Type genus

Josia Hübner (1819).

Josiidae Piepers and Snellen, 1900: 26.

Josiinae Kiriakoff, 1950: 81.

Josiini Miller and Otero, 1994: 340.

Most Josiini are easy to recognize. Their unique tympanal structure (figs. 313B, 313C, 326) is definitive. More obscure, but equally effective synapomorphies include the hingelike joint at the base of each antennal ramus (fig. 312E, F). Forewing and HW veins M₃ and CuA₁ are invariably stalked, and there is never a FW stridulatory organ. The wing patterns are simple and always aposematic (pls. 26–Plate 27Plate 28Plate 29Plate 30Plate 31Plate 32Plate 33Plate 3435), usually comprising a dark ground color with a yellow or orange transverse band or longitudinal stripe. The larvae are unique for Dioptinae in showing a glassy head surface (fig. 358A, B), rather than a pebbled one (fig. 357A–E), and in going through only four instars (Miller and Otero, 1994) rather than five or six. The pupae form a loose net (pl. 39I), rather than being attached to the substrate completely exposed as in Dioptini, and they never possess hook-shaped setae on the abdominal dorsum (compare figs. 360C, E).

Redescription (Adults)

Labial palpus moderate in length, porrect (e.g., fig. 298D), never extremely short, rarely (Phintia) elongate and sickle shaped (fig. 328A, D); eye usually large and bulging (e.g., fig. 331A–D), occasionally somewhat small (fig. 336A, B); male antenna bipectinate, rami usually long, occasionally short; each ramus attached to shaft by a hingelike joint (fig. 312E, F); female antenna ciliate; tegula large (fig. 326); wing pattern simple (pls. 26–Plate 27Plate 28Plate 29Plate 30Plate 31Plate 32Plate 33Plate 3435), usually comprising a dark brown to blackish ground color, and a single yellow or orange transverse band or longitudinal stripe; some genera (Notascea, Scea) with a large orange triangle at FW base; FW stridulatory organ absent; FW and HW veins M₃ and CuA₁ stalked, rarely approximate, never separate; metathoracic tympanum deeply enclosed, forming a large, internal kettledrum-shaped cavity, opening to outside via a small, round opening (fig. 313B, C); tympanal membrane large, oriented horizontally; female St7 large but simple, never scaleless and setose; male terminal segments not heavily sclerotized, Tg8 and St8 not forming a jawlike structure; features of male genitalia relatively uniform; transtilla simple, arms joining to form a plate above aedeagus; Barth organ usually large, sometimes (Proutiella) extremely so, occasionally small (Josia); aedeagus large, globose at base, ductus ejaculatorius simplex opening dorsally; vesica frequently extremely large; cornuti large and spinelike, caltrop cornuti absent, rarely (Notascea) present.

(Immatures)

Larvae. Head surface glassy (fig. 358A, B); head usually black or white with black patterning (e.g., pl. 39H); thorax and abdomen patterned with white, yellow, and purplish maroon (pls. 39H, 39J–Q, 40B); A8 usually white, forming a conspicuous hump; A10 proleg small, held aloft, never stemapodiform; A10 crochets present, but reduced in size and number; four instars; caterpillar universally associated with Malpighiales (table 6), most commonly feeding on Passifloraceae.

Table 6

Known host plants of the Josiini

Subgeneric assignments for Passiflora species are based on Killip (1938) and MacDougal (1994).

Pupae. Formed in a loose net, often among detritus, not attached to substrate exposed; abdomen lacking hook-shaped dorsal setae; body reddish brown (pl. 39I), never patterned.

PROUTIELLA MILLER, new genus

Figures 284–Figure 285Figure 286Figure 287288; plate 26

Type species: Josia tegyra 142Druce, 1899.

Diagnosis

At first glance, Proutiella is an unlikely assemblage of moths (pl. 26). Even though the genus contains only nine species, wing-pattern variation is extensive. For example, the wings of P. vittula bear little resemblance to those of P. ilaire, and P. repetita looks different from P. simplex. Nevertheless, morphology within the group is remarkably homogeneous, showing numerous synapomorphies not found elsewhere in the Josiini. Proutiella species can be recognized by the following combination of traits: labial palpus short (fig. 284A, B), ascending to slightly above clypeus; Lp1 curving strongly upward; Lp2 gently curved, approximately equal in length to Lp1; scales on thoracic pleuron extremely short, appressed, always with a metallic blue-gray sheen; FW vein M₁ fused with base of radial sector (fig. 284C), arising from radius beyond outer margin of transverse band (slightly within outer margin in P. latifascia); FW and HW elongate (pl. 26), especially in males; abdominal venter whitish with a wide, gray-brown stripe along midline.

The male and female genitalia of Proutiella are especially distinctive. Among a long list of apomorphic features, two characters in males stand out: the BO is extremely large, and its membrane is delicate (e.g., fig. 288A); and the aedeagus is narrowly sinuate, almost always with a ratchetlike apex (e.g., fig. 286C). Females exhibit a signum structure unique for the Josiini (e.g., fig. 285E), and the dorsum of the CB bears a signumlike longitudinal sclerite (e.g., fig. 286E).

Description

Male. FW length  =  10.5–16.5 mm. Head (fig. 284A, B): Labial palpus short, either porrect, or ascending to immediately above clypeus; Lp1 relatively long, curved, with a loose fringe of longish scales below; Lp2 approximately as long as Lp1 or slightly longer, closely scaled on dorsal and lateral surfaces, longer scales below; Lp3 small, ovoid; clypeus broadly scaleless; scales of front short, appressed, pointing downward from below antennal bases; eye moderately large, postgena wide or moderately so; scales on vertex short, pointing anteriorly; antenna bipectinate, terminal 15–20 annulations simple.

Thorax: Epiphysis short, narrow, extending slightly beyond tibia apex; tibia short, approximately one-half length of femur; tegula fairly short, delicate, distal portion narrow; scales on pleuron extremely short, appressed; metathoracic tympanum kettledrum shaped, drum smaller than in most Josiini; tympanal membrane enclosed, fairly small, round, facing horizontally.

Forewing (fig. 284C; pl. 26): Elongate, acutely angled at apex; veins Rs₂–Rs₄ in the pattern [2+3]+4; M₁ fused for a long distance with base of radial sector, arising below base of Rs₁, beyond outer margin of transverse FW band; DC much longer than one-half FW length; veins M₃ and CuA₁ stalked; wing pattern highly variable, FW ground color black to blackish brown, or blackish brown/iridescent blue; always with a light yellow or orange-yellow transverse band beyond DC, band in P. vittula oblique.

Hind wing (fig. 284C; pl. 26): Elongate, acutely angled at apex; M₃ and CuA₁ stalked; HW ground color black to blackish brown, central area either concolorous with outer margin (P. infans, P. ilaire, P. repetita), with a white central area (e.g., P. tegyra), or with an orange-yellow longitudinal stripe (P. vittula).

Abdomen: Elongate, tapered distally; dorsum gray-brown, venter lighter gray-brown, with a pair of sublateral, white or whitish, longitudinal stripes.

Terminalia (figs. 285A–D; 286A–D; 287A, C–E; 288A–C, E): Tg8 shorter than St8, tapering distally; anterior margin of Tg8 slightly concave at midline, sometimes with short anterolateral apodemes; posterior margin of Tg8 with a large, wide, extremely deep U-shaped mesal excavation (in P. simplex and P. vittula excavated to near anterior margin); St8 wide anteriorly, strongly tapered distally, anterior margin broadly convex with a short mesal apodeme, posterior margin with a small, narrow mesal notch; socii/uncus complex with an extremely narrow attachment to tegumen, almost hinged; uncus beaklike, bent downward, with large posterior flanges above base at bend, apex truncate; socii variable, never as long as uncus, sometimes narrow (e.g., P. vittula), occasionally short and laterally compressed (e.g., P. simplex), or rarely (P. ilaire) compressed with dorsal flanges; tegumen narrow, much taller than vinculum, sides roughly parallel; vinculum short, extremely thin, especially across ventral margin of genitalia; saccus apparently absent; valva extremely large, costal margin upright; BO extremely large, occupying almost entire valva, greatly expanded along outer margin, membrane fragile; dorsal margin of BO narrowly sclerotized, sclerite appearing twisted; costa narrow to moderately wide, sometimes (P. simplex, P. ilaire) with a prominent flattened process on upper margin; apex membranous; arms of transtilla with a large, bladelike flange at base, arms pointing ventrally, meeting at midline to form a large, deeply concave medial plate, its posterolateral margins curled inward; aedeagus narrow, sinuate; base of aedeagus narrow, not bulbous (as in other Josiini); ductus ejaculatorius simplex arising anteriorly (arising dorsally in other Josiini); apex of aedeagus curved upward, forming a ratchetlike structure with 3–6 sclerotized, transverse ridges; vesica relatively small, somewhat wider than aedeagus, oriented upward and angled somewhat anteriorly; a group of short, spinelike cornuti near base, and a second group of extremely short, sometimes minute, cornuti beyond those.

Female. FW length  =  11.0–17.0 mm. Head: Similar to male, except Lp2 shorter; antenna ciliate.

Thorax: Similar to male.

Forewing: Longer and broader than in male (pl. 26), apical angle not as acute, transverse band wider.

Hind wing: Longer and broader than in male, outer margin more rounded; frenulum with two bristles.

Abdomen: Shorter and deeper than male, dorsal and ventral coloring similar.

Terminalia (figs. 285E, 286E, 287B, 287F, 288D): Tg7 extremely large, wide, slightly longer than St7, shape almost quadrate, anterior and posterior margins simple; St7 narrower posteriorly, anterior margin simple, posterior margin barely concave; Tg8 comprising a narrow, transverse, lightly sclerotized band, membranous at midline; AA short, extremely thin; A8 pleuron lightly sclerotized; PP long, straight, extremely thin; PA moderate in size, sclerotized, densely spiculate along posterior margin, posterior margin quadrate or nearly so, dorsal lobe absent; PVP long, wide, with a few longitudinal striae, surface densely spiculate; DB short to extremely short, dorsoventrally compressed, sinuate, lightly sclerotized; CB a large, elongate oval, narrowed and sclerotized at base, sometimes (e.g., P. adiante, P. repetita) goblet shaped; dorsum of CB with a narrow, elongate or oval-shaped sclerite running from near base, its internal surface finely spiculate, sclerite with a seam along midline; signum forming a small, concave, cup-shaped structure, bearing 2–15 short, internal spines along lateral margins; DS arising dorsally at base of CB; base of DS wide, then gradually tapering, coiled.

ETYMOLOGY

This genus name honors Louis Beethoven Prout (1864–1943), former curator of Lepidoptera at the Tring Museum, later the Natural History Museum, London (BMNH). It was Prout who first unified the genera of Dioptinae in a single group, assembling them from far-flung parts of the Lepidoptera classification. Without his astounding insights into dioptine systematics and taxonomy (Prout, 1918, 1920, 1922), this paper could not have been written.

Distribution

In Central America, Proutiella species occur in Costa Rica and Panama. In South America, they are found throughout the Amazon Basin as well as in eastern Brazil, with P. vittula (Rio de Janeiro) the southernmost representative of the genus. Proutiella is relatively rare in collections so distribution data is scarce, but the genus is apparently endemic to lowland forests. The highest recorded altitude is a Panamanian example of P. ilaire (USNM), collected by Gordon Small in 1963 at Cerro Campana, at an altitude of 900 meters. Similarly, a MPM specimen of P. vitulla was collected in 1938 by P. Gagarin at 900 meters near Petropolis Brazil. The majority of Proutiella specimens were captured well below 500 meters.

Biology

Nothing is known regarding the biology of Proutiella.

Discussion

Previous authors (Prout, 1918; Hering, 1925; Bryk, 1930; Forbes, 1931, 1939a) placed the species now in Proutiella in the large, then polyphyletic, genus Josia Hübner. My analyses offer a different solution; not only should these taxa be removed, but they form a clade widely separated from Josia in the josiine hierarchy. They are placed in Proutiella here as new combinations.

Proutiella arises as the basalmost lineage of the entire Josiini (fig. 283). This new genus thus plays a pivotal role in the phylogeny of the tribe. It will be important to collect as much information as possible regarding the biology of Proutiella. Discovering its host associations will provide an important missing piece in the puzzle of josiine host-plant evolution.

Figure 283

Cladogram of exemplar species in the Josiini showing generic assignments; relationships for the Dioptini are shown in figure 3.

A large suite of characters supports the monophyly of Proutiella (BS  =  21; fig. 2). In the genitalia (figs. 285–Figure 286Figure 287288), these include the following: male Tg8 with an extremely deep, U-shaped excavation along posterior margin; base of uncus with anteroventral angles, apex of uncus usually minutely truncate; socii various, not elongate and thin, often laterally compressed; upper margin of BO sclerotized, scalloped; aedeagus narrow and sinuate, often with an apical “ratchet”; female CB with an elongate, denticulate strip near base, strip bearing a seam along its midline; signum small, smooth, concave, with a few short, thickened spines; base of CB with a smooth, broadly sclerotized region to which DS is attached; PVP finely spiculate, with longitudinal striations.

Most Proutiella species are obscure. For example, Proutiella infans, occurring from Venezuela and French Guiana south to Amazonian Brazil, is rare, but is nevertheless represented in most museum collections. Its identity had never before been established, so the species has gone essentially unrecognized until now. Of the other Proutiella species, each is known from only a handful of specimens.

One of the most interesting aspects of Proutiella is the dramatic variation in hind-wing pattern that occurs (pl. 26), contrasted against relatively uniform genital morphology within the clade. All included species exhibit a lemon-yellow or orange-yellow transverse FW band, set against a dark ground color, a pattern widespread in the Josiini. However, the HW of Proutiella shows considerable variability. Here, we see a white central area in P. tegyra and relatives, an orange-yellow longitudinal stripe in P. vitulla, a uniformly iridescent blue HW in P. ilaire, and a purplish-brown HW in P. latifascia and P. infans. Proutiella repetita is unique in showing a yellow spot at the HW apex. Proutiella thus provides a microcosm of wing-pattern evolution, occurring within a small, tight-knit clade. The genus seems to provide an ideal model for analyzing the underlying factors of pattern change, and showcases the Josiini as a group with remarkable potential for evolving divergent wing-pattern types, even among closely related species.

Type Locality

Ecuador, Sarayacu.

Type

Holotype ♀, leg. Buckely (BMNH).

Discussion

I know this species exclusively from the female holotype (pl. 26), which was not dissected. Judging from its size and wing pattern, P. esoterica is closely related to P. tegyra, or might be a junior synonym. Both moths have Ecuador as their type locality. Four species within Proutiella—esoterica, jordani, simplex, and tegyra—show the same basic wing pattern (pl. 26). All are small moths with a blackish FW ground color, a yellow-orange transverse FW band, and a white HW central area. Although the taxonomy of this group is unresolved, my suspicion is that there are fewer than four valid species. However, until the relevant types have been dissected, I retain P. esoterica as a species, following previous authors. As Prout (1918) noted in his original description of P. esoterica, its wing length (FW  =  15.5 mm) exceeds that of P. tegyra females (FW length  =  11.5–13.5 mm). Furthermore, the white area in the HW of P. esoterica is larger than in any specimen of jordani, tegyra, or simplex I have seen.

