp. Impatiens_Vietnam

Vietnam, Cuc Phuong NP, Impatiens clavigera Hook. f. (Balsaminaceae)

sp. Lasianthus_Borneo

Indonesia, Kalimantan Timur, Lasianthus Jack sp. (Rubiaceae)

sp. Lasianthus_Sabah

Malaysia, Sabah, Lasianthus Jack sp. (Rubiaceae)

sp. Paedaeria_Taiwan

Taiwan, Paederia foetida L. (Rubiaceae)

Australian species

sp. Psychotria_Australia

Australia, Queensland, Psychotria simmondsiana F.M.Bailey (Rubiaceae)

sp. Morinda_Australia

Australia, Queensland, Morinda jasminoides A.Cunn. (Rubiaceae)

sp. Hibbertia_Australia

Australia, West Australia, Hibbertia Andrews (Dilleniaceae)

The type species of Holocacista has recently been diagnosed in the context of a study of European (and North American) Vitaceae miners (van Nieukerken et al. 2012b), when the genus was still considered monotypic. Here we briefly diagnose it against other species in the genus, without a full redescription. It should be noted that only material from Italy and Bulgaria has been examined in detail. Morphological details (Figs 21, 23, 24, 26) and venation (Fig. 29) are described under the generic treatment. For a full synonymy we refer to van Nieukerken et al. (2012b).

(Fig. 9). Wingspan 4.0–4.5 mm. Antenna ringed, 15 segments; head and thorax bronze grey. Forewing fuscous to black, with golden silvery pattern consisting of four spots, costals distal to dorsals, the first costal and dorsal sometimes united as oblique fascia; a distinct fringe line, fringe silvery white. Differs from South African H. capensis and H. salutans by more golden shining spots and distinct first costal spot (almost absent or reduced in other species); from H. varii and other species by distinct fringe line and ringed antenna.

Male genitalia (Fig. 105, and Figs 48–50 in van Nieukerken et al 2012b). Total length vinculum + tegumen 630–720 μm, phallus 575–630 μm. Pecten with 8–10 teeth. Juxta more elaborate than in South African Vitaceae miners, deeply bifurcate. Phallus without spines on phallocrypt, wrinkled. Female genitalia illustrated by van Nieukerken et al (2012b, Figs 51–52) and Cean (2014).

Host plants. Vitaceae: Vitis vinifera, wild and cultivated, possibly also on cultivated Parthenocissus Planch. (new record from Russia, Kalmykiya).

Leafmines (Figs 87, 88). The egg is inserted usually close to a major vein, probably on leaf underside. The mine is first a gallery, turning from once to several times around the oviposition site and then extends, often along a vein as a rather straight linear mine, occasionally as a serpentine mine; distally enlarging into a small blotch. The frass is black forming a broken line, often not exactly in the middle of the mine; in thicker leaves it may be wider; in the blotch the frass is dispersed; the larva cuts out a case of about 3.3–4 mm × 2.0–2.5 mm.

Widespread in southern Europe, Turkey and Central Asia (van Nieukerken et al. 2012b), now also recorded in Romania (Cean 2014). Probably only native in eastern part of its current distribution area.

Adults and leafmines: Bulgaria: 1♂, 7 adults [sex not determined], Sliven, 5.iv.1928, P. Tschorbadjiev, Genitalia slide JCK7867 (coll. Natural History Museum Sofia). Italy: 28 adults (4♂, 1♀ dissected), Vicenza, Borghetto, experimental vineyard, leafmines on Vitis vinifera, 2007, emerged i–ii.2009, M. Baldessari (RMNH); 1♂ (dissected), 7♀, many leafmines, larvae, ibidem, 19.viii.2013, EvN2013904, emerged 11.ix–3.x.2013, M. Baldessari (RMNH). Russia: leafmines only, Kalmykiya, Elista, Citypark, 26.ix.2000, leafmines on Parthenocissus, V. Zolotuhin (coll. Zolotuhin).

Holotype male, South Africa (Western Cape), Paarl NW, De Heuvel estate, 180 m, 16.i.2013, leafmines on Vitis vinifera cv ‘Regal’, EvN2013004, emerged 27.i.2013, E.J. van Nieukerken & H. Geertsema, Genitalia slide EvN4622, DNA extracted (RMNH.INS.24622) (RMNH).

Externally Holocacista capensis is almost inseparable from other South African Vitaceae-feeding Holocacista species, including H. salutans. Absence or reduction of the first costal spot in the male, however, may be an indication that the specimen might be H. capensis; only study of genitalia allows a firm identification. For differences with H. varii, see there. The only South African Vitaceae-feeding “real” Antispila species is much larger and has more antennal segments (ca. 26). In male genitalia the configuration of the small number of spines on phallocrypt in combination with the ventrally curved phallus appendix is characteristic, otherwise very similar to H. salutans and some of the unnamed Rhoicissus miners. Leafmines characterised by the very contorted first part of the mine, which is straighter or shorter in the other species; currently the only known leafminer on Vitis in South Africa.