Proutiella taxa exhibiting a black, yellow, and white pattern could potentially be confused with Ephialtias abrupta (pl. 29), a josiine with similar wings. However, in Ephialtias FW vein M₁ arises from the DC close to the base of the radial sector (fig. 311H), rather than being stalked with it, as occurs in all Proutiella (fig. 284C). Differences in labial palpus shape and genital morphology provide additional means for separating moths in these two genera.

Figure 284

Morphology of Proutiella (♂♂). A, head of P. vittula, lateral view; B, head of P. ilaire, lateral view; C, P. ilaire wings (illustration by J.S. Miller).

Distribution

Ecuador (BMNH).

Dissected

None.

Proutiella ilaire (Druce), new combination

Figures 284B, 284C, 285; plate 26 [EX]

Figure 285

Genitalia of Proutiella ilaire (♂ JSM-1380, ♀ JSM-1381). A, ♂ genitalia; B, aedeagus; C, ♂ Tg8; D, ♂ St8; E, ♀ genitalia (illustration by A. Trabka).

Ephialtias ilaire Druce, 1885a: 147, pl. 13, fig. 26.

Type Locality

Panama, Bugabá, 800–1500 ft.

Type

Syntype ♂, leg. Champion (BMNH).

Discussion

Proutiella ilaire can be distinguished by its iridescent, dark blue-gray ground color in the wings (pl. 26). The blue is usually somewhat more brilliant in the HW than in the FW. Proutiella ilaire shows similarities to P. repetita from Colombia, which differs in being slightly larger and in having a small, light yellow spot at the HW apex (pl. 26). However, morphological study suggests that, within Proutiella, P. ilaire is most closely related to P. latifascia; the two share the presence of a long, thin process on the costa of the male valva (figs. 285A, 287A).

In Costa Rica, P. ilaire is known almost exclusively from the Osa Peninsula (fig. 4), on the Pacific coast near the border with Panama. That material is undoubtedly conspecific with the P. ilaire type (from northwestern Panama). However, a single specimen (shown here as P. ilaire, pl. 26) was captured by Gunnar Brehm (April 2003) near Puerto Viejo (fig. 4), on the Caribbean side of the Cordillera Central. In lieu of obtaining additional material, I have not performed dissections to determine whether eastern material represents an undescribed species. If previous experience with Costa Rican Dioptinae holds true, the two are distinct.

Hering (1925) described turbida as a subspecies of ilaire, and Bryk (1930) listed that name as an ilaire synonym. I studied the male and female ZMH syntypes of P. turbida—both from northwestern Colombia—and have concluded that this taxon is instead a synonym of P. latifascia Prout (1918). Proutiella latifascia shows a characteristic purplish iridescence in the otherwise dark brown HW, and the FW band is wider than in P. ilaire. The turbida types show both traits clearly. Furthermore, they were collected in the same habitat as P. latifascia (see below).

Distribution

Costa Rica (AMNH, EME, INBio, LACM, SMNS); Panama (BMNH, CAS, USNM, ZMH); Colombia (AMNH, BMNH, USNM).

Dissected

♂, Costa Rica, Golfito, 24 Jul 1957, leg. Truxal & Menke, LACM (genitalia slide no. JSM-1486, wing slide no. JSM-1666); ♂, [no data], Wm Schaus Collection, USNM (genitalia slide no. JSM-1380); ♀, Panama, Chiriquí, Wm. Schaus Collection, USNM (genitalia slide no. JSM-1381); ♀, Panama, Canal Zone, Barro Colorado Island, 8 Jan 1945, leg. C. Michener, AMNH (genitalia slide no. JSM-1487).

Proutiella infans (Walker), new combination

Figure 286; plate 26

Figure 286

Genitalia of Proutiella infans (♂ JSM-1534, ♀ JSM-1535). A, ♂ genitalia; B, ♂ St8; C, aedeagus; D, ♂ Tg8; E, ♀ genitalia (illustration by J.S. Miller).

Scea infans Walker, 1856: 1647.

Type Locality

Brazil, Pará.

Type

Holotype ♀, ex Saunders Collection (OUMNH).

Discussion

Proutiella infans is superficially indistinguishable from two Lyces species (pl. 29)—ena (Boisduval) and enoides (Boisduval). Previous authors placed all of these in Josia. During the course of over 25 years of research on the Dioptinae, I was unsure of the identity of infans; no specimens had been identified as such in collections, whereas ena was apparently common. Not until Walker's type was finally located at the Oxford University collection (October 2006) could the characters of P. infans be established. It then became clear that not only are infans and ena distinct species, but they belong in different genera of the Josiini. Re-examination of material that had historically been identified as L. ena showed that, while ena is indeed more common, examples of P. infans reside in small numbers scattered throughout the world's major collections.

The two taxa are superficially so similar that specimens must be examined carefully to identify them. Two characters are particularly useful: First, in Lyces ena M₁ arises from the anterolateral angle of the DC, close to, but separate from, the base of the radial sector (see fig. 320I). In P. infans, on the other hand, M₁ is stalked with Rs₁–Rs₄, arising well out on the radial sector (see fig. 284C). This trait can be used to separate all members of Proutiella and Lyces. A second character is useful as well. The abdominal venter in P. infans is white with a wide, gray longitudinal stripe along its midline. In contrast, the venter in L. ena is completely white or grayish white, without a longitudinal stripe. The wings of P. infans are also somewhat more elongate, especially in males.

Within Proutiella, P. infans and P. latifascia show similar wing patterns. The best way to distinguish these is by their size. The FW length of P. infans (11.0–14.0 mm) is much shorter than that of P. latifascia (16.0–17.0 mm). In addition, Lp1 is white in P. infans but light gray-brown in P. latifascia. The two species do not overlap in distribution, with P. infans endemic to the Upper Amazon, including French Guiana, but P. latifascia occurring in southwestern Colombia and western Ecuador. Aside from this, P. infans cannot be confused with any other Proutiella species.

Proutiella infans shows a distribution roughly congruent with that of its look-alike, Lyces ena. One difference is that L. ena has been collected in Panama, whereas P. infans is known exclusively from South America. Based on all the material of P. infans I could locate in museum and private collections (15 specimens in total), the species occurs from the mouth of the Amazon upriver to Santarém and Manaus, and even further west to Ariquemes in Rondônia, at the headwaters of the Rio Madeira. Proutiella infans has also been collected in French Guiana and Venezuela.

It seems likely that careful searching through collections will reveal many more examples of P. infans than are currently known. For example, the collection at the BMNH contains hundreds of specimens identified as L. ena, but none of P. infans; some of their material is undoubtedly the latter. Unfortunately, at the time of my most recent visit to the London collection (March 2005), I was unaware of the identity of P. infans. Indeed, the genus Proutiella was an unknown entity at that time. It will be important for Lepidopterists to sort through their material identified as L. ena, separating out the ones in which FW vein M₁ is fused with the radial sector—these are P. infans.

Distribution

Brazil (BMNH, CMNH, USNM, ZMH); French Guiana (BHC, USNM); Venezuela (AMNH, MNHN).

Dissected

Holotype ♀ (genitalia dissection JSM-1736); ♂, Brazil, Rondônia, 62 km S Ariquemes, Fazenda Rancho Grande, 165 m, 18–26 Apr 1991, 10°32′S, 62°48′W, leg. Ron Leuschner, AMNH (genitalia slide no. JSM-1534); ♀, Brazil, Rondônia, 62 km S Ariquemes on Linea C-20, 22 km W B-364, 10°32′S, 62°48′W, 11 Aug 1995, leg. A.D. Warren, AMNH (genitalia slide no. JSM-1535).

Proutiella jordani (Hering), new combination

Plate 26

Josia (Phintia) jordani Hering, 1925: 525, fig. 70g.

Type Locality

None stated.

Type

Holotype ♀ (ZMH).

Discussion

The only confirmed example of P. jordani known to me is the female holotype (pl. 26), which bears no data other than a label reading “66”. Based on examination of that specimen, wing-pattern characters differentiating P. jordani from P. tegyra involve the shape of the white central area of the HW and the shape of the yellow FW band. In P. jordani, the white HW area is more elongate at its apex, and the inner and outer margins of the yellow FW band are indented, rather than smooth. A third species—P. simplex Walker (pl. 26)—is also extremely similar to P. jordani; it shows an elongate white area, acute at its apex, but differs in having smooth borders around its yellow FW band. The female genitalia of all three taxa show differences, especially regarding the shape of the signum and the size of the PVP. I therefore retain them as valid species, following Hering (1925) and Bryk (1930).

Distribution

Not known.

Dissected

Holotype ♀ (genitalia slide no. JSM-1763).

Proutiella latifascia (Prout), new combination

Figure 287A–E; plate 26

Figure 287

Genitalia of Proutiella. A, ♂ of P. latifascia (JSM-1536); B, ♀ of P. latifascia (JSM-1537); C, aedeagus of P. latifascia; D, ♂ Tg8 of P. latifascia; E, ♂ St8 of P. latifascia; F, holotype ♀ of P. simplex (JSM-1737).

Josia latifascia Prout, 1920 (replacement name for Ephialtias lativitta Warren, 1901: 424): 509.

Type Locality

Ecuador, Cachabí.

Type

Holotype ♂, leg. Rosenberg, low country, January 1897 (BMNH).

Josia Phintia turbida Hering, 1925: 525, fig. 71f. Revised synonymy; formerly a synonym of ilaire Druce.

Type Locality

Colombia, Río San Juan.

Type

Syntype ♂/♀ (ZMH).

Discussion

My knowledge of P. latifascia, until now an obscure taxon, has accumulated only within the past year. Altogether, seven specimens have been assembled. A male on loan from the MNHN (JSM-1536) matches the type of P. latifascia precisely, and both are from western Ecuador. Two AMNH females from Nariño, Colombia, are undoubtedly conspecific with these. The state of Nariño, in extreme southwestern Colombia, is contiguous with these Ecuadorian localities. The types of turbida Hering, newly recognized as a synonym of latifascia (see paragraph below), were collected on the Río San Juan in northwestern Colombia. Thus, all known examples of P. latifascia were captured along the Pacific slope of the Colombian and Ecuadorian Andes. Recorded elevations range between 250 and 500 meters.

Proutiella latifascia is similar in appearance to P. ilaire (pl. 26), but P. latifascia is generally larger and exhibits a purplish, rather than blue, iridescence in the wings. The species closest to P. latifascia in wing coloration is P. infans, but the latter is tiny and impossible to confuse with these.

Prout (1920), who united lativitta Warren (1901) and lativitta Walker (1869) together in Josia, proposed latifascia as a replacement for Warren's name. My examination of the ZMH types shows that turbida Hering (1925), formerly regarded as a synonym of ilaire (Bryk, 1930), is instead a synonym of latifascia.

Distribution

Ecuador (BMNH, MNHN); Colombia (AMNH, ZMH).

Dissected

♂, Ecuador, Santo Domingo de Los Colorados, 1930, leg. R. Benoist, MNHN (genitalia slide no. JSM-1536); ♀, Colombia, Nariño, La Guayacana, 250 m, Jul 1946, leg. Von Schneidern, AMNH (genitalia slide no. JSM-1537).

Proutiella repetita (Warren), new combination

Plate 26

Ephialtias repetita Warren, 1905: 314.

Type Locality

Colombia, Guadalite, Cundinamarca.

Type

Holotype ♂, leg. de Mathan, August 1903 (BMNH).

Discussion

Proutiella repetita is distinctive in exhibiting an iridescent blue-black FW and HW ground color, a yellow transverse band in the FW, and a small yellow spot at the HW apex. Its sympatric relatives—P. ilaire and P. latifascia—lack the HW spot.

Proutiella repetita is relatively rare in collections; apart from a series of seven specimens at the BMNH, I know this taxon from only eight other examples worldwide. Based on what little material is available for these three species, the following summarizes their known geographical distributions: Proutiella repetita is restricted to Colombia; P. ilaire occurs from Colombia north to Costa Rica; P. latifascia is endemic to Colombia, but exclusively on the Pacific slope of the Andes. Proutiella repetita is apparently endemic to Colombia's Amazonian side.

Distribution

Colombia (BMNH, CAS, OUMNH, USNM, ZMH).

Dissected

♂, Colombia, Chiriguano District, Lake Sapatoza region, Aug–Sep 1924, leg. C. Allen, BMNH (genitalia slide no. JSM-312); ♀, Colombia, Santa Marta, BMNH (genitalia slide no. JSM-311).

Proutiella simplex (Walker), new combination

Figure 287F; plate 26

Ephialtias simplex Walker, 1856: 1648.

Type Locality

Brazil, Pará.

Type

Syntype ♀, ex Saunders Collection (OUMNH).

Discussion

Proutiella simplex shows female genitalia (fig. 287F) different from those of P. tegyra (fig. 288D), the species with which it could most easily be confused. In P. simplex, the signum bears small internal spicules; these are larger in P. tegyra. Another difference involves the shape of the papillae anales; the lobes are quadrate in P. tegyra females but become narrow dorsally in P. simplex. The easiest way to separate the two species is by their wing lengths—females of P. simplex (FW length  =  14.0–15.0) are larger than P. tegyra (FW length  =  11.5–13.5 mm).

Figure 288

Genitalia of Proutiella tegyra (♂ JSM-1422, ♀ JSM-318). A, ♂ genitalia; B, aedeagus; C, ♂ St8; D, ♀ genitalia; E, ♂ Tg8.

However, the identity of P. simplex remains somewhat uncertain. I have not dissected verified males. For the time being, I restrict the identity of this species to material collected in the state of Pará (Brazil), along the eastern portion of the Amazon River. Using this criterion, only seven specimens of P. simplex are known worldwide—two female syntypes at the OUMNH (both dissected), four females at the BMNH, and a single male in the ZMH collection.

Distribution

Brazil (BMNH, OUMNH, ZMH).

Dissected

Syntype ♀ (genitalia slide no. JSM-1737); syntype ♀, Brazil, Pará, OUMNH (genitalia slide no. JSM-1748).

Proutiella tegyra (Druce), new combination

Figure 288; plate 26 [EX]

Josia tegyra 142Druce, 1899: 298.

Type Locality

“Ecuador”.

Type

Syntype ♀, leg. Whitely (BMNH).

Josia (Phintia) tessmanni Hering, 1925: 525, fig. 70g. New synonymy.

Type Locality

Peru, Río Pachitea.

Type

Syntype ♂/♀ (ZMH).

Discussion

Numerous small Proutiella species show a narrow yellow transverse FW band and a white HW central area (pl. 26). Bryk (1930) listed five—esoterica Prout (1918), jordani Hering (1925), simplex Walker (1856), tegyra 142Druce (1899), and tessmanni Hering (1925). In an attempt to make taxonomic sense of this situation, I dissected material from sites across the Amazon Basin. Genitalia comparisons confirm my suspicion that fewer valid species exist than had heretofore been recognized. However, I took a conservative approach, sinking only one of the five names; tessmanni is now a junior synonym of tegyra.

According to my concept, P. tegyra exhibits slight intraspecific differences in the size and shape of the yellow FW band, as well as in the shape of the white HW area, but these are not correlated with differences in genital morphology. Proutiella tegyra is the smallest member of this species complex, being significantly smaller than P. esoterica and P. simplex, but only slightly smaller than P. jordani. I did not dissect the types of tegyra, esoterica, or jordani, and there remains a strong possibility that even fewer valid species are involved. Resolution of that issue will require revisionary work.