Male (Figs 1, 3). Head face and vertex covered with appressed, metallic, silvery-white scales, more brownish grey on vertex. Palpi porrect, white; base of proboscis covered with white scales. Antenna with 16 segments, ringed, each flagellomere with a basal fuscous scale ring and apical white scale ring on upper side, scales on underside all white. Legs grey, tarsi mostly yellowish white, especially on underside. Thorax and forewings ground colour grey brown, slightly irrorate, caused by scales being dark tipped and paler at base. A silver-white pattern on forewing consists of a triangular dorsal spot at 1/4, usually associated with a minor spot of just a few scales at costa, that may be joined to dorsal spot, or even completely absent; a second triangular dorsal spot at 1/2, reaching almost to middle of wing; a triangular costal spot just beyond middle, always separate; fringe line very distinct, demarcated by dark-tipped scales. Terminal fringe silvery white. Hindwings pale grey. Underside of wings fuscous, with white spots visible. Abdomen lead coloured, including vestiture on external genitalia.

Female (Fig. 2). Antenna with 16 segments. Colour pattern distinct from male: scales almost uniformly dark fuscous with purplish tinge, resulting in darker, velvety wing colour and contrasting silvery-white pattern; first costal and dorsal spots joined to form a narrow fascia, wider at dorsum; second dorsal and costal spots as in male; fringe line distinct, scales forming cilia line with slightly paler bases. Abdomen almost black, narrowly pointed posteriorly.

Measurements. Male: forewing length 1.8–2.3 mm (2.0 ± 0.1, 20) (1 dwarf of 1.55 mm forewing length excluded), wingspan: 3.9–4.9 mm. Female: forewing length 1.9–2.1 mm (2.0 ± 0.1, 14), wingspan 4.0–4.6 mm.

Male genitalia (Figs 37–49, 94–96, 107–110, 115). Total length vinculum + tegumen 425–625 µm. Vinculum (S IX) long, reaching anterior margin of segment VI. Tegumen (Figs 41, 46, 110) well sclerotised, with medial, slightly-bilobed posterior projection, one sensilla on each lobe; tegumen dorsally with groups of microtrichia, and two lateral lobes with setae or sensilla; a poorly-sclerotised structure below tegumen may be a reduced uncus. Valva (Figs 40, 43, 108) narrow, apex blunt, with stalked pectinifer halfway to inner margin, pecten comprising 8–11 blunt sensilla, usually same number on both valvae, but sometimes a difference of one. Valva length (without transtilla) 200–230 µm. Transtilla (Figs 41, 109) with long sublateral processes and medial spatulate posterior process, with rounded corners. Juxta elongate, as a narrow ventral process of phallus, attached on phallus near phallocrypt spines (Fig. 48). Phallus (Figs 39, 42, 44, 45, 48, 49, 94–96, 107) long and narrow, ca. 340–425 µm long. Phallocrypt (manica) with some strongly-sclerotised conical spines, arranged asymmetrically; in lateral view (Figs 44, 94–96) one dorsally, curved ventrad, a similar strong one ventrally curved dorsad, latter with 3–4 smaller spines in a row anteriorly; in ventrally mounted specimens spines appear mostly on right side, where phallus is constricted. Phallus outer tube with ventrally-curved appendix ca.103–150 µm long (measured along curve).

Female genitalia (Figs 50–53). Length of anterior apophyses 800–900 μm (n=5), posterior apophyses 880–935 μm (n=5). Oviscapt with 5–6 cusps on either side (Figs 51, 52). Ductus spermathecae with many wide convolutions, spermathecal papilla with circle-shaped sclerotisation (Fig. 50).

Host plants. Vitaceae: Rhoicissus digitata (L.f.) Gilg. & M. Brandt and various South African grown cultivars of Vitis vinifera (e.g., Chardonnay, Chenin Blanc, Red Globe, Régal).

Leafmines (Figs 70–75, 79). The egg is inserted on the leaf underside, usually within 1–2 mm from a vein, rarely slightly farther. Freshly expanded foliage is preferentially selected for oviposition, but as egg laying proceeds from early spring to late autumn, it also oviposits on older leaves, even those showing previous feeding. The majority of the mines on Vitis (75% of 160 mines from six samples) start at the leaf edge, but even there the egg is always near the vein in the tip of a lobe; some mines originate close to the leaf midrib. Also, the few studied mines on Rhoicissus start at the leaf(let) tip. The mine starts as a much contorted narrow gallery, often first in a zigzag pattern with U-turns, eventually enlarging into an irregular wide gallery or a blotch. The frass is brown in the early mine, later black, in a rather thin line in the centre of the gallery; later the frass is in clumps in a wider central line. The whole mine occupies a small area of ca. 12–15 mm long, of which the size depends on leaf thickness; in thin leaves mines are appreciably longer and wider. Mines are very often clustered in groups of 3–5 or even more. The larva cuts out an elliptic case of about 2.5–4.1 mm (3.4 ± 0.3, n=34) × 1.5–3.1 mm (2.3 ± 0.3, n=34) mm wide, ratio 1.2–1.8 (1.5 ± 0.1).

The moth is multivoltine; the first adults appear during early spring (September to October) and a single generation lasts from three to four weeks; peak numbers are reached during February and March at the height of the grape picking season. Moths are still present in April; the last were seen early May; many cocoons overwinter in leaf litter, dropping to the ground and pupating amongst leaf litter or attached to stems and trellises from April onwards, and yielding moths from September onwards. Larvae are present almost continuously from November to early May when the leaves start to wither and drop. Larvae have only once been collected on Rhoicissus, these in March. When fully grown, larvae descend from the mines to attach their cocoons upon landing on a variety of objects such as other leaves, berries of grape bunches, trellises or on the bark of the vine itself (Fig. 76).

Moths aggregate and mate in the heat of the day (1100–1400 hrs) on exposed vine foliage, but prefer to oviposit in the shaded canopy conditions under which table grapes are grown; wine grapes, grown in an open cultivation system and fully exposed to the sun are rarely, or at least less seriously attacked.

(Fig. 115). On native Rhoicissus as yet only found once: South Africa, Western Cape (Wilderness). On cultivated Vitis from South Africa: Western Cape, Northern Cape and Gauteng.

We barcoded eight specimens, including the Holotype. All barcodes belong to Barcode Identification Number (BIN):ACG9027, the largest intraspecific distance is 1.4%, between one specimen collected in Gauteng and the rest, collected in the Western Cape.

DNA-Barcode of Holotype, HELA103-14 (658 basepairs):

AACTTTATATTTTATTTTTGGTATTTGAGCGGGATTAGTAGGAACATCAATAAGTTTATTAATTCGTGCTGAATTAGGAATCCCTGGGTCCTTAATTTCTAATGATCAAATTTATAATACTATTGTTACAGCTCATGCATTTATTATAATTTTTTTTATAGTTATACCTATTATAATTGGAGGATTTGGAAATTGATTAGTTCCGTTAATATTAGGAGCCCCAGATATAGCATTTCCTCGTCTTAATAATATAAGTTTTTGACTCCTTCCCCCATCTTTAACATTATTAATTTCAAGAAGATTAGTTGAAATGGGATCAGGAACTGGATGAACTGTCTATCCACCTTTATCTTCCAATATTGCCCATATGGGAACTTCTGTGGATTTAACTATTTTTTCTTTACATTTGGCTGGAATTTCATCTATTTTAGGAGCTGTAAATTTTATTACAACAATTATTAATATAAAACCAGTTAGAATAATATATAATCAACTTTCTTTATTTGTTTGATCTGTGGGTATTACAGCTTTATTACTATTATTATCTTTACCTGTATTAGCTGGAGCTATTACTATATTATTAACTGATCGAAATTTAAATACTTCTTTTTTTGACCCTATGGGAGGAGGAGACCCTATTCTATATCAACATTTATTT

The only wild Rhoicissus on which mines of H. capensis were collected, was identified by Vári in his notebook as R. revoilii. The single leaf we studied could belong to this species or to R. digitata, which is very similar. On the basis of the distribution (Palgrave and Palgrave 2002), we conclude that the latter is the most likely, since R. revoilii is not known to occur in the Western Cape.

Several reared adults were used in 2013 for a rearing experiment on potted plants of Rhoicissus rhomboidea (E.Mey. ex Harv.) Planch., bought in the Netherlands. Although the adults lived for several days, no traces of mines were found. Either the species is unsuitable as a hostplant, or these potted plants contained remnants of insecticides. Later, we were more successful with rearing larval offspring from Vitis-grown adults from Wellington on potted Rhoicissus digitata in the laboratory in Stellenbosch (for resulting leafmines see Fig. 79). The main aim of this preliminary study was to detect whether moths reared on V. vinifera would readily breed on (caged) R. digitata; moths emerging from grapevine leaf litter or sampled foliage were released into the caged Rhoicissus. The latter was readily infested, often resulting in the entire leaf being consumed by the larvae.

Other live cocoons were sent in 2013 to Lund, Sweden, emerged there, and have been used for pheromone studies (Wang et al. 2015).