Specimens of P. tegyra can be found in most major collections, but usually only a few examples exist at each. The species appears to occur throughout lowland forests of the Upper Amazon Basin. It has not been recorded in the lower Amazon, home to P. simplex.

Distribution

Ecuador (AMNH, BMNH); Peru (AMNH, CMNH, ZMH); French Guiana (JCC); Brazil (AMNH, BMNH, CMNH, USNM).

Dissected

♂, Peru, Middle Río Ucayali, 20 Sep 1923, H. Bassler Collection, AMNH (genitalia slide no. JSM-282); ♂, Brazil, São Paulo de Olivença, Sep 1935, leg. S. Waehner, BMNH (genitalia slide no. JSM-317); ♂, Peru, Río Pachitea, leg. G. Tessmann, ZMH (genitalia slide no. JSM-1422); ♂, Peru, Achinamiza, 25 Dec 1925, F6001, H. Bassler Collection, Acc. 33591, AMNH (genitalia slide no. JSM-1727); ♂, Peru, Yurimaguas, Río Huallaga, 3 Apr 1920, AMNH (genitalia slide no. JSM-1726); ♀, Ecuador, Sucumbíos, Yasuní, Garzacocha, 6 Aug 1992, leg. P. DeVries, AMNH (genitalia slide no. JSM-1725); ♀, Peru, Achinamiza, 11 Oct 1927, H. Bassler Collection, AMNH (genitalia slide no. JSM-283); ♀, Brazil, Amazonas, Bates Collection, BMNH (genitalia slide no. JSM-318).

Proutiella vittula (Hübner), new combination

Plate 26

Ephialtias vittula 254Hübner, 1823: 15, figs. 265, 266.

Type Locality

“Brazil”.

Type

Not seen.

Josia adiante Walker, 1854: 303.

Type Locality

None.

Type

Syntype ♂, ex E. Doubleday Collection (BMNH).

Ephialtias carneata Warren, 1901: 440.

Type Locality

Rio de Janeiro, Brazil.

Type

Holotype ♂ (BMNH).

Discussion

Attempts to resolve the identity of L. vittula have proved to be perplexing. As is true of all Hübner types, the type is either lost or destroyed. What remains are 254Hübner's color figures (1823: fig. 265, 266). These show a moth with a steeply angled, orange-yellow transverse FW band, and a thin orange-yellow HW stripe, acute at its apex. After extensive study of hundreds of specimens from collections around the world, comparing each carefully with Hübner's illustration, I have established the identity of P. vitulla to the best of my ability.

Unfortunately, the specimens I conclude should be called P. vittula are not those identified as such in museums. Material identified as “Josia vittula” in major collections, often appearing in large series, invariably constitutes Lyces angulosa Walker (pl. 29). My research suggests that P. vittula itself is rare—I have seen approximately a dozen specimens, all collected in eastern Brazil. Within the Josiini, true vittula belongs in Proutiella, whereas what has been called vittula belongs in Lyces.

The wing pattern of P. vitulla is divergent with respect to those of other Proutiella species. This phenomenon can perhaps be explained by the evolution of mimetic resemblance with co-occurring Lyces species in southeastern Brazil. The wings of Lyces constricta (pl. 29) are so similar to those of P. vitulla, that these taxa can only be separated by comparing wing venation—if M₁ arises from the radial sector (e.g., fig. 284C) the moth is P. vitulla; if M₁ arises from the upper angle of the DC (e.g., fig. 320I) the species belongs in Lyces.

Distribution

Brazil (AMNH, BMNH, CMNH, MPM, NMW, OUMNH, USNM, ZMC).

Dissected

♂, Brazil, “Mus. Westerm.”, ZMC (genitalia slide no. JSM-367); ♀, Brazil, Minas Gerais, ZMC (genitalia slide no. JSM-368).

GETTA WALKER, [1865] 1864

Figures 289–Figure 290Figure 291Figure 292Figure 293Figure 294Figure 295Figure 296297; plates 26, 27

Figure 289

Morphology of Getta. A, head of G. unicolor ♂, lateral view; B, head G. niveifascia ♂, lateral view; C, head of G. niveifascia ♂, frontal view; D, head G. niveifascia ♂, posterior view; E, G. baetifica ♂ labial palpus; F, G. baetifica ♀ wings; G, G. niveifascia ♂ tegula; H, G. baetifica ♂ wings (illustration by J.S. Miller).

Figure 290

Scanning electron micrographs of Getta baetifica ♂. A, area surrounding androconial organ on FW ventral surface, near anal margin of wing; B, androconia on FW ventral surface; C, scaleless area surrounding androconial organ on FW ventral surface; D, base of scent scale on FW ventral surface; E, typical scales on HW dorsal surface; F, base of typical scale on HW dorsal surface.

Figure 291

Scanning electron micrographs of Getta baetifica male, HW dorsal surface. A, scaleless corrugated area surrounding androconial organ; B, close-up of scaleless area; C, scent scales in androconial organ; D, scent scales; E, base of scent scale; F, surface of scent scale.

Figure 292

Genitalia of Getta niveifascia (♂ JSM-366, ♀ JSM-563). A, ♂ genitalia; B, ♂ St8; C, aedeagus; D, ♂ Tg8; E, ♀ genitalia (illustration by A. Trabka).

Figure 293

Genitalia of Getta tica, sp. nov. (♂ JSM-1519, ♀ JSM-564). A, ♂ genitalia; B, ♂ Tg8; C, aedeagus; D, ♀ genitalia; E, ♂ St8.

Figure 294

Costa Rica, showing the known distribution of Getta tica, sp. nov.

Figure 295

Genitalia of Getta turrenti, sp. nov. (paratype ♂ JSM-569, paratype ♀ JSM-1733). A, ♂ genitalia; B, ♂ Tg8; C, aedeagus; D, ♀ genitalia; E, ♂ St8, ventral view.

Figure 296

Mexico and northern Central America, showing the known distribution of Getta turrenti, sp. nov.

Figure 297

Genitalia of Getta unicolor (♂ JSM-1523, ♀ JSM-1734). A, ♂ genitalia; B, ♀ genitalia; C, ♂ Tg8; D, aedeagus; E, ♂ St8.

Getta Walker [1865] 1864: 139. Type species: Getta niveifascia Walker, 1864 (by monotypy).

Diagnosis

Of the six described species in Getta, five of them—baetifica, niveifascia, tica, turrenti, and unicolor—show features on the male wings making them instantly identifiable as members of this genus. In these, the FW DC is short and extremely narrow, and FW vein M₁ is stalked far out onto the radial sector (fig. 289H). The HW is an unusual triangular shape, showing an extreme angle near its apex, and its DC is also short, less than one-third the wing length. However, by far the most striking aspect of their male wings is the presence of a unique androconial system. This organ comprises two large, ovoid patches of yellowish beige deciduous scales near the wing bases (fig. 289H)—one on the FW ventral surface inside the anal margin (fig. 290A–C), and the other on the HW dorsal surface near its anterior margin (fig. 291). The two patches correspond in location, and perhaps overlap when the moth is in flight (Prout, 1918). Surrounding each patch is a broad area, shiny and devoid of scales (fig. 290A–C; fig. 291A, B). Such an androconial organ is not found elsewhere in the Notodontidae. In all Getta species (males and females), the FW ground color is dark with a yellow or cream-colored transverse band beyond the DC, and the HW is uniformly dark, lacking pattern.

The sixth species, G. ennia, does not exhibit all the features listed above, and is more difficult to identify as a member of Getta. Most significantly, males do not possess the conspicuous androconial patches on the FW and HW. However, close examination of G. ennia males reveals deciduous, light gray scales in the corresponding wing regions. These are not nearly as showy as in the others, and a scaleless area does not surround them. Furthermore, the FW and HW discal cells in G. ennia are small, but are not as extremely reduced. Other Getta apomorphies of the male wings, including the stalking of FW vein M₁ and the triangular HW, do occur.

Getta females are more difficult to recognize than males. The most important characteristics are: FW vein M₁ stalked for a long distance along radial sector (fig. 289F); and FW DC shorter than one-half the wing length. At the distal margin of the DC, the UDC is displaced basally a considerable distance from the LDC. Because of their similar sizes and wing patterns, females of Getta unicolor and Phavaraea rejecta are particularly easy to confuse. In addition to the wing-venation traits listed above, a reliable way to separate them is to compare their labial palpi. In G. unicolor, the palpus terminates in a broad, laterally compressed tuft (fig. 289A) that hides Lp3; this tuft is smaller in females, but is clearly present. The palpus of P. rejecta, on the other hand, is without a terminal scale tuft. Instead, Lp2 is tightly covered with short scales, and Lp3 is prominent, drooping slightly downward (fig. 305A).

Redescription

Male. FW length  =  15.0–23.0 mm. Head (fig. 289A–E): Labial palpus porrect, ascending to immediately above clypeus; Lp1 curving upward, with a loose fringe of long, yellow scales below (brown in G. ennia); Lp2 over one and a half times as long as Lp1, slightly curved upward, either tightly covered with short scales (G. niveifascia, G. ennia), or more loosely covered (G. baetifica, G. tica, G. unicolor) and with a diffuse fringe of longer scales below; Lp3 short, bullet shaped, or extremely short, almost round; scales of front variable, long or short, usually pointing ventromedially, in G. unicolor and G. ennia forming a prominent central tuft; eye large and round, bulging outward, postgena extremely narrow; scales on vertex pointing anteriorly; antennal scape with a short scale tuft on inner surface; antenna bipectinate, rami somewhat laterally compressed, terminal 24–30 annulations simple.

Thorax (fig. 289G): Epiphysis apically acute, extending to apex of tibia; tegula almost equal in length to mesoscutum, distal portion elongate, ventral process acute, sulcus strong; metathoracic tympanum kettledrum shaped, region below and behind tympanal opening scaleless; tympanal membrane large, enclosed, oriented horizontally.

Forewing (figs. 289H, 290; pls. 26, 27): Elongate, apical angle unusually acute; vein Rs₁ arising from radial sector; veins Rs₂–Rs₄ in the pattern [2+3]+4, except in G. unicolor (see species account); M₁ fused with Rs₁–Rs₄ for a long distance, arising below Rs₁; DC shorter than one-third FW length and extremely narrow (in G. ennia less than one-half FW length and somewhat narrow); veins M₃ and CuA₁ long stalked; all species (except G. ennia) with a large, ovoid patch of yellowish beige, deciduous androconia on ventral surface near base, straddling anal fold and extending posteriorly to 1A+2A; androconial patch surrounded by a wide, scaleless area between cubitus and anal margin; ground color black or blackish brown, with a yellow or white transverse band beyond DC, extending from behind costa at middle of wing anterior margin to tornus.

Hind wing (figs. 289H, 291; pls. 26, 27): Broad and short; expanded forward from near base to form an angled leading edge; outer margin near apex forming an acute angle; Rs and M₁ fused; M₃ and CuA₁ long stalked; all species (except G. ennia) with a large, ovoid patch of yellowish beige, deciduous androconia on dorsal surface near base, straddling Sc+R and extending posteriorly through DC to cubitus.

Abdomen: Long or moderately long, apex acute, without dorsal spots or lateral stripes.

Terminalia (figs. 292A–D; 293A–C, E; 295A–C, E; 297A, C–E): Tg8 short, anterior margin with a shallow mesal excavation, posterior margin with an extremely wide, U-shaped excavation; St8 narrower than Tg8, tapered posteriorly; anterior margin of St8 with a greatly elongate, thin mesal apodeme, sometimes reaching to near middle of St6; posterior margin of St8 with a deep U-shaped mesal excavation (G. baetifica, G. tica, G. ennia), a shallow U-shaped excavation (G. unicolor), or a W-shaped excavation (G. niveifascia); socii/uncus complex with a narrow attachment to tegumen, except wide in G. ennia; uncus long, fairly narrow, apex slightly truncate, sometimes with a dorsal crest (G. baetifica, G. tica), or a transverse, basal collar (G. ennia); socii variable, long and thin or laterally compressed with bladelike dorsal flanges; tegumen shorter than vinculum, sometimes greatly so, abruptly wider in dorsal section; vinculum taller than tegumen, narrow; saccus apparently absent; valva large, mostly membranous, BO extremely large, occupying over three-fourths of valva; longitudinal sclerite of BO sometimes with a pointed, sclerotized process above (G. baetifica, G. tica); costa sclerotized, narrow, fairly short; distal portion of costa forming a sclerotized, thornlike apical process (blunt in G. unicolor); arms of transtilla narrow, bent sharply downward from attachment to valva, meeting in manica to form a concave mesal sclerite; aedeagus fairly short, wide, slightly angulate at base below; apex of aedeagus gradually attenuated, sometimes with a sharply downcurved distal process; vesica as long as aedeagus, sometimes much longer, bearing spinelike cornuti of varying size and thickness; G. unicolor with unique, short, scoluslike cornuti near middle.

Female. FW length  =  15.0–25.0 mm. Head: Similar to male; antenna ciliate, shaft wide in middle section.

Thorax: Similar to male.

Forewing (fig. 289F; pl. 26): Longer, wider and much broader than male, outer margins more rounded; transverse band wider; venation similar to male, except DC not as short, slightly less than one-half FW length, and not as narrow; androconial organ absent.

Hind wing (fig. 289F; pl. 26): Longer than male, without expanded anterior area; DC not small; androconial organ absent; frenulum comprising two bristles.

Abdomen: Shorter and much wider than male.

Terminalia (figs. 292E, 293D, 295D, 297B): Tg7 large, wide, trapezoidal, slightly narrower posteriorly, anterior and posterior margins simple; St7 narrower than Tg7, similar in shape; Tg8 apparently absent, dorsum of A8 completely membranous; AA moderately long, sometimes short; A8 pleuron membranous, with a narrow sclerite running along anterior margin; PP long and thin, bent downward slightly; PA small, membranous, with a small dorsal lobe; PVP short, wide, concave, its surface minutely spiculate; DB short, dorsoventrally compressed, lightly sclerotized; CB a large, elongate oval, variously sclerotized at base; CB bearing a large, sclerotized ventral plate, its surface coriaceous, its margins sometimes spinose; signum comprising an elongate dorsal band, wider anteriorly, its internal surface coarsely dentate (G. baetifica, G. tica, G. ennia), or a teardrop-shaped, heavily dentate sclerite with a central knob (G. niveifascia); DS arising near base of CB, on right side from a small, membranous appendix.

Distribution

Getta spans almost the entire distribution of the Josiini from southern Mexico south to Bolivia and from the Andes east to French Guiana and Amazonian Brazil. Most of the species inhabit lowland areas; the highest recorded altitude comes from G. baetifica, which has been collected at 1750 meters on the western slopes of the Ecuadorian and Colombian Andes.