Adults and leafmines: South Africa, Gauteng: 3♂, Pretoria, Roodeplaat, 1245 m, leafmines Vitis vinifera, emerged 10–12.x.1990, S. Marais, Genitalia slide EvN4264, DNA extracted (RMNH.INS.24264) (HG, RMNH); 3♂, 5♀, same locality, emerged 4–14.iv.2012, D. Visser (HG, RMNH); 2♀ [5 more specimens in TMSA], Pretoria, emerged 2–6.xi.1950, L. Vári, Genitalia slide TM6830, Wing slide TM 2414 (TMSA); 1♀ [1 more specimen in TMSA], 1 herbarium sheet with 6 leafmines on 4 leaves, Pretoria, 8.iii.1953, Ac. no. 660, leafmines on V. vinifera, emerged 10–31.iii.1953, L. Vári (TMSA); 1♂, 1♀, 1 herbarium sheet with ca. 13 leafmines in 4 leaves, Pretoria, in own garden, 21.x.1953, Ac. no. 866, leafmines on V. vinifera, emerged 26.x–3.xi.1953, L. Vári (TMSA); 1♂, larvae and leafmines, Roodeplaat exp. Farm, 1168 m, 23.i.2013, leafmines V. vinifera, EvN2013025–026, E.J. van Nieukerken & S. Richter, 1 larva DNA extracted (RMNH.INS.29586). Northern Cape: 9♂, 14♀ [unmounted], Vaalhartz Research Stn., Jan Kempdorp near Kimberly, 27.ii.1980 [emergence date?], W. v.d. Westhuyzen (TMSA). Western Cape: 2♂, 3♀, Cape Town, Woodstock, cocoons collected on V. vinifera, 26.ii.2012, emerged 5–13.iii.2012, M. Wohlfarter; 1♂, Oudtshoorn, March 1998, on urban vine, H. Geertsema (HG); 81♂, 58♀, Paarl, nr Windmeul, 168 m, leafmines/cocoons on V. vinifera, emerged 1.ii–30.iii.2012, H. Geertsema, Genitalia slides EvN4260♂, 4261♀, 4262♂, 4263♀; complete adults on slide EvN4445♂, 4446♂, 4447♂, DNA extracted (RMNH.INS.24260, 24261, 24262, 24263, 24445, 24446, 24447) (HG, RMNH); 3♂, 2♀, Paarl NW, Nelson estate, 125 m, 15.i.2013, leafmines on Vitis vinifera cv ‘Chenin Blanc’, EvN2013002, emerged 18.i–1.ii.2013, E.J. van Nieukerken & H. Geertsema; 6♂, 10♀, 4 larvae, ibidem, 130 m, leafmines on V. vinifera cv ‘’Chardonnay’, EvN2013003, emerged 24.i–5.ii.2013, Genitalia slide EvN4624♀, DNA extracted (RMNH.INS.24624), larvae RMNH.INS.2956265; 6♂, 11♀, 6 larvae, Paarl NW, De Heuvel estate, 180 m, 16.i.2013, leafmines on V. vinifera cv ‘Regal’, EvN2013004, emerged 20–27.i.2013, E.J. van Nieukerken & H. Geertsema, larvae RMNH.INS.29578–83; 2♂, 3♀, ibidem, leafmines on V. vinifera cv ‘Red globe’, EvN2013005, emerged 26.i–4.ii.2013; 1♂, 3♀, ibidem, 25.i.2013, leafmines on V. vinifera cv ‘Red globe’, EvN2013030, emerged 1–6.ii.2013, E.J. van Nieukerken & H. Geertsema (RMNH, HG); 5 adults, Somerset, 23.ii.2012, cocoons collected on V. vinifera, emerged 27.ii–5.iii.2012, O. Lotter; 1♀, Wellington, emerged 25.xii.2014, leafmines on V. vinifera, L. Torrance (HG). 3♂, 1♀ [8 more specimens in TMSA], 1 leaf with 6 mines, Wilderness, Kaaimans River, 15.iii.1954, Ac. no. 1093, leafmines on Rhoicissus digitata [in notebook Vári as R. revoilii], emerged 4–5.iv.1954, L. Vári, Genitalia slide EvN4381♂, DNA extracted (RMNH.INS.24381) (TMSA).

[leafmines and larvae collected, no adults kept in collection]. South Africa, Western Cape: 11♂, 21♀ [reared from 50 cocoons in Lund, Sweden and used for pheromone studies], Paarl NW, De Heuvel estate, 180 m, 25.i.2013, leafmines on Vitis vinifera cv ‘Regal’, EvN2013029, emerged 2–15.ii.2013, E.J. van Nieukerken & H. Geertsema; several adults, Wellington, emerged xii.2014, ex V. vinifera laboratory bred on R. digitata, L. Torrance (HG).

Externally H. salutans hardly differs from H. capensis, but the male usually has a costal spot at 1/3 from base, albeit very small. The only consistent characters to separate it from H. capensis are in the male genitalia: the row of larger spines dorsally on the phallocrypt, whereas H. capensis has a row ventrally and just a single spine ventrally; also the shape of the transtilla H. salutans differs from that in capensis. The leafmines of H. salutans have the gallery mine with wider frass, more clumped and not zigzag as in H. capensis.

Male (Fig. 4). Head: face and vertex covered with appressed, metallic, silvery-white scales, more brownish grey on vertex. Palpi porrect, white; base of proboscis covered with white scales. Antenna with 16 segments, ringed, each flagellomere with a basal fuscous scale ring and apical white scale ring on upper side, scales on underside all white. Legs grey, tarsi mostly yellowish white, especially on underside. Thorax and forewings ground colour grey brown, slightly irrorate, caused by scales being dark tipped and paler at base. A silver-white pattern on forewing consists of a triangular dorsal spot at 1/4 from base, a smaller spot at costa, sometimes joined to dorsal spot as a narrow fascia; a second triangular dorsal spot at 1/2, reaching almost to middle of wing; a triangular costal spot just beyond middle, always separate; fringe line very distinct, demarcated by dark-tipped scales. Terminal fringe silvery white. Hindwings pale grey. Underside of wings fuscous, with white spots visible. Abdomen lead grey, including vestiture on external genitalia.

Female. Antenna with 16 segments. Colour pattern distinct from male: scales almost uniformly dark fuscous with purplish tinge, resulting in darker, velvety wing colour and contrasting silvery-white pattern; first costal and dorsal spots always joined to form a narrow fascia, wider at dorsum; second dorsal and costal spots as in male; fringe line distinct, scales forming cilia line with slightly paler bases. Abdomen almost black, narrowly pointed posteriorly.

Measurements. Male: forewing length 1.7–2.3 mm (2.0 ± 0.2, 6), wingspan: 4.0–5.0 mm. Female: forewing length ca. 2.0 mm (n=3), wingspan ca. 4.5 mm.

Male genitalia (Figs 54–58, 97–100, 111, 113). Total length vinculum + tegumen ca. 460–490 µm (n=3). Vinculum (S IX) long, reaching anterior margin of segment VI. Tegumen and uncus well sclerotised, with two medial projections, probably representing tegumen and uncus, dorsalmost projection very similar to tegumen of H. capensis, ventral one truncate, slightly excavated posteriorly, with serrate margins. Valva narrow, apex blunt, with stalked pectinifer halfway along inner margin, pecten comprising 8–10 blunt sensilla. Valva length (without transtilla) ca. 165–215 µm. Transtilla with long sublateral processes and medial spatulate posterior process, with produced lateral corners (Fig. 111). Juxta (Fig. 56) elongate, as a narrow ventral process of phallus, attached to phallus near phallocrypt spines. Phallus (Figs 56, 97–100) long and narrow, ca. 390–430 µm long. Phallocrypt (manica) with two rows of strongly-sclerotised conical spines, arranged symmetrically; in lateral view seen dorsally, all curved ventrad, more than 6–7 spines in a row; a group of small spines posterior to these. Phallus outer tube not constricted, with ventrally-curved appendix of ca. 105–125 µm long (measured along curve).

Female genitalia (Figs 65, 66). Length of anterior apophyses 850 μm (n=1), posterior apophyses 890 μm (n=1). Oviscapt not yet studied in ventral view. Ductus spermathecae with many wide convolutions, spermathecal papilla with circle-shaped sclerotisation (Fig. 66).

Host plants. Vitaceae: Rhoicissus digitata (L.f.) Gilg. & M. Brandt, R. revoilii Planch., R. tomentosa (Lam.) Wild. & R.B. Drumm. and Cissus cornifolia (Baker) Planch. Records from R. tridentata (L. f.) Wild. & R. B. Drumm. subsp. cuneifolia (Eckl. & Zeyh.) Urton require confirmation (see below).

Leafmines (Figs 85, 86). The egg is inserted on the leaf underside, usually close to a vein; some mines start at the leaf edge. The mine starts as a much contorted narrow gallery with all convolutions close to each other, hardly leaving leaf tissue between them. Later, the mine enlarging into an irregular wide gallery or a blotch. The frass is black throughout, clumped and almost filling the gallery, but with space between the clumps. Mines are very often clustered in groups. The larva cuts out an elliptic case of about 3 mm long and 2 mm wide.

Voltinism. Larvae have been found from March to June, in September and again from December to January; adults usually emerge between 3–8 weeks after collecting of leafmines; probably multiple overlapping generations.

(Fig. 115). South Africa: KwaZulu-Natal, Limpopo and Zimbabwe: Masvingo. Records from Gauteng (Pretoria) need confirmation, several leafmines from Rhoicissus tridentata resemble those of H. salutans on other hosts, but we have yet no proof from adults that they are this species.