Biology

Life histories have been discovered for three Getta species: G. baetifica (Ecuador), G. niveifascia (French Guiana), and G. tica (Costa Rica). In all three, the larvae are unusual in their association with Passiflora species belonging to the subgenus Astrophea (table 6). Members of this fascinating subgenus—comprising 50 species confined to tropical areas of South America, with only two taxa ranging north into Central America (Escobar, 1994)—grow as woody shrubs (pls. 40C, 41B) and trees rather than as herbaceous vines, the growth form more typical for Passiflora (Killip, 1938; Holm-Nielsen et al., 1988). Astrophea species invariably grow in extremely moist areas, often overhanging stream banks (pl. 40A). The leaves of these plants can be extremely large (pl. 41C); the leaves of Passiflora macrophylla, one of the hosts of G. baetifica, sometimes attain a length of nearly two feet. Escobar (1994) provided a key to all described species in Astrophea.

The biology of Getta baetifica was previously discussed in a phylogenetic study of the Josiini (Miller, 1996). In that paper it was pointed out that an association with Astrophea also occurs within the Heliconiinae (Nymphalidae). It will be fascinating to investigate whether these two unrelated groups of Lepidoptera—the josiines and heliconiines—show similar patterns of Passiflora use.

Getta larvae are striking. The white, yellow, and blackish-purple pattern on the body is bold (pl. 40B). Furthermore, the body integument in Getta caterpillars bears an almost glassy sheen, an unusual characteristic for the Josiini.

Discussion

Characters from adult morphology suggest that Getta falls at the base of a larger clade that includes two additional genera (Clade 19; fig. 7)—Polyptychia and Phavaraea. All told, this lineage contains only 12 species (appendix 2), and an argument could be made to subsume them within a single genus. However, I defer to previous treatments of the group in retaining those genera; each is easily recognizable.

Members of the clade show several remarkable biological attributes. First, like Getta, Polyptychia is restricted to the Passiflora subgenus Astrophea (table 6). It will be important to focus field efforts on finding life histories for members of Phavaraea. One expects that their larvae are associated with Astrophea as well. The second unusual feature of this clade is the seemingly rampant evolution of androconial systems within it. Getta shows complex androconial patches on the anterior margins of the FW and HW (figs. 289H, 290, 291), but other species within this lineage show androconia on the HW anal margin (e.g., Phavaraea rejecta; fig. 306A), in the center of the HW (e.g., Polyptychia hermieri; figs. 298E, 299, 300), on the legs (Polyptychia hermieri; pl. 27), and on the dorsal (P. dilatata) as well as ventral (e.g., Phavaraea poliana; pl. 27) abdominal surfaces. How and why such a fantastic collection of androconial systems evolved within a single small clade will provide a fascinating subject for future research.

Figure 298

Morphology of Polyptychia hermieri (♂). A, head, lateral view; B, head, frontal view; C, head, posterior view; D, head, lateral view; E, wings (illustration by J.S. Miller).

Figure 299

Scanning electron micrographs of Polyptychia hermieri (♂), HW dorsal surface. A, brush of bristlelike scales covering central androconial pocket; B, base of brush, with a few androconia from central pocket showing below; C, base of brush scales; D, apex of a single brush scale; E, apex of a brush scale; F, surface of a brush scale.

Figure 300

Scanning electron micrographs of Polyptychia hermieri (♂), HW dorsal surface. A, close-up of brush scale surface; B, central androconial pocket, showing bare area where deciduous androconia have been removed; C, deciduous androconia of central pocket; D, apex of deciduous androconium; E, base of deciduous androconium from central pocket; F, close-up of deciduous androconium base.

The species content of Getta proposed in this paper differs significantly from previous classifications (Prout, 1918; Hering, 1925; Bryk, 1930). Changes involve the addition to Getta of G. unicolor Hering, taken from Polyptychia, and the description of two new Getta species related to G. baetifica. Furthermore, I transfer clite Walker from Getta to Erbessa, a rather dramatic leap across the dioptine classification (appendix 2). Three species of Getta were listed in the dioptine catalogue (Bryk, 1930); now there are six. Although there is some doubt as to whether ennia belongs in Getta since males do not show the elaborate FW and HW androconia (see discussion for G. ennia), cladistic analyses (fig. 283) support the hypothesis of previous authors in assigning it here.

When Walker (1864) described Getta, the type species became G. niveifascia Walker by monotypy. Later, Prout (1918) placed niveifascia in synonymy with clite Walker (1854). However, my studies show that clite belongs in Erbessa. As a consequence, the type of Getta reverts to G. niveifascia Walker.

Type Locality

Ecuador, Angamarca.

Type

Syntype ♀ (BMNH).

Discussion

My attempts to understand the taxonomy and geographical distribution of Getta baetifica provide a classic example where, what at first seemed to be a single species with a broad distribution, was instead revealed to be a group of cryptic species. Two of those are newly described here. Such cases, typical for the Dioptinae, are perhaps the general rule for most neotropical moth groups.

After dissecting moths along a north-south gradient from Mexico to Ecuador, I discovered two new species closely related to G. baetifica: G. tica (eastern Costa Rica and Panama) and G. turrenti (southern Mexico and northeastern Guatemala). The distribution of G. baetifica itself can now be specified as being Chocoan, extending along the western slope of the Ecuadorian and Colombian Andes at elevations between 500 and 1750 meters. The northern limits of G. baetifica remain unclear; western Panamanian material must be studied to determine whether it should be referred to G. baetifica or G. tica. At Colón, on the Caribbean coast of Panama, G. tica occurs.

Getta baetifica seems to be relatively common. However, as holds true for many Josiini, the larvae are much easier to find than adults; on various occasions, Elicio Tapia and I have collected large numbers of G. baetifica caterpillars feeding on either Passiflora macrophylla or P. arborea in western Ecuador. Adults were rarely observed at those sites. Getta baetifica is somewhat difficult to distinguish from the two newly described Central American species. Means for separating them are discussed in the species descriptions (below), and are presented in the key to Getta species (above).

A BMNH record for G. baetifica, based on a single specimen—“Rio, Brazil, Stg.”—seems doubtful. Similarly, a Bolivian record at the ZMH (Corvico) needs further study. No additional material is available to verify these seemingly far-flung localities.

Distribution

Colombia (AMNH, BMNH, CMNH, LACM, USNM, ZMH); Ecuador (AMNH, BMNH, CMNH, CUIC, LACM, OUMNH, MNHN, MUSM, USNM).

Dissected

♂, Ecuador, Cotopaxi, Las Pampas, 1500 m, 12 Aug 1993, leg. E. Tapia, ex Passiflora arborea, AMNH (genitalia slide no. JSM-711); ♂, Ecuador, Cotopaxi, Las Pampas, 1500 m, 5 Sep 1993, leg. E. Tapia, AMNH (genitalia slide no. JSM-705); ♂, Ecuador, Paramba, USNM (wing slide no. JSM-1358); ♂, Colombia, Chocó, Tambito, 1700 m, Aug 1946, leg. Von Schneidern, AMNH (genitalia slide no. JSM-1522); ♀, Ecuador, Dos Puentes, 1750 ft, Mar 1930, leg. W.J. Coxey, CMNH (genitalia slide no. JSM-566); ♀, Ecuador, Esmeraldas, El Placer, 550 m, 25 Dec 1993, leg. Elicio Tapia, AMNH (genitalia slide no. JSM-1521); ♀, Colombia, Río Aguacatal, 1600 m, 15 Jul 1908, USNM (wing slide no. JSM-1357).

Getta ennia Druce

Plate 26 [EX]

Getta ennia 142Druce, 1899: 295.

Type Locality

“Brazil”.

Type

Syntype ♀, leg. Smith (BMNH).

Discussion

Since its original description (142Druce, 1899), ennia has been included in Getta, even though males lack the conspicuous FW and HW androconial patches found in other species. My cladistic analyses confirm this placement (fig. 283). Supporting evidence includes: the tuft of scales on the antennal scape; the presence of an elongate anterior apodeme on male St8; the presence of a thornlike process on the valva apex, and the presence of a coriaceous sclerite on the female CB. Getta ennia also shows three important wing venation characters of the genus: FW vein M₁ is stalked with Rs₁–Rs₄; HW veins Rs and M₁ are completely fused; and the male FW DC is short and narrow.

The androconia found in other Getta species were thus either lost in G. ennia, or this species represents the basal element of the genus and they evolved subsequently. As is noted in the Diagnosis for Getta (above), the wings of G. ennia males seem to possess a rudimentary androconial system. This could have led to the evolution of more elaborate structures.

Getta ennia shows additional characteristics setting it apart from other Getta species. All other taxa have Lp1, the lower portion of the occiput, and often the propleuron, yellow. In G. ennia these regions are uniformly dark gray, like the rest of the body. The labial palpi of G. ennia are thinner than those of other Getta species. The shape of the male tegumen and socii uncus complex are also unusual, as is the structure of the ventral plate in the female corpus bursae.

An AMNH male from Bolívar in southern Venezuela (leg. Carl Ferraris) is blacker than most examples of Getta ennia; the abdomen is entirely black above, with the venter being broadly dark gray. Its genitalia (JSM-567) differ slightly from those of G. ennia, suggesting that this specimen represents an undescribed species.

Distribution

Brazil (AMNH, BMNH, CMNH, ZMH); Peru (BMNH); Ecuador (BMNH); Venezuela (AMNH).

Dissected

♂, Brazil, Pará, leg. A.M. Moss, BMNH (genitalia slide no. JSM-1406); ♂, Brazil, Humayta, Río Madeira, Jul–Sep 1906, leg. W. Hoffmanns, BMNH (genitalia slide no. JSM-1407); ♂, Brazil, Manacapurú, Mar 1926, leg. S.M. Klages, C.M. Acc. 8607, CMNH (genitalia slide no. JSM-562); ♀, Brazil, Humayta, Río Madeira, Jul–Sep 1906, leg. W. Hoffmanns, BMNH (genitalia slide no. JSM-1408); ♀, Amazons, L. de Villafranca, coll. 477, BMNH (genitalia slide no. JSM-295).

Getta niveifascia Walker, revised status

Figures 289B–D, 289G, 292; plate 26 [EX]

Getta niveifascia Walker, 1864: 140.

Type Locality

Peru, Nauta, “Amazon Region”.

Type

Syntype ♂, ex Bates Collection (BMNH).

Getta probles Prout, 1918: 427.

Type Locality

Peru (E), Chanchamayo, 1000–1500 m.

Type

Holotype ♀, leg. Watkins (BMNH).

Discussion

Material identified as Getta clite in museum collections has been incorrectly named. I dissected the BMNH type of clite and discovered it to be a species of Erbessa (tribe Dioptini). Getta niveifascia, formerly a synonym of ennia (Bryk, 1930), is thus raised from synonymy. The wing patterns of E. clite (pl. 3) and G. niveifascia (pl. 26) are easily confused, but obvious features of their morphology, such as tympanal structure and wing venation, differ drastically. This misidentification, ascribed to Kirby (1892), was followed by all subsequent authors. Prout (1918) described probles as a subspecies of “Getta clite”. My dissections of the types confirm that probles and niveifascia are conspecific.

Preliminary studies suggest that G. niveifascia occurs throughout the Amazon Basin, occurring at a fairly wide range of altitudes. The type locality, Nauta, is near Iquitos on the Río Marañon at an elevation of 100 meters. Other examples, such as Prout's type of probles, were collected at 1000–1500 meters in the eastern Andes.

However, dissections hint that G. niveifascia could ultimately represent a case analogous to what was discovered for G. baetifica; what appeared at first to be a single species, is in fact a series of cryptic species. French Guiana specimens of G. niveifascia show slight differences from Peruvian material. In males the costa of the valva is straighter and the vesica of the aedeagus is larger. Rather than describe material from French Guiana as distinct, I await future revisionary work and treat all this material as a single species.

Getta niveifascia was reared in French Guiana (table 6) on Passiflora candida (pl. 41B, subgenus Astrophea).

Distribution

Peru (AMNH, BMNH, MUSM); Bolivia (CUIC); Ecuador (BMNH); Colombia (AMNH); Brazil (USNM, VOB); French Guiana (AMNH, BHC, OUMNH); Guyana (CMNH); Venezuela (MNHN).

Dissected

Syntype ♂ (genitalia slide no. JSM-1403); ♂, Peru, La Merced, 2500 ft, 5 Jun 1903, leg. Watkins & Tomlinson, BMNH (genitalia slide no. JSM-1365); ♂, Peru, Madre de Dios, Tambopata Reserve, 200 m, Dec 1994, leg. S. Fratello, day-collecting (genitalia slide no. JSM-1462); ♂, Colombia, Cauca Valley, AMNH (genitalia slide no. JSM-366); ♂, French Guiana, Kaw, PK 31.5, 25 Sep 1992, leg. J.A. Cerda, AMNH (genitalia slide no. JSM-1367); ♂, “French Guiana”, BMNH (genitalia slide no. JSM-1405); ♂, French Guiana, Fourgassié, 20 Jan 1985, no. 98, leg. B. Hermier, AMNH (genitalia slide no. JSM-1524); ♀, Brazil, Santarém, Rio Tapajós, Pará, Nov 1921, USNM (genitalia slide no. JSM-563); ♀ holotype of probles Prout, Peru, Chanchamayo 1000–1500 m, BMNH (genitalia slide no. JSM-1525); ♀, Peru, Moyobamba, 1888, leg. M. de Mathan, BMNH (genitalia slide no. JSM-1366); ♀, French Guiana, Kaw, PK 41, 25 Aug 1994, leg. J.A. Cerda, AMNH (genitalia slide no. JSM-1368); ♀, French Guiana, Matoury, Oct 1981, leg. J.Y. Gallard, BHC (genitalia slide no. JSM-1735).

Getta tica, new species

Figures 293, 294; plates 26, 40B

Diagnosis

Specimens of Getta tica from Costa Rica and Panama, have until now been regarded as Central American examples of G. baetifica (type locality: Angamarca, Ecuador). The two are extremely similar in appearance (pl. 26). Differences include a more intensely colored and wider orange-yellow transverse FW band in G. tica, and a purplish blue, rather than turquoise blue, iridescence. However, the most reliable means for separating the two requires study of their genitalia. The male vesica is much longer in G. baetifica, whereas the ventral plate of the female CB is wider and more rounded in G. tica (fig. 293D). The third taxon in this species complex, G. turrenti, sp. nov, from northern Central America, is even more purple in the FW and HW. Its genitalia are distinctive, showing an especially wide, short vesica (fig. 295C) and a large basal appendix of the female CB (fig. 295D).

Description

Male. FW length  =  19.5 mm. Head: Labial palpus moderately long, porrect, curving slightly upward to immediately below middle of front; Lp1 short, curving upward, bright orange-yellow, with a ventral fringe of long yellow scales, dorsum black; Lp2 slightly longer than Lp1, glossy bluish black, with longer, yellow scales on ventral surface near base; Lp3 short, bullet shaped, bluish black; scales of front charcoal gray with a bluish iridescence; frontal scales long, pointing down from antennal bases, then horizontally to meet at midline; occiput charcoal gray; eye large and round, sparsely covered with microscopic setae; vertex black with a purplish-blue luster, scales long, pointing anteriorly; antenna bipectinate, rami fairly short, tightly spaced; scape with a fanlike tuft of charcoal gray, slightly iridescent blue scales; dorsum of antennal shaft tightly covered with glossy, blackish brown scales.

Thorax: Area of prothorax below proboscis yellow; legs and pleural region covered with long, hairlike, dark gray/iridescent turquoise scales; patagium, tegula, and dorsum covered with long, hairlike, charcoal-gray or iridescent cobalt-blue scales; scales along outer margins of tegula particularly long and fine.

Forewing: (Dorsal) Ground color black to blackish gray (pl. 26); wing veins lighter gray; a patch of iridescent, Windsor blue to violet blue scales in basal fourth, extending from cubitus anteriorly to anal margin posteriorly; an orange-yellow transverse band extending from midpoint of leading edge, slightly behind costa, to immediately short of tornus; transverse band crossing fork of M₁ and Rs₁–Rs₄, as well as fork of M₃ and CuA₁, ending beyond apex of 1A+2A. (Ventral) Ground color gray to blackish gray (pl. 26); no iridescent blue patch at base; a large, ovoid androconial organ near base, covered with elongate, deciduous, yellowish-beige scales; area surrounding androconial patch, extending from DC to slightly below 1A+2A, scaleless, with a cellophane-like appearance.

Hind wing: (Dorsal) Ground color blackish gray to gray (pl. 26); from base to near outer margin covered with iridescent, Windsor blue to violet-blue scales; iridescent area extending from cubitus anteriorly to anal margin posteriorly; a large, ovoid androconial organ at base near anterior margin, covered with elongate, yellowish-beige scales; organ straddling Sc+R, its anterior margin falling short of wing's leading edge, its posterior margin passing through DC to touch cubitus; a wide area surrounding androconial organ scaleless, its surface glistening gray. (Ventral) Ground color blackish gray to charcoal gray (pl. 26); posterior half of wing, from CuA₂ to anal margin, with a faint, bluish iridescence.

Abdomen: Dorsum uniformly charcoal gray, with a Windsor blue to turquoise iridescence; venter dark gray, with a turquoise iridescence.

Terminalia (fig. 293A–C, 293E): Tg8 narrower than Tg7, wider than St8, anterior margin with a wide, shallow mesal excavation, posterior margin with an indistinct, wide, U-shaped mesal excavation; St8 narrower posteriorly, lateral margins slightly constricted in distal third; anterior margin of St8 with an extremely long, thin mesal apodeme, almost as long as rest of sternum, posterior margin with a deep, U-shaped mesal excavation; socii/uncus complex with a fairly wide attachment to tegumen; uncus bent downward near base, distal portion long, thin, bent downward at apex, with a large, laterally compressed dorsal crest along midline near base; socii slightly shorter than uncus, each with a dorsal crest (matching that on uncus) near base, apices acute; ventral margin of genitalia slightly convex; valva bases broadly sclerotized, each with a knob on inner surface ventrally; BO large, outer margin broadly rounded; longitudinal sclerite of BO sinuate, with a large, sclerotized, toothlike ridge at dorsum; a large scale tuft in manica below transtillar arms; base of aedeagus angulate (not rounded), apex gradually narrow; vesica long, slightly narrower than aedeagus, with an almost 90° bend at distal third; vesica with a small, dentate, transverse sclerite at base near apex of aedeagus, a compact group of large, spinelike cornuti on right side before bend, and a large group of straight, spinelike cornuti on dorsal surface in distal third (beyond bend).

Female. FW length  =  21.5–25.0 mm. Head, thorax, and wings similar to male, except: Pleuron with a wide, longitudinal band of orange-yellow from behind eye to mesepimeron; antenna ciliate, shaft widening from base, then gradually tapered; wings longer, broader, and more rounded; androconial organs absent; blue iridescence of FW base and HW more extensive, orange-yellow transverse band of FW wider.

Terminalia (fig. 293D): Tg7 large and wide, slightly narrower at posterior margin, margins simple; St7 narrower than Tg7, slightly narrower posteriorly, margins simple; CB large, rounded, with a long, narrow dorsal sclerite, widening to form a broad Y at its apex, internal surface of sclerite minutely dentate; CB with a large, coriaceous, ventral sclerite near base, its lateral margins broadly rounded; junction between CB and DB forming a small, melanized, laterally compressed appendix on left side; DS attached at junction of DB and CB on right side; DB narrow; posterior margin of PA simple.

Etymology

This species name comes from the moth's larval host plant, Passiflora tica. The word tica is a Costa Rican term for “farmer”.

Distribution

Getta tica occurs on the Caribbean slope of the Cordillera de Talamanca at elevations between 750 and 1100 meters, in regions of high humidity. Established localities in Costa Rica are confined to a small area (fig. 294) either in, or in close proximity to, Parque Nacional Braulio Carrillo. Getta turrenti occurs in southern Mexico and Guatemala (fig. 296), and an as yet undescribed species exists in Honduras (see Discussion for G. turrenti). The southernmost limits of G. tica have yet to be determined. A USNM female from Colón, on the Caribbean coast of Panama, is here provisionally considered to be an example of G. tica. The identity of Getta from the Pacific side of the Cordillera in Panama also remains unresolved.

Biology

In Costa Rica, G. tica is associated with Passiflora tica (pl. 40A, C–E). As is true of the other known hosts of Getta, this plant belongs in the Passiflora subgenus Astrophea (table 6). Getta tica shares this host with the butterfly Heliconius eleuchia (Gómez and Gómez, 1981; DeVries, 1987). Getta tica females have been observed flying between 10 a.m. and 12 a.m., laying eggs on the undersides of new P. tica leaves (J. Corrales, October 2004). The host plants grow along shaded mountain streams on nearly vertical embankments.

On January 23, 2007, Jorge Corrales and I collected eggs of Getta tica, as well as those of Heliconius eleuchia, at the following locality: Costa Rica: Heredia: San Miguel, Bosque Alegre, Laguna de Hule, 750 m, 513200-253300 LN. Two batches of eggs were found on Passiflora tica plants overhanging the Río Cariblanco (pl. 40A). In addition, numerous dead larvae of all instars were observed on P. tica leaves. They appeared to have been killed by a white fungus. The eggs hatched approximately one week later, but these larvae succumbed to the white fungus as they matured. None formed pupae or adults.

Discussion

Although consistent wing-pattern differences appear to exist, the best way to separate G. baetifica and G. tica is by comparison of their genitalia. In both, the vesica of the aedeagus is elongate with a sharp bend part way out. The portion beyond the bend bears a large group of spinelike cornuti on its dorsal surface. This portion of the vesica in G. baetifica is long and relatively thin (longer than one-half the vesica length). In contrast, the portion beyond the bend in G. tica is shorter (fig. 293C), less than one-half the vesica length. In females, the large ventral plate of the CB is shorter and more broadly rounded in G. tica (fig. 293D). These genitalia traits, though subtle, seem to be reliable for separating the two species. Their Astrophea host plants differ as well; G. baetifica has been recorded on Passiflora macrophylla and P. arborea (Ecuador), whereas caterpillars of G. tica feed exclusively on P. tica.

The identity of material from Panama requires additional study. Based on its wing coloring, a female specimen from Colón (USNM) is almost certainly G. tica. This location, on the Caribbean coast, is a logical extension of the Costa Rican habitat for G. tica, on the Caribbean side of the Cordillera Central. Specimens from the western side of the central ridge (USNM) in Panama are of indeterminate status. They potentially represent the northernmost range of G. baetifica, known from the Pacific slope in Colombia and Ecuador.

Holotype

Male (pl. 26). Costa Rica: San José: Parque Nacional Braulio Carrillo, La Montura, 1100 m, larva 1–2 instar coll. 13 Dec 1995, adult 30 Dec 1995, voucher no. 95-JFCM-22, leg. J.F. Corrales (INBio; genitalia slide no. JSM-1519). The type is deposited at the AMNH.

Paratypes

Costa Rica: San José: 1♀, Parque Nacional Braulio Carrillo, La Montura, 1100 m, larva 1–2 instar coll. 11 Dec 1995, adult 29 Dec 1995, voucher no. 95-JFCM-24, leg. J.F. Corrales (INBio); 1♀, La Montura, Braulio Carrillo, 1000 m, 30 Jul 1982, leg. P. DeVries (AMNH; genitalia slide no. JSM-564); 1♀, La Montura, Braulio Carrillo, 1000 m, 27 July 1982, leg. P.J. DeVries (AMNH; genitalia slide no. JSM-1520); 1♀, Parque Nacional Braulio Carrillo, 1050 m, 26 Jun 1980, leg. G.B. Small (USNM); 1♀, 2 Jul 1980, leg. G.B. Small (USNM).

Other Specimens Examined

Panama: Panama: 1♀, Colón, 1500 ft, 23 Feb 1969, leg. G.B. Small (USNM).

Dissected

2♂♂, 2♀♀.

Getta turrenti, new species

Figures 295, 296; plate 26

Diagnosis

Of the three blue-winged Getta species—baetifica, tica, and turrenti—the latter shows the greatest amount of purple (pl. 26); Getta baetifica is more greenish blue, whereas G. tica tends toward cobalt blue. Getta turrenti also shows distinctive features of its genitalia, and these provide the most reliable means for separating it from the others. The male vesica of G. turrenti (fig. 295C) is shorter and wider than in its close relatives. The cornuti at the base of the vesica are larger and more numerous, and the aedeagus itself is wider. At the base of the CB in females, all three species exhibit a small, folded appendix with a thickened membrane. In G. turrenti (fig. 295D) this appendix is much larger than in either G. baetifica or G. tica.

Description

Male. FW length  =  20.5–21.0 mm. Head: Labial palpus long, porrect, curving slightly upward to immediately above clypeus; Lp1 moderately long, curving upward, scales light yellow to lemon yellow, with a loose fringe below, dorsum blue black; Lp2 slightly longer than Lp1, bluish brown to blue-black; Lp3 short, bullet shaped, blue-black; scales of front dark brown with a steely blue iridescence; frontal scales long, pointing down from antennal bases, then horizontally to meet at midline to form a conical frontal tuft; occiput brown; eye large and round, sparsely covered with microscopic setae; vertex brown, scales long, pointing anteriorly; antenna bipectinate, rami long, tightly spaced; scape and dorsum of antennal shaft covered with glossy, purplish-brown scales.

Thorax: Area of prothorax below proboscis yellow; legs closely covered with gray-brown scales with a blue iridescence; pleuron covered with long hairlike and shorter scales, all with a blue iridescence; patagium, tegula, and dorsum covered with long, dark brown to dark gray-brown scales, these with a turquoise blue iridescence; distal scales of tegula long.

Forewing: (Dorsal) Ground color blackish gray; wing veins slightly lighter gray; a patch of iridescent, violet-blue scales in basal fourth, extending from cubitus anteriorly to anal margin posteriorly; an orange-yellow to lemon-yellow transverse band extending from midpoint of leading edge, behind costa, to slightly short of tornus; transverse band crossing fork of M₁ and Rs₁–Rs₄, as well as fork of M₃ and CuA₁, ending past apex of 1A+2A. (Ventral) Ground color gray-brown to charcoal gray; no iridescent area; a large, ovoid androconial organ near base, covered with elongate, deciduous, yellowish-beige scales; area surrounding androconial patch, extending from DC to immediately below 1A+2A, scaleless, with a cellophane-like appearance.

Hind wing: (Dorsal) Ground color gray-brown to dark charcoal gray; basal half covered with iridescent, violet-blue scales; iridescent area extending from cubitus anteriorly to anal margin posteriorly; a large, ovoid androconial organ at base near anterior margin, covered with elongate, yellowish-beige scales; organ straddling Sc+R, its anterior margin falling short of wing's leading edge, its posterior margin passing through DC to touch cubitus; a wide area surrounding androconial organ scaleless, its surface glistening light gray. (Ventral) Ground color gray-brown to charcoal gray; posterior half of wing, from CuA₂ to anal margin, with a faint, bluish iridescence.

Abdomen: Dorsum uniformly brownish gray, with a turquoise iridescence; venter gray, with a turquoise iridescence.

Terminalia (fig. 295A–C, 295E): Tg8 narrower than Tg7, wider than St8, anterior margin with a wide, shallow mesal excavation, posterior margin with an indistinct, wide, U-shaped mesal excavation; St8 narrower posteriorly, lateral margins slightly constricted in distal third; anterior margin of St8 with an extremely long, thin mesal apodeme, almost as long as rest of sternum, posterior margin with a deep, U-shaped mesal excavation; socii/uncus complex with a fairly wide attachment to tegumen; uncus bent downward near base, distal portion long, thin, bent downward at apex, with a large, laterally compressed dorsal crest along midline near base; socii almost as long as uncus, each with a dorsal crest (more acute dorsally than one on uncus) near base, apices acute; tegumen not as tall as vinculum, narrow at attachment of valva below, abruptly wider above, dorsal margin transverse; vinculum narrow, tall; ventral margin of genitalia transverse; valva bases broadly sclerotized, each with a knob on inner surface ventrally; BO large, outer margin broadly rounded; longitudinal sclerite of BO sinuate, with a large, sclerotized, toothlike ridge at dorsum; base of aedeagus angulate (not rounded), apex gradually narrow; vesica relatively short, wider than aedeagus, forming a broad, club shaped appendix distally; vesica with a transverse sclerite at base near apex of aedeagus, a loose group of large, spinelike cornuti on right side near base, and large group of straight, robust, spinelike cornuti on dorsal surface at apex.

Female (fig. 26). FW length  =  21.0–23.0 mm. Head, thorax, and wings similar to male, except: pleuron with a wide, longitudinal light yellow band from behind eye to mesepimeron; antenna ciliate, shaft widening from base, then gradually tapered; wings longer, broader, and more rounded than in male; androconial organs absent; blue iridescence of FW base and HW more extensive.

Terminalia (pl. 295D): Tg7 extremely large and wide, slightly narrower at posterior margin, margins simple; St7 equal in length to Tg7, narrower, lateral margins constricted in distal third, posterior third narrowing slightly; CB relatively narrow; CB with a relatively wide, short dorsal sclerite, broadening at its apex, internal surface of sclerite coarsely dentate; CB with a large, coriaceous, ventral sclerite near base, its lateral margins broadly rounded; junction between CB and DB forming a large, melanized, laterally compressed appendix on left side, appendix with an upright fold on its anterior surface; DS attached at junction of DB and CB on right side; DB wide; posterior margin of PA lobate near dorsum; other features as in genus description.

Etymology

This species is named in honor of Rafael Turrent, who kindly made his collection of Mexican Lepidoptera available to me for study. His material yielded many interesting Dioptinae, including most of the known examples of Getta turrenti.

Distribution

Getta turrenti is known from the Caribbean side of the Sierra Madre (fig. 296), in an area spanning roughly 1000 km. Its habitat extends from Orizaba Mexico, E-SE of Veracruz, south to Cayuga Guatemala. Getta turrenti has been captured at four Mexican sites—Orizaba and Oaxaca in the state of Veracruz, as well as Chajul and San Quintin. The two latter localities are in Chiapas near the Guatemala border, on the Río Lacantum. All of these specimens were collected at relatively low elevations; Chajul is at 150 meters altitude.

A female from Truxillo, Honduras (AMNH; F.C. Nicholas Collection), on the Caribbean coast, represents a fourth iridescent blue Getta species. Its FW stripe is unusually wide, its HW is strikingly purple, and its genitalia (JSM-1741) are distinctive. Known from only a single specimen, I refrain from describing this taxon until additional material becomes available.