Meyrick (1921) described Antispila salutans from five specimens from “Natal, Durban, in October (Janse).” They were part of a much larger series, of which the labels in the Ditsong Museum (former Transvaal Museum) give more information: all from Durban, collected by v.d. Merwe, Coll. Janse. Some are dated 10.10.18 [in hand], others x.19 or xi.19 [in print], those dated x.19 also have a label with the text “Ac. n. 453” [Accession number 453]. Many specimens have a cocoon added, showing that they have been reared. The five specimens in London are merely labelled (in Meyrick’s hand) “Durban, Natal, AJTJ, 10 [or 11].19”. Meyrick usually replaced original labels of specimens that he kept for his own collection with shorter ones in his own hand or print (Clarke 1955). Meyrick (1921) wrote in the introduction of his paper: “The types of those new species received from Mr A. J. T. Janse are contained in his collection, ….” Three specimens of Antispila salutans are placed in the type collection in Pretoria under the type numbers 109–111. In addition to the locality label, they have a type label in red ink, with name and type number, and an additional label, printed in black, with a 4-digit number, split in two rows. Such labels are always attached to Janse specimens studied by Meyrick, but no registry book is available in Pretoria with more information (Martin Krüger, personal communication). So if we regard these as real syntypes, even though one of these was labelled with ‘xi’, thus from November, only two of the specimens in London can be regarded as the remaining syntypes, and all other specimens are just topotypical, but not types (see also Razowski and Krüger 2007). For the time being we cannot make a decision as to which specimens in London are actual syntypes, and suggest that the male specimen in the type collection in Pretoria with “Type No. 109” is probably the best candidate to be selected as Lectotype, preferably during a full revision. For now a Lectotype selection does not seem necessary.

Unfortunately, we have not been able to find information on the rearing and hostplant of van der Merwe’s series. The Accession number 453 on some labels had previously been misinterpreted as a number of Vári, who labelled all his reared material with such numbers [probably following up on Janse’s system, but with new numbers]. The leaf with mines that was pinned in the salutans box belongs to Bridelia cathartica Bert. (Euphorbiaceae), has probable Coleoptera mines, and has Vári’s number 453. Obviously, this has nothing to do with the heliozelid. There are no notebooks of Janse left that could shed light on this number (Martin Krüger, personal communication).

Unfortunately, we did not find recent material of this species and are therefore as yet unable to give the DNA barcode.

Syntypes. South Africa, KwaZulu-Natal: 1♂, “DURBAN / 10.10 / v.d. Merwe 18/ Coll. Janse” [black print, date in hand, black cadre]; “24 / 93” [black print]; “Antispila /salutans /Type No. 109”[hand, red ink, “Type No” in print]. 1♂, “DURBAN / 10.10 / v.d. Merwe 18/ Coll. Janse” [black print, date in hand, black cadre]; “24 / 93” [black print];”Antispila /salutans /Cotype No. 111” [hand, red ink, “Cotype No” in print]. 1 adult, “DURBAN / v.d. Merwe xi.19/ Coll. Janse” [black print, black cadre]; “29/ 28” [black print]; “Antispila /salutans /Cotype No. 110” [hand, red ink, “Cotype No” in print]. 4 specimens including 2 possible Syntypes in London, not examined.

Adults and leafmines: South Africa, Kwazulu-Natal, 2♂, 1♀, Durban, emerged x and xi.1919, van der Merwe [ex coll. Janse], genitalia slide TM4023 (♂), wing slide TM1585 (♀); 5♂, 3♀, 1 leaf with 8 mines, Jozini Dam [Pongolapoortdam], Lebombo Mts., 14.i.1965, Ac. no 2788, leafmines on Rhoicissus tomentosa, emerged 27.i–5.ii1965, L. Vári, genitalia slides EvN4384 (♂),EvN4668 (♀); 1♂, 2♀, Umhlanga Rocks, 9–16.vi.1968, Ac. no 2944, leafmines on Rhoicissus revoilii, emerged 2–5.viii.1968, L. Vári; 1♂, 1♀, Umhlanga Rocks, 25.iii.1975, Ac. no 3342, leafmines on Rhoicissus revoilii, emerged 10–11.iv.1975, L. Vári, genitalia slide EvN4383 (♂); Limpopo: 1♂, 1♀, 1 leaf with 3 mines, Cyprus Farm, nr. Ofcolaco, 20.ix.1960, Ac. no 2247, leafmines on Rhoicissus tomentosa, emerged 11–13.x.1960, L. Vári; 6 leafmines on 6 leaves, Debengeni, De Hoek, Waterfalls, 15.vi.1954, Ac. no. 1329, leafmines on Rhoicissus revoilii, L. Vári; 1♂, 5 mines on 2 leaves, Louis Trichardt, 17.iii.1964, Ac. no 2693, leafmines on Rhoicissus tomentosa, emerged 31.iii–5.iv.1964, L. Vári. Zimbabwe, Masvingo: 2♂, 11 mines on 11 leaves, Lundi, 22.iv.1956, Ac. no 1916, leafmines on Cissus cornifolia, emerged 1–30.vi.1956, L. Vári; genitalia slide EvN4386 (♂) (all TMSA).