Discussion

Getta turrenti can be readily distinguished from G. tica and G. baetifica: The head, thorax, and wing ground color of G. turrenti tend toward dark brown or blackish brown rather than black; the iridescent FW and HW areas are slightly more violet colored, especially at their peripheries; the male antennal pectinations are noticeably longer in G. turrenti than in either G. tica or G. baetifica. When the genitalia of these three species are compared, obvious differences set G. turrenti apart.

In all likelihood, caterpillars of G. turrenti feed on whatever Astrophea species are endemic to that region of Central America. It will be important to search along streams and other likely Astrophea habitats, in hopes of discovering the life history of this species.

Holotype

Female (pl. 26). Mexico: Chiapas: San Quintin, 16.4°N, 91.35°W, 14–18 Oct 1977, leg. Peter Hubbell. The type is deposited at the LACM.

Paratypes

Mexico: Chiapas: 1♂, Chajul, Río Lacantum, May 1981, leg. R. Turrent (LACM; genitalia slide no. JSM-565); 2♂♂, Chajul, Río Lacantum, leg. R. Turrent (ARTC); 2♀♀, San Quintin, 16.4°N, 91.35°W, 14–18 Oct 1977, leg. Peter Hubbell (LACM; genitalia slide no. JSM-1733). Veracruz: 1♀, Tuxpango, 5 km S Orizaba, Aug 1938, leg. Mario del Toro (ARTC); 2♂♂, Oaxaca, Río Sarabía, S Tuxtepec, leg. R. Turrent (ARTC).

Other Specimens Examined

Guatemala: 1♂, Cayuga, May, Schaus & Barnes Collection (USNM; genitalia slide no. JSM-569).

Dissected

2♂♂, 1♀.

Getta unicolor (Hering), new combination

Figures 289A, 297; plate 27

Polyptychia unicolor Hering, 1925: 530.

Type Locality

Peru, Yurimaguas.

Type

Holotype ♀ (ZMH).

Discussion

This species had been a mystery since my research on the Dioptinae first began. Hering (1925) described unicolor in Polyptychia, and it was retained there by Bryk (1930). Several years ago, I had identified a French Guyana female (BHC) as Polyptychia unicolor. That female and the female holotype were thought to be the only known specimens. However, during a visit to the BMNH (March 2005), a male of what I thought was an undescribed Getta species was discovered among their unidentified material. Its membership in Getta was clear—the moth bore the characteristic FW and HW androconia (pl. 27). Upon further study, this male, from Bartica, Guyana, exhibited a set of unusual characters: First, FW vein M₁ is long stalked with the radial sector, a condition that does not occur in Polyptychia (fig. 298E) but provides one of the diagnostic features for Getta (fig. 289F, H). Second, the apex of Lp2 bears an extremely wide, dense, fanlike tuft (fig. 289A). Third, the moth has ridgelike tufts running vertically along the lateral margins of the front, extending from the clypeus up to the vertex between the antennal bases (fig. 289A). This set of traits is unique for the Dioptinae. Subsequent comparisons showed that the French Guiana unicolor female exhibits the same wing venation, tufted labial palpus, and frontal crests as the Guyana male. This female was recompared with the unicolor type, and the mystery was solved—unicolor Hering (1925) belongs in Getta as a new combination, rather than in Polyptychia. Had Hering seen a unicolor male, he would not have made that error.

With this knowledge in hand, I searched available museum material and located a total of eight G. unicolor specimens (2♂♂, 6♀♀ including the type). Even though the moth is rare, it shows a surprisingly broad distribution. Getta unicolor can be found across the Guyana Shield from Venezuela east to French Guiana, and seemingly occurs throughout the Upper Amazon Basin, from Peru north to Colombia. As additional specimens of G. unicolor accumulate, it will be interesting to determine whether the species is truly widespread, or alternatively, whether it comprises a species complex, similar to the case of G. baetifica and its close relatives.

Getta unicolor (pl. 27) is unmistakable. Its tufted palpi (fig. 289A) are a giveaway. The wings show a faint blue iridescence, but not the dazzling blue found in G. baetifica. Furthermore, G. unicolor is much larger (♂ FW length  =  21.0–22.5 mm) than either G. niveifascia (♂ FW length  =  11.0–19.0 mm) or G. ennia (♂ FW length  =  17.0 mm), the only other species with which its wing pattern might be confused. The genitalia in both sexes of G. unicolor are also unique (fig. 297). It is interesting to note an unusual group of small cornuti on the male vesica. When viewed with a dissecting scope, these are shaped like scoli, found on the bodies of saturniid caterpillars. Such scolus-like cornuti occur nowhere else in the Notodontidae as far as I am aware.

Distribution

Peru (BMNH, ZMH); Ecuador (FNHM); Colombia (BMNH); Venezuela (LACM); French Guiana (BHC, MNHN); Guyana (BMNH).

Dissected

♂, Guyana, Bartica, BMNH (genitalia slide no. JSM-1523); ♀, Colombia, Florida, Río Putumayo, Apr 1932, leg. G. Klug, BMNH (genitalia slide no. JSM-1526); ♀, French Guiana, C. Bar Collection, BMNH (slide no. JSM-1734).

The following species has been transferred from Getta: clite Walker to Erbessa Walker

POLYPTYCHIA C. AND R. FELDER, 1874

Figures 298–Figure 299Figure 300Figure 301Figure 302Figure 303304; plate 27

Figure 301

Scanning electron micrographs of Polyptychia hermieri (♂), HW dorsal surface. A, surface of deciduous androconium from central pocket; B, anal margin, showing hairlike scales covering androconial fold; C, fold of anal margin; D, deciduous wooly scales from androconial fold; E, surface of wooly scale; F, apex of tibial spur, showing serrate margins.

Figure 302

Genitalia of Polyptychia fasciculosa (♂ JSM-1359, ♀ JSM-1360). A, ♂ genitalia; B, ♂ St8; C, ♂ Tg8; D, aedeagus; E, ♀ genitalia (illustration by A. Trabka).

Figure 303

Genitalia of Polyptychia hermieri, sp. nov. (♂ JSM-1505, ♀ JSM-716). A, ♂ genitalia; B, ♀ genitalia; C, ♂ St8; D, aedeagus; E, ♂ Tg8, dorsal view.

Figure 304

Known distributions of newly described South American Josiini: Josia neblina, Notascea brevispinula, Notascea loxa, Phintia broweri, Polyptychia hermieri, and Scea torrida.

Polyptychia C. and R. Felder, 1874: pl. 104, fig. 7. Type species: Polyptychia fasciculosa C. and R. Felder, 1874 (by monotypy).

Diagnosis

Males of Polyptychia are highly distinctive, being the only dioptine with conspicuous white, hairlike androconia along the length of the hind femora and tibiae (pl. 27). Males are also unique in possessing two androconial organs on the HW (fig. 298E)—the first located immediately beyond the DC along the cubital vein, and the second in a large fold on the anal margin. Although both organs enclose deciduous androconia, they are structurally different. The one along the anal margin contains long gray bristles (fig. 301B, C), and under those, a large pocket of white, wooly, apparently sticky scales (fig. 301D, E). The second pouch, beyond the DC, encloses an elaborate androconial organ containing thickened, fleshy-looking deciduous scales (figs. 300B–F, 301A). At its basal end is a brush of long, thick, bristlelike scales (figs. 299, 300A). No other dioptine possesses a two-part androconial system of this type.

Females can be distinguished by their small yellow spot immediately below the HW apex (pl. 27). This spot varies in size, being conspicuous in some species but almost absent in others. It is usually slightly larger on the HW ventral surface. Another trait for distinguishing Polyptychia (both sexes) from other Josiini is the configuration of the FW radials (fig. 298E); Rs₄ branches from the radial sector below veins Rs₁–Rs₃. This also occurs in a single species of Getta—G. unicolor (pl. 27)—but nowhere else in the Dioptinae. Polyptychia and G. unicolor can be separated because M₁ is long stalked with the radial sector in the latter, whereas in Polyptychia M₁ arises from the anterolateral angle of the DC. Their androconial systems differ dramatically.

Redescription

Male. FW length  =  18.5–23.0 mm. Head (fig. 298A–D): Labial palpus long, porrect, gradually upturned to immediately above clypeus; Lp2 slightly longer than Lp1, Lp3 somewhat elongate, bullet shaped; clypeus scaleless; frontal scales pointing downward from antennal bases, ventral few pointing horizontally toward midline above clypeus; eye large, bulging outward, gena narrow, scaleless; scales on vertex pointing anteriorly, with a faint part between antennal bases; antenna bipectinate, rami relatively short, approximately 20 terminal annulations simple.

Thorax (fig. 301F): Epiphysis relatively short, narrow, its apex falling well short of tibial apex; femur and tibia of metathoracic leg covered with short scales on outer surface, inner surface covered with a dense brush of long, white, hairlike androconia (pl. 27), longest and most conspicuous on posterior surfaces; tegula long, dorsal portion narrow, ventral process acute, the two sections divided by a strong transverse sulcus; metathoracic tympanum of the kettledrum type, membrane large, enclosed, oriented horizontally.

Forewing (fig. 298E; pl. 27): Shaped like an elongate triangle, drawn out toward apex, apical angle acute, outer margin almost straight; vein Rs₁ stalked with Rs₂–Rs₄; Rs₄ arising from radial sector below veins Rs₁–Rs₃; Rs₁–Rs₃ in the pattern 2+[3+4]; M₁ arising from anterolateral angle of DC near base of Rs₁–Rs₄, UDC extremely short; DC less than one-half FW length; veins M₃ and CuA₁ stalked; ground color dark brown to blackish brown, with a yellow transverse band beyond DC.

Hind wing (figs. 298E, 299, 300, 301; pl. 27): Triangular, apical angle acute; Rs and M₁ fused; two complex androconial organs present: the first, a deep, slitlike pouch located beyond DC, situated between M₂ and M₃+CuA₁, ending before HW outer margin; slit enclosing bat shaped, deciduous androconia, these covered by a long brush of coarse scales arising together from anterior end of slit; second androconial organ comprising a pouch, located along a deeply folded anal margin, edged with a row of extremely long, hairlike scales; pouch enclosing a large packet of white, wooly, deciduous scales; veins M₃ and CuA₁ obscured by central (anterior) androconial pouch, their apices appearing as small folds near margin.

Abdomen: Elongate, apex acute; dorsum slightly darker than venter, no stripe present along pleuron.

Terminalia (figs. 302A–D, 303A, 303C–E): Tg8 short, slightly wider anteriorly than posteriorly, anterior margin with a pair of broad, anterolateral apodemes; St8 slightly longer than Tg8, with acute anterolateral angles, sternum somewhat wider anteriorly than posteriorly; socii/uncus complex with a relatively wide attachment to tegumen; socii laterally compressed, variable in size and shape (see description of P. hermieri); tegumen as tall as vinculum, relatively narrow, forming hump shaped shoulders at dorsum; vinculum narrow; saccus absent; valva large, roughly forming an upright, elongate oval; BO large, occupying most of valva; pleats of BO on inner surface enclosing fine, hairlike androconia; a ventral pocket on valva's lateral surface enclosing long, thick, bristlelike androconia, the base of each set in a darkened socket; costa narrow at base, becoming much wider distally, forming large scoop-shaped (P. hermieri) or flattened (P. fasciculosa) processes; apex of valva lightly sclerotized, curled inward; region above BO and below apex membranous, with fine, transverse striations; transtillar arms narrow, oriented either slightly downward or horizontally, meeting to form a small concave structure in anellus at midline; spiculate patch above transtilla (some Josiini) absent; aedeagus broad, anterior end slightly constricted, apex with a tiny ventral tooth; vesica extremely large, broadly transverse, covered with long, spinelike cornuti; ductus ejaculatorius simplex sinuate, curled.

Female. FW length  =  19.0–25.0 mm. Body characters as in male except: antenna ciliate; FW longer and broader, less acutely angled at apex; HW broader, lacking androconial organs; apex with a small, yellow, comma-shaped submarginal spot, its anterior margin touching Rs, its posterior margin touching M₃; abdomen much broader, truncate.

Terminalia (figs. 302E, 303B): Tg7 longer than Tg6, similar in length to St7, lateral margins broadly excavated near midpoint, anterior margin simple, posterior margin simple, with a wide sclerotized band; St7 slightly narrower at posterior margin than at anterior one, otherwise unmodified; Tg8 sclerotized posterior margin rounded, forming a rooflike structure over PA; AA long; A8 pleuron lightly sclerotized; PA small, membranous, bearing long setae, PA margins with a prominent dorsal lobe; PP long, thin, curving downward; area of PVP extremely wide, lightly sclerotized, margins obscure, surface spiculate; DB short and wide, dorsoventrally compressed; body of CB roughly ovoid, with numerous melanized folds in membrane; base of CB with a large, wide, sclerotized, flattened appendix on dorsum, its surface coriaceous, appendix somewhat U-shaped in posterior view, DS attached at its posterior margin; signum large, shield shaped, concave in lateral view, its internal surface coarsely dentate along lateral margins.

Distribution

Polyptychia is strictly South American. The group is most common in the Upper Amazon Basin of Brazil and in northern South America, especially the Guyana Shield. Much of the material of P. fasciculosa in museum collections was captured near the turn of the 20th century at Suapure, in central Venezuela on the Orinoco River. A single example of P. hermieri, sp. nov., is known from Trinidad (CMNH). Polyptychia is an inhabitant of lowland rainforests; to my knowledge, not a single specimen has been collected above 500 meters in elevation. The southernmost Polyptychia record is from Fazenda Rancho Grande (Rondônia), at the headwaters of the Rio Madeira.

Biology

Polyptychia, like its close relative Getta, is associated with Passiflora species in the subgenus Astrophea (table 6). Polyptychia hermieri, sp. nov., was reared in French Guiana on Passiflora candida (pl. 41B). It appears likely that the entire clade to which Getta and Polyptychia belong (Clade 19; fig. 7; see below) will be found in association with Astrophea. Host plants remain unknown for Phavaraea, the other genus of that lineage.

Caterpillars of P. hermieri (pl. 39H) resemble Getta larvae (pl. 40B), but show a more intricate series of blackish transverse lines across the abdominal dorsum. They are large, feeding along the edges of the tough, dry leaves of P. candida.

Discussion

According to my phylogenetic analyses, Polyptychia belongs in a subclade of the Josiini (Clade 19; fig. 7) that includes Getta and Phavaraea, and totals 12 species (appendix 2). Within the clade, Polyptychia arises above Getta as the sister group to Phavaraea. These results should be considered provisional, however. Certain apomorphies imply that Getta and Polyptychia might instead be sister groups. Potential synapomorphies for these two are their diamond-shaped male hind wings, and the fusion of HW veins Rs and M₁ (figs. 289H, 298E). To truly understand relationships within this lineage, it will be necessary to treat all 12 species in a single analysis. Ideally, such a study would employ characters from immature stages and DNA.

The androconial systems in Polyptychia are perplexing. Nowhere else in the Dioptinae do androconia occur on the legs, and although HW androconial organs occur in Getta and Phavaraea, it is difficult to imagine that these are homologous with the elaborate central pocket of Polyptychia. Perhaps developmental research could provide insights into this remarkable phenomenon.

Type Locality

Colombia, “Bogotá”.

Type

Holotype ♂ (BMNH).

Thyrgis ceron 142Druce, 1899: 295.

Type Locality

Colombia, San Martín, Llanos of Río Meta.

Type

Syntype ♀, leg. G.D. Child (BMNH).

Discussion

All previous authors have recognized a single species of Polyptychia—fasciculosa—occurring throughout South America. It was not until I located the Felder type of fasciculosa at the BMNH (pl. 27) that the existence of two species became clear. The type is interesting for several reasons: First, it is large (FW length  =  23.0 mm) compared to most examples of so-called fasciculosa in collections. Second, the Felder specimen is almost certainly the corresponding male of ceron 142Druce (1899), the female BMNH type of which is also large (FW length  =  25.0 mm). Their wing patterns and body coloration are identical. The type locality of P. fasciculosa, “Bogotá”, is widely cited in Felder descriptions, but many Lepidopterists doubt the veracity of this report. According to them, the true provenance of Felder types could encompass a range of Colombian localities. However, the BMNH type of ceron provides more detail. That specimen was collected in Colombia at San Martín (“Llanos of Río Meta”). This locality is at 400 meters elevation near the headwaters of the Río Meta, which eventually joins the Orinoco in Venezuela.

With the identity of P. fasciculosa firmly established, I then compared dissections of so-called “fasciculosa” across numerous South American localities. These revealed the existence of a second Polyptychia species, described below as Polyptychia hermieri, sp. nov. My research suggests that P. fasciculosa occurs in the Upper Amazon Basin of Brazil, from the headwaters of the Amazon north to the Llanos drainage of the Meta and Orinoco rivers, through Colombia and Venezuela. Polyptychia hermieri, on the other hand, occurs from Trinidad and French Guiana south to the lower portion of the Amazon River. Its western limits in Brazil, though not precisely known, extend upriver at least as far as Santarém. Means for separating the two taxa are provided in the Polyptychia species key (above). Their male genitalia are dramatically different.

Adding to the complexity, a third species of Polyptychia exists. I know this taxon from a single male collected at Rondônia, Brazil. Label data for that specimen are as follows: Brazil, Rondônia, 62 km S Ariquemes, 165 m, Faz. Rancho Grande, 10°32′S, 62°48′W, 27 Aug–8 Sep 1994, leg. Ron Leuschner, LACM (genitalia slide no. JSM-1739). I refrain from describing this taxon until additional material becomes available.

Distribution

Brazil (AMNH, BMNH, CMNH, CUIC, LACM, USNM, ZMH); Peru (MPM, BMNH); Colombia (BMNH, USNM, ZMH); Venezuela (BMNH, USNM).

Dissected

♂, Venezuela, Suapure, 25 Apr 1900, USNM (genitalia slide no. JSM-1359, wing slide no. JSM-1361); ♂, Brazil, São Paulo de Olivença, Mar 1883, leg. M. de Mathan, BMNH (genitalia slide no. JSM-1504); ♂, Venezuela, 14 Apr 1899, USNM (genitalia slide no. JSM-1506); ♀, Colombia, Muzo, 1915, Apollinaire Coll., USNM (genitalia slide no. JSM-1508); ♀, Venezuela, Suapure, 11 Jul 1899, USNM (genitalia slide no. JSM-1742).

Polyptychia hermieri, new species

Figures 298–Figure 300301, 303, 304; plates 27, 39H [EX]

Diagnosis

Although Polyptychia hermieri is superficially similar to P. fasciculosa (pl. 27), it is distinguished by its darker wing and body coloring, and its more richly pigmented yellow transverse band. Three other characters are useful for separation: the frontal scales of P. hermieri show a strong, iridescent blue luster when observed from above, whereas in P. fasiculosa the front is brown, with only a faint hint of blue. Second, males differ because the brush of long, bristlelike scales in the central HW androconial patches (figs. 298E, 299) are light brown in P. fasciculosa, but more cream colored in P. hermieri. Finally, the two can often be separated by their size. The FW length of P. hermieri is generally smaller that that of P. fasciculosa (see species key), but overlap exists. When their genitalia are compared (figs. 302, 303), the species are instantly recognizable.

Description

Male. Forewing length  =  18.5–21.0 mm. Head (fig. 298A–D): Lp1 yellow on outer surface, black on inner surface, with a long, loose fringe below, inner black scales of fringe longer than yellow ones; Lp2 tightly scaled with black to dark gray black scales, a few yellow scales near base; Lp3 dark gray black; scales of front blackish brown, with an iridescent blue tinge; occiput blackish brown above, a patch of long, yellow scales below eye; an additional patch of long yellow scales below palpus and above procoxa; vertex blackish brown, with an iridescent blue cast, especially near antennal bases; scape blackish brown; antennal shaft wide, dorsum densely covered with blackish brown scales.

Thorax (fig. 301F): Prothoracic and mesothoracic legs blackish brown on inner surfaces, gray-brown on outer ones; coxa of metathoracic leg covered with long, blackish brown scales; femur and tibia of metathoracic leg covered with short, gray-brown scales on outer surfaces, inner surfaces covered with a dense brush of long, white, hairlike androconia, longest and most conspicuous on posterior surfaces of tibiae; spurs gray to gray-brown; pleuron covered with blackish brown, hairlike scales, having a slight iridescent blue tinge; patagium black to blackish brown; tegula blackish brown, fringed with concolorous long, hairlike scales; dorsum dark brown to blackish brown.

Forewing (fig. 298E; pl. 27): (Dorsal) Ground color dark blackish brown, almost black; a wide, yellow to orange-yellow transverse band beyond DC, band starting behind costa anteriorly and ending slightly short of tornus posteriorly; inner and outer margins of transverse band irregular, lightly irrorated with black scales. (Ventral) Similar to dorsal surface, except ground color lighter, blackish brown to dark chocolate brown; transverse band wider than on dorsal surface, lighter yellow.

Hind wing (figs. 298E, 299, 300, 301A–E; pl. 27): (Dorsal) Ground color blackish brown, slightly lighter in tone than FW, wing surface less densely scaled; anterior (central) androconial organ enclosing brownish-yellow, bat-shaped, deciduous androconia, these covered by coarse bristles; second androconial organ along anal margin edged with a row of extremely long, smoky gray, hairlike scales, pouch enclosing a large packet of white, wooly deciduous scales. (Ventral) Ground color uniformly dark brown; slitlike androconial pouch represented below by a large, irregular fold.

Abdomen: Uniformly dark gray-brown to blackish brown above, with an iridescent blue tinge; color slightly lighter gray-brown on venter.

Terminalia (fig. 303A, 303C–E): Anterolateral apodemes of Tg8 truncate; posterior margin of Tg8 membranous, slightly concave; anterior margin of St8 with a broadly forked mesal apodeme, posterior margin with a shallow, U-shaped excavation; uncus broad, robust, beaklike, laterally compressed above to form a large dorsal crest; socii wide, laterally compressed, with a broad, crestlike flange at base, apices forming wide, scooplike structures; ventral margin of genitalia simple, almost horizontal; distal portion of costa forming a broad, scoop-shaped process, apex itself lightly sclerotized, tapered; arms of transtilla oriented downward; anterior end of aedeagus forming a large rounded knob, apex with a tiny ventral tooth; vesica extremely large, broadly transverse, oriented almost horizontally (more steeply angled in P. fasciculosa), a transverse band of short, spinelike cornuti on right side near base (not present in P. fasciculosa).

Female (pl. 27). Forewing length  =  19.0–21.5 mm. Body characters similar to male except HW with a comma shaped yellow spot below apex; androconia absent.

Terminalia (fig. 303B): Tg8 sclerotized except for a membranous seam along midline; AA long, recurved; DB with a strong fold anteriorly, near base of CB; body of CB roughly ovoid; coriaceous, U-shaped appendix larger and taller than in P. fasciculosa, its dorsal angle more acute; signum with a prominent, external ridge along midline.

Etymology

This species is named in honor of Bernard Hermier, from Cayenne, French Guiana. Bernard kindly hosted my visit to French Guiana in 1994, during which we found larvae of Polyptychia hermieri feeding on Passiflora candida in disturbed habitat along the road from Cayenne to Coralie. Bernard's spectacular Lepidoptera collection contains paratypes of P. hermieri, not to mention untold material of rare and undescribed Lepidoptera from French Guiana. Study of his collection was crucial to completion of this publication.

Distribution

My studies reveal that P. hermieri is not restricted to French Guiana, but also occurs in northeastern Brazil (fig. 304). Its distribution extends from French Guiana southeast to the mouth of the Amazon, then west upriver at least as far as Obidos and Santarém. It has also been collected on the Rio Tapajós, which joins the Amazon at Vila Franca and Santarém. Dissections of material from further upriver at São Paulo d'Olivença, on the Rio Solimões, show these to be examples of P. fasciculosa. The latter thus appears to be more western in distribution. A female from Trinidad bears a wing pattern and genital morphology (JSM-1744) suggesting it to be P. hermieri, which extends the species' range west across the Guyana Shield.

In addition to the material listed below, there are numerous examples of P. hermieri from French Guiana and Brazil in the BMNH collection. Unfortunately, I was not able to obtain label data for those specimens, all of which are relatively ancient.

Discussion

When larvae of P. hermieri were discovered in French Guiana (May 1994), it was my assumption that we had found P. fasciculosa. Over 10 years later (March 2005), it became clear that we had collected a new species, described here as P. hermieri.

Distinctions between the two species first became apparent through comparison of male genitalia, which differ in the shape of St8 (figs. 302B, 303C), as well as the shape of the uncus, socii and valval costa (figs. 302A, 303A). The vesica also differs (figs. 302D, 303D), with that of P. hermieri being oriented nearly horizontal to the aedeagus, but that of P. fasciculosa oriented at a steeper angle. Female differences are less obvious. These involve the configuration of the large, coriaceous sclerite at the base of the corpus bursae (figs. 302E, 303B). Characters of the vestiture useful for separation are noted in the Diagnosis (above).

Adults and larvae of P. hermieri were discussed and figured in Miller (1996), although there they were then unknowingly referred to as P. fasciculosa. The caterpillars (pl. 39H) feed on Passiflora candida (pl. 41B; table 6).

Holotype

Male (pl. 27). French Guiana: Cayenne: Pk 57.5 on road to Coralie, 14 May 1994, leg. J.S. Miller, reared from larva on Passiflora candida, emerged Jun 4 1994. The type is deposited at the AMNH.

Paratypes

French Guiana: Cayenne: 1♂, Pk 57.5 on road to Coralie, 14 May 1994, leg. J.S. Miller, C. Snyder & L.D. Otero, reared from larva on Passiflora candida, emerged Apr 1994 (AMNH); 1♂, Kaw, Pk 57, 4 Oct 1996, á vue, leg. J.A. Cerda (BHC); 1♀, Kaw, Pk 37.5, 1 Oct 2000, á vue, leg. J.A. Cerda (BHC); 1♀, Kaw, Pk 37.5, 24 Sep 2000, á vue, leg. J.A. Cerda (BHC); 1♀, Kaw, Pk 38, 12 Oct 1996, leg. M. Duranron (BHC); 1♂, Fourgassié, 20 Jan 1985, leg. B. Hermier (AMNH; genitalia slide no. JSM-715); 1♂, Pk 57.5 on road to Coralie, 14 May 1994, leg. J.S. Miller, reared from larva on Passiflora candida, emerged 4 Jun 1994 (AMNH; genitalia slide no. JSM-1505, wing slide no. JSM-1669); 1♀, Piste de Coralie, km 10, 7 Mar 1992, at light, leg. B. Hermier (AMNH; genitalia slide no. JSM-716).

Other Specimens Examined

French Guiana: Cayenne: 1♀, Cayenne, CAS (genitalia slide no. JSM-1507). Brazil: Pará: 1♂, 1♀, Itaituba, Rio Tapajós, Mar, leg. H. Fassl (BMNH; male genitalia slide no. JSM-1509); 1♀, Serra Tiracambu, 450 m, Jun 2003, á vue, leg. P. Jauffret (BHC); 1♀, Taperinha, leg. Fassl (USNM; genitalia slide no. JSM-1360, wing slide no. JSM-1362); 2♂♂, Obidos, Dognin Collection (USNM); 3♀♀, Santarém, leg. Fassl (USNM); 1♀, Braganza, leg. H.B. Merrill (USNM); 1♀, Barreiras, leg. Fassl (USNM); 2♂♂, 1♀, L. de Villafranca (USNM); 1♂, 1♀, Rio Manes, Dognin Collection (USNM). Trinidad: 1♀, Holland Collection (CMNH; genitalia slide no. JSM-1744).

Dissected

4♂♂, 3♀♀.

The following has been removed from Polyptychia: unicolor Hering to Getta Walker

PHAVARAEA WALKER, 1854

Figures 305–Figure 306Figure 307Figure 308Figure 309310; plates 27, 28

Figure 305

Morphology of Phavaraea (♂♂). A, head of P. rejecta, lateral view; B, head of P. rejecta, lateral view; C, head of P. rejecta, frontal view; D, head of P. rejecta, posterior view; E, head of P. dilitata, lateral view; F, head of P. dilatata, frontal view; G, head of P. dilatata, posterior view (illustration by J.S. Miller).

Figure 306

Morphology of Phavaraea (♂♂). A, wings of P. rejecta; B, wings of P. dilatata (illustration by J.S. Miller).

Figure 307

Genitalia of Phavaraea dilatata (♂ JSM-346, ♀ JSM-347). A, ♂ genitalia; B, ♂ St8; C, aedeagus; D, ♂ Tg8; E, ♀ genitalia (illustration by A. Trabka).

Figure 308

Genitalia of Phavaraea poliana (♂ JSM-303, ♀ JSM-304). A, ♂ genitalia; B, ♀ genitalia; C, ♂ Tg8; D, aedeagus; E, ♂ St8, ventral view.

Figure 309

Male genitalia of Phavaraea rectangularis (JSM-1775). A, genitalia; B, aedeagus; C, Tg8; D, St8.

Figure 310

Genitalia of Phavaraea rejecta (♂ JSM-1353, ♀ JSM-1355). A, ♂ genitalia; B, ♀ genitalia; C, aedeagus; D, ♂ Tg8; E, ♂ St8 (illustration by A. Trabka).

Phavaraea Walker, 1854: 315. Type species: Centronia rejecta Geyer 1832 (in 254Hübner, 1832) 4: 18, [pl. 115], figs. 663, 664 (by monotypy).

Sagaris Walker, 1854: 313. Type species: Josia styge Walker, 1854 (by monotypy); J. styge is a junior subjective synonym of Centronia rejecta Geyer, 1832; a junior homonym of Sagaris Panzer (Insecta, Hymenoptera; see Watson et al., 1980).

Milodora Boisduval, 1870: 85. Type species: Centronia rejecta Geyer, 1832 (by subsequent designation by Prout, 1918: 426); a junior objective synonym of Phavaraea Walker, 1854.

Scedros Walker, 1854: 316. Type species: Josia dilatata Walker, 1854 (by monotypy).

Apocinesia Bryk, 1930: 55. Type species: Josia dilatata Walker, 1854 (by original designation); a junior objective synonym of Scedros Walker, 1854 (Watson et al., 1980). Revised synonymy.