Holocacista varii is the only species similar to H. capensis that occurs probably commonly in the natural habitats near the grape growing areas of Western Cape and thus could potentially be confused with it. It is distinctly larger, and the forewings are more shining bronze than those of H. capensis. Moreover, the male and female have a complete fascia at 1/3 from forewing base that is not narrower at the costa and the antennae are not ringed. In male genitalia, H. varii lacks the larger spines on the phallocrypt, and has a more developed juxta; further, the dorsal row of spines on the tegumen is characteristic and the shape of the transtilla differs. The female genitalia have more elaborate sclerotisations, and the apophyses are longer.

Male (Fig. 5). Head face and vertex covered with appressed, metallic, pale-bronze scales. Palpi porrect, white; base of proboscis covered with white scales. Antenna with ca. 20 segments, uniform bronze brown, scales on underside all white. Legs grey, tarsi mostly yellowish white, especially on underside. Thorax and forewings grey brown with some bronze lustre, with silver-white patterning; an oblique fascia at 1/4, hardly narrower at costa; a slightly triangular dorsal spot at 1/2, not reaching middle of wing; a triangular or squarish costal spot just beyond middle; fringe line not very distinct, demarcating scales not conspicuously dark tipped. Terminal fringe silvery white. Hindwings pale grey. Underside of wings fuscous. Abdomen lead grey, including vestiture on external genitalia.

Female (Fig. 6). Antenna with ca. 19 segments. Colour pattern different from male: scales more uniformly bronze brown, with strong lustre, and contrasting silvery-white pattern.

Measurements. Male: forewing length 2.4–2.8 mm (2.5 ± 0.1, 6), wingspan: 5.0–5.7 mm. Female: forewing length 2.1–2.6 mm (n=3), wingspan 4.5–5.6 mm.

Male genitalia (Figs 59–64, 104, 112, 114). Total length vinculum + tegumen ca. 670 µm. Vinculum (S IX) long, reaching anterior margin of segment VI. Tegumen (Figs 60, 61, 114) well sclerotised, with medial, blunt posterior projection, with several setae; tegumen dorsally with a transverse keel with many strong spines in posterior direction. Valva (Fig. 63) narrow, basally wider, apex blunt, with stalked pectinifer halfway along inner margin, pecten comprising 7 or 8 blunt sensilla. Valva length (without transtilla) ca. 265 µm. Transtilla (Figs 64, 112) with relatively long sublateral processes and medial spatulate posterior process, indented posteriorly. Juxta (Fig. 62) well developed, split into an elongate process ventral to phallus and a furcate process dorsal to phallus. Phallus (Figs 62, 104) long and narrow, ca. 540 µm long. Phallocrypt (manica) slightly spinulose posteriorly, no strong spines present. Phallus in lateral view a distinctly-curved outer tube with ventrally-curved appendix, the latter almost straight, ca. 120 µm long (measured as curve).

Female genitalia (Figs 67–69). Length of anterior apophyses ca. 935 μm (n=1), posterior apophyses 1020 μm (n=1). Oviscapt with 4 or 5 cusps at either side (Fig. 69). Ductus spermathecae with many wide convolutions, spermathecal papilla with circle-shaped sclerotisation, other elaborate sclerotisations in vestibulum.

Host plants. Geraniaceae; most commonly found on Pelargonium cucullatum (L.) L’Hérit., a common plant in Fynbos of Cape Peninsula and the western part of the Western Cape; single records on P. panduriforme Eckl. & Zeyh., P. hispidum Willd. and P. citronellum J.J.A. Van der Walt.

Leafmines (Figs 89–90). The egg is inserted at any place of the leaf underside, usually not far from a vein. The mine starts towards a vein and then follows the vein as a narrow gallery of ca. 2 cm, eventually rather suddenly enlarging into a more or less triangular or elongate blotch, usually after the mine makes a turn of 180°. In thin leaves the blotch can be very elongated. The frass in the early mine is a narrow black line, in the blotch the frass is typically clumped near the entrance. Mines occur either singly or with a few together on one leaf. The larva cuts out an elliptic case of about 3.4–3.7 mm long × 1.6–2.2 mm wide. The larva probably descends with its cocoon into leaf litter before pupation.

Voltinism. Larvae are found between mid-September and April, and are apparently absent in winter; adults usually emerge between 3–6 weeks after collecting the larvae, suggesting there are multiple overlapping generations. The specimen from Worcester (October) is the only record of an adult taken in the wild.

(Fig. 116). South Africa: Western Cape, Eastern Cape (new record). The species is abundant on Table Mountain and in the Cape Peninsula, but it is here also recorded from other Fynbos localities around Stellenbosch and Worcester, and Vári also reared it from Zuurberg Pass in Eastern Cape, so we assume a wide distribution in Western and Eastern Cape.

We barcoded four specimens (three from Table Mountain, one from Stellenbosch), with a maximum intraspecific distance of 1.55%, within the same population on Table Mountain. The BIN is BOLD:ACG8941.

In the original description Mey (2011) mentioned that the valval pecten has 13–15 spines, a number that we cannot confirm from the few specimens studied. He further considered the phallus appendix as a cornutus. This appendix, however, is not attached to the vesica, but is an unmovable extension of the phallic tube. Mey (2011) placed this species in Antispilina on the basis of its venation, overlooking the fact that Holocacista has the same venation. Antispilina is a small Palearctic genus, feeding on Polygonaceae, with some new species in the course of description (B.W. Lee et al. in preparation). The placement of this species in Holocacista, which is diverse in Africa, makes much more sense. Possibly the ancestor of H. varii shifted hosts from Vitaceae to Pelargonium.

Many of the specimens listed under material have only been briefly examined by EvN during his visit to the Ditsong Museum, hence the absence of indication of sex.

Eastern Cape: 1 adult, Zuurberg Pass, south slopes, 22.iii.1954, 11 mines, Pelargonium sp., 1 adult emerged 20.iv.1954, L. Vári (TMSA). Western Cape: 5 adults, Bloubergstrand, 3.x.1974, Ac. no. 3308, leafmines on Pelargonium, emerged 26–30.x.1974, L. Vári (TMSA); 2 adults, ibidem, 13.x.1975, Ac. no. 3496, leafmines on Pelargonium, emerged 14.xi.1975, L. Vári (TMSA); 6 adults, Cape of Good Hope Nature Reserve, 26.x.1966, Ac. no. 2851, leafmines on Pelargonium, emerged 21–23.xi.1966, L. Vári (TMSA); 8 adults, ibidem, Ac. no. 288527.x.1967, emerged 20–22.xi.1967, L. Vári (TMSA); 4 adults, Cape Peninsula, Bakoven, 29.x.1975, Ac. no. 3499, leafmines on Pelargonium, emerged 17–19.xi.1975, L. Vári (TMSA); 5 adults, Cape Peninsula, Hout Bay, 11.xi.1954, Ac. no. 1357 leafmines on Pelargonium cucullatum, emerged 7–21.xii.1954, L. Vári (TMSA); 2 adults (paratypes), ibidem, 14.ix.1966, Ac. no. 2846, emerged 21.xi.1966, L. Vári (TMSA); 10 adults (paratypes), Cape Peninsula, nr. Muizenberg, Steenberg, 10.xi.1979, Ac. no. 3764, leafmines on Pelargonium, emerged 15.xi–10.xii.1979, L. Vári (TMSA); 1♂, 1♀, 1 adult (paratypes), Cape Town, Kirstenbosch, 17.xi.1954, Ac. no. 1365, leafmines on Pelargonium cucullatum, emerged 11–14.xii.1954, L. Vári (MHUB, TMSA); 1 adult, ibidem, emerged 5–29.xii.1954, A.J.T. Janse (TMSA); 2 adults (paratypes), ibidem, 14.ix.1962, Ac. no. 2535, emerged 1–17.x.1962, L. Vári (TMSA); 1♂, 1♀, ibidem, 23.xi.2014, leafmines on P. citronelium, emerged 11.xii.2014, L. Torrance & H. Geertsema (USEC); 1♂, 1♀, ibidem, 23.xi.2014, leafmines on P. cucullatum, emerged 11–14.xii.2014, L. Torrance & H. Geertsema (USEC).

6♂, 4♀ (paratypes), 1 adult, Cape Town, slopes Table Mt., 4.iii.1954, Ac. no. 1047, leafmines on Pelargonium cucullatum, emerged 26.iii–3.iv.1954, L. Vári (TMSA); 9 adults, Noordhoek, leafmines on Pelargonium, pupa 10–14.v.1984, emerged 28–31.v.1984, H. Geertsema (TMSA); leafmines, Stellenbosch, Botanical Garden, 122 m, 27.i.2013, EvN2013032, leafmines on Pelargonium panduriforme, E.J. van Nieukerken (RMNH); 1 larva, ibidem, 27.i.2013, EvN2013033, leafmines on Pelargonium hispidum, E.J. van Nieukerken (RMNH); 3 larvae, Stellenbosch, Jonkershoek, 390 m, 18.i.2013, EvN2013017, leafmines on Pelargonium cucullatum, E.J. van Nieukerken & H. Geertsema (RMNH); 3♂, 4♀, 3 larvae, Table mountain NP, Cecilia, nr Klaasenskop, 385 m, 19.i.2013, EvN2013022, leafmines on Pelargonium cucullatum, emerged 11–22.ii.2013, E.J. van Nieukerken (RMNH); 1♂, 1♀, 1 larva, Table mountain NP, Cecilia, parking lot, 180 m, 19.i.2013, EvN2013024, leafmines on Pelargonium cucullatum, emerged 24.i–12.ii.2013, E.J. van Nieukerken & H. Geertsema (RMNH); 1♂, Worcester, Fairy Glen, 15–19.x.1966, L. Vári & Potgieter (TMSA).

Leafmines, observation, Western Cape, Ashton, 2014, L. Torrance & H. Geertsema.