Diagnosis

Phavaraea, which contains the largest species in the Josiini—female FW lengths of P. rejecta reach 28.0 mm—belongs in a clade with two additional genera, Getta and Polyptychia (Clade 19; fig. 7). All of these taxa exhibit generally similar, simple wing patterns (pls. 26–Plate 2728). Males of this clade exhibit a startling array of androconial systems, located on various parts of the body. The HW of Phavaraea and Polyptychia possesses a pocket of androconia enclosed in a deeply folded anal margin (figs. 298E, 306A). Getta stands apart. There, the androconial system is two-parted (fig. 289H), with a region of scent scales located along the FW anal margin (ventral surface), as well as a region on the HW anterior margin (dorsal surface).

Thus, presence of an androconial fold along the HW anal margin indicates membership in either Phavaraea or Polyptychia. These genera can then be separated by wing venation. Phavaraea exhibits the FW radial venation typical of most Dioptinae, where Rs₄ arises from the radial sector well beyond the base of Rs₁ (fig. 306A, B). Polyptychia shows an apomorphic configuration, with Rs₄ arising basal to Rs₁ (fig. 298E). Females of the three genera are slightly more difficult to identify. Getta can be distinguished because the FW DC is short (fig. 289F), and HW veins Rs and M₁ are fused. The remaining challenge is to separate females of Phavaraea and Polyptychia. Again, the position of FW vein Rs₄ relative to Rs₁, alluded to for males, applies.

Redescription

Male. FW length  =  18.0–27.5 mm. Head (fig. 305): Labial palpus long, reaching upward to above middle of front; Lp1 moderately long, wide, slightly curved, sometimes with a short ventral ridge of scales; Lp2 long and straight, from one to one and a half times as long as Lp1, tightly scaled, in P. rectangularis with a bladelike ventral ridge of scales; Lp3 short to somewhat elongate, drooping slightly downward; clypeus narrow, scaleless; frontal scales long, erect, somewhat loose, swooping downward from antennal bases, then coming together to form a ridgelike tuft above clypeus at midline; eye large, bulging outward, anterior portion of gena forming a scaleless triangle, posterior portion absent; scales on vertex short, semierect, without a part along midline; antenna bipectinate, rami short to moderately long (mesal row of rami much shorter than lateral row in P. rectangularis), terminal 15–30 annulations simple; antennal shaft wide in basal half, then gradually narrowed to apex.

Thorax: Epiphysis long and narrow, almost four-fifths length of tibia, falling slightly short of tibia apex; legs tightly covered with short scales; pleuron with a mixture of short and long scales; tegula long, dorsal portion narrow, ventral process acute, these two sections divided by a strong transverse sulcus, central portion tightly scaled, margins fringed with hairlike scales; metathoracic tympanum kettledrum shaped, membrane large, enclosed, oriented horizontally.

Forewing (fig. 306A, B; pls. 27, 28): Elongate, apical angle acute, outer margin strongly convex, tornus rounded; Rs₁ arising from radial sector; Rs₂–Rs₄ in the pattern [2+3]+4, Rs₄ arising well beyond Rs₁; M₁ arising from anterolateral angle of DC near base of radial sector, or short stalked with Rs₁–Rs₄ (P. dilatata); DC approximately one-half FW length; M₃ and CuA₁ short stalked, or arising together from posterolateral angle of DC; ground color dark brown to blackish brown, with a lemon-yellow or white transverse band crossing distal margin of DC.

Hind wing (fig. 306A, B; pls. 27, 28): Shape highly variable, outer margin smoothly convex (P. poliana, P. rejecta), gently scalloped (P. rectangularis) or greatly expanded (P. dilatata); a large fold along anal margin forming an androconial pouch, this enclosing a set of long, bristlelike scales, as well as a covering of shorter, fleshy scales; vein 3A obscured by androconial pouch.

Abdomen: Elongate, or moderately so; dorsum tightly covered with short scales, these having a slight iridescent blue tinge; no stripe present along pleuron; P. dilatata with four pencils of greatly elongate, dark brown to black bristles arising from pleuron on A4–A7, these extending almost to apex of A8; venter densely covered with extremely short, mattlike scales, these either white (P. rectangularis), blackish purple (P. poliana, P. rejecta), or gray-brown (P. dilatata).

Terminalia (fig. 307A–D, 308A, 308C–E, 309, 310A, 310C–E): Tg8 moderate in length, wider anteriorly than posteriorly, anterior margin concave, with a pair of short apodemes at lateral angles, posterior margin slightly concave, or with a large, deep U-shaped mesal excavation (P. rectangularis); St8 longer than Tg8, widest at anterior margin; anterior margin of St8 broadly convex, bearing a mesal apodeme of varying length and shape; posterior margin of St8 with a small narrow, or large deep, V-shaped mesal notch; socii/uncus complex narrow, widest at attachment to tegumen; socii thin, shorter than uncus, curving slightly downward or gently sinuate; uncus long and narrow, curved downward near apex or strongly recurved; tegumen variable in height, shorter than vinculum (P. poliana, P. rejecta) or taller than vinculum (P. dilatata, P. rectangularis); vinculum narrow; ventral margin of genitalia slightly convex; juxta small, crescent shaped; valva extremely large, membranous, upright, apices extending dorsally above uncus; BO occupying most of valva; pleats of BO on valva inner surface enclosing fine, hairlike androconia; a ventral pocket on valva's lateral surface enclosing long, thick, bristlelike androconia, these nearly as long as valva itself; costa narrow, its dorsal margin either gently convex, or rarely downcurved (P. rectangularis), usually spinose at apex; region above BO and below apex membranous, with fine, transverse striations; transtillar arms narrow, oriented sharply downward, or wide and horizontal (P. dilatata), meeting to form a concave, transverse sclerite at midline; aedeagus highly variable is size and shape, rounded at base, apex with a variously developed ventral process, this long, robust and strongly curved in P. dilatata; vesica variable in shape, either shorter than aedeagus (P. dilatata) or much longer than aedeagus (P. rectangularis); vesica covered with short, thornlike and long, spinelike cornuti.

Female. FW length  =  22.0–28.0 mm. Body characters as in male, except androconia absent. Additional differences include: Antenna ciliate; FW broader and longer, less acutely angled at apex; HW outer margin rounded, tornus simple; abdomen wider and shorter, truncate distally.

Terminalia (figs. 307E, 308B, 310B; P. rectangularis not examined): Tg7 equal in length to St7 or slightly longer; Tg7 widest at anterior margin, gradually narrower distally, anterior and posterior margins simple; St7 tapered toward distal margin, or much wider distally with lateral margins heavily sclerotized (P. dilatata); anterior margin simple, posterior margin gently or strongly concave; Tg8 almost completely membranous near midline, lateral portions either thin and straplike, or wide and robust; AA short and thin, or robust; A8 pleuron membranous; PA relatively small, lightly sclerotized, setae on dorsal margin longer than others; PP thin and straight, moderately long, longer than AA; PVP either mostly membranous, with numerous longitudinal striae, or large and densely sclerotized (P. dilatata); DB short and wide, membranous, dorsoventrally compressed; CB large, an elongate oval, with numerous folds in membrane; base of CB bearing a large appendix, its surface densely melanized, with numerous folds; dorsum of CB bearing a long, undulating, straplike sclerite along midline; DS attached dorsally at base of CB; signum variable in shape and location, either a large, transverse sclerite located ventrally near base of CB (P. poliana, P. rejecta), or a round sclerite in CB dorsum (P. dilatata).

Distribution

The described species of Phavaraea are endemic to lowland forests of South America. They occur across the Guyana Shield, as well as in the Upper Amazon Basin from Brazil and Peru north to Colombia. However, the genus is also now known to exist in Central America; a single USNM specimen, of an undescribed species, was collected in Panama near El Llano (330 m), on the Caribbean side of the Cordillera de San Blas. The southernmost record for Phavaraea is Rio de Janeiro, a locale where P. dilatata has been commonly collected. Material from Rio identified in collections as P. rejecta, instead appears to represent an undescribed species.

Biology

The life history of Phavaraea has not been discovered. The search for their host plants should begin with Passiflora species in the subgenus Astrophea (table 6), known hosts for Polyptychia and Getta, members of the larger josiine clade to which Phavaraea belongs (Clade 19; fig. 7).

Discussion

Even though Phavaraea, as defined here, contains only four species, considerable controversy is involved. Two of these taxa—rejecta and poliana—have resided comfortably in the genus since Prout (1918). Phavaraea dilatata, on the other hand, has undergone a tortured taxonomic history, seemingly because Lepidopterists have long attempted to showcase its strange appearance. When Walker (1854) originally described dilatata, he highlighted its uniqueness by placing it in its own subgroup, “Group 15. Scedros”, within what at that time was the genus Josia. As an aside, Walker's concept of Josia included 45 species assigned to 16 subgroups, but only some of those taxa were actually Josiini. Many of his so-called Josia species have since been moved to different dioptine genera, such as Erbessa, while still others have been assigned to the Arctiidae and Geometridae.

Nevertheless, Walker characterized dilatata by its extremely broad HW anal margin, as well as by having an abdomen “with a tuft of black hairs on each side towards its tip”. Prout (1918) then placed dilatata in Sagaris, now Phavaraea (see Watson et al., 1980). Hering (1925) followed suit. Finally, Bryk (1930) erected the genus Apocinesia, a junior objective synonym of Scedros (Watson et al., 1980), with dilatata as its sole included species. Following this convoluted path, I here place Scedros as a synonym of Phavaraea, once again uniting dilatata with rejecta and poliana, as Prout (1918) had suggested so long ago.

Contributing to the controversy, I here add a fourth Phavaraea species—rectangularis. This exotic taxon, described in the Arctiidae (Toulgoët and Navatte, 1997), instead belongs in the Josiini; here, it is referred to Phavaraea as a new combination. While acknowledging that P. rectangularis belongs somewhere in the Phavaraea + Polyptychia + Getta clade (figs. 7, 283), its precise position is in doubt. After considering the available options, I have placed it in Phavaraea, but this hypothesis is far from satisfying. Phavaraea now contains four species that exhibit an unsettling morphological diversity. Nowhere is this more apparent than in their androconial systems. While all Phavaraea males possess an androconial pouch on the HW anal margin, P. rejecta and P. poliana also show short, iridescent androconia on the abdominal venter of A2–A6 (pls. 27, 28). Phavaraea dilatata males lack this ventral system, but show dorsal abdominal androconia on A4–A7, as well as huge lateral bristles on the pleuron of A4–A6 (pl. 28). Male and female genitalia within Phavaraea reflect comparable structural diversity.

The classification proposed here should thus be regarded as provisional. It undoubtedly offers advantages over having a species, such as rectangularis, assigned to the wrong lepidopteran family. Nevertheless, much more remains to be learned regarding the members of this clade. Ultimately, the most stable solution for classifying these Josiini might be to recognize only two genera—one for those with androconia on the anterior margins of the FW and HW (Getta; 6 spp.), and one for those with a pouch on the HW anal margin (Polyptychia + Phavaraea; 6 spp.). Comprehensive species-level analysis would then be needed to resolve the relative positions of these taxa.

Type Locality

Brazil.

Type

Syntype ♂, Bates Collection (BMNH).

Discussion

Phavaraea dilatata can be distinguished from other members of the genus because the FW is more elongate (pl. 28), and the FW veins are lined with light brown scales as they run through the dark brown ground color; in other Phavaraea the FW veins are concolorous. Males exhibit a complex, and unique, set of androconia providing for easy recognition. First, the HW is extremely full with a greatly expanded anal margin (fig. 306B), held folded around the abdomen (Walker, 1854). Within this fold, the HW margin bears long, pedicellate, light gray androconia interspersed with even longer, brown lanceolate ones. On the abdomen itself there are two types of androconia: (1) long, blackish-brown bristles arising from the pleuron of A4–A6; and (2) elongate blackish-brown scales on the dorsum of A5–A7. The blackish bristles alone make the moth instantly recognizable. How the various parts of this male system function together during courtship is unknown.

Phavaraea dilatata is relatively rare. My survey of the world's collections uncovered a total of 24 specimens (19♂♂, 5♀♀), most of these captured well over 100 years ago. The largest series, at the ZMH, contains six males and two females, originally from the Staudinger Collection, all in beautiful condition (Brazil, Novo Friburgo). Three specimens document the species' occurrence in Suriname (1♂) and French Guiana (1♂, 1♀), while the remainder were collected in eastern Brazil, in the vicinity of Rio de Janeiro. The genitalia of moths from these rather disjunct localities are identical.

Distribution

Brazil (BMNH, CAS, MCZ, MNHN, OUMNH, VOB, ZMH); French Guiana (CAS); Suriname (BMNH).

Dissected

♂, Suriname, BMNH (genitalia slide no. JSM-346); ♂, Brazil, Espírito Santo, R.H. Stretch Collection, CAS (genitalia slide no. JSM-1743, wing slide no. JSM-1667); ♀, Brazil, Rio de Janeiro, San Fidelis, Sto. Antonio dos Brotos, 1876–1882, leg. August Vincent de Lyon, BMNH (genitalia slide no. JSM-347).

Phavaraea poliana (Druce)

Figure 308; plate 27

Eucyane poliana Druce, 1893: 285.

Type Locality

Guyana, Essequibo River.

Type

Syntype ♀, leg. Whitely (BMNH).

Discussion

Phavaraea poliana (pl. 27) shows a wing pattern similar to that of P. rejecta (pl. 28), except that the transverse FW band is white rather than being yellow-orange, and the band in P. poliana tends to be wider. Other than wing color, the easiest way to separate P. poliana and P. rejecta is by the color of their labial palpi (see Phavaraea species key). Males of P. poliana exhibit a stunning patch of brilliant, iridescent purple androconia on the venter of A2–6 (pl. 27), with matt-black androconia on A7 and A8. In P. rejecta, the androconia on A2–A6 are dull blackish purple (pl. 28), and those on A7 and A8 are gray black.

Distinguishing features of the genitalia include the following: In males, the anterior apodeme on St8 is truncate in P. poliana (fig. 308E), but bifid in P. rejecta (fig. 310E). Females show stronger differences. The signum in P. rejecta is a narrow transverse sclerite (fig. 310B), raised from the outer surface of the CB. In contrast, the signum of P. poliana is a large, broad plate (fig. 308B).

Like Phavaraea rejecta, P. poliana is quite rare—I know it from 15 specimens. The two species are sympatric, but P. poliana shows a more restricted distribution, occurring in the Guyana Shield as well as the lower portion of the Amazon River in Brazil.

Distribution

Brazil (AMNH, CMNH, CUIC, VOB, ZMH); Guyana (BMNH); Suriname (BMNH); French Guiana (BHC, BMNH).

Dissected

♂, French Guiana, BMNH (genitalia slide no. JSM-303); ♀, Suriname (“interio”), Sep 1892, leg. C.W. Ellacombe, BMNH (genitalia slide no. JSM-304).

Phavaraea rectangularis (Toulgoët and Navatte), new combination

Figure 309; plate 28

Pseudophaloë rectangularis Toulgoët and Navatte, 1997: 316–317.

Type Locality

French Guiana, Relais de Patawa, Piste de Kaw, PK 36, 19.

Type

Holotype ♂, leg. J. Cerda, Apr 1995 (MNHN).

Discussion

This species is here transferred from Pseudophaloë (Arctiidae: Pericopinae) to Phavaraea as a new combination. Phavaraea rectangularis, endemic to French Guiana, is highly unusual and exceedingly rare. The moth was not discovered until the mid-1990s. As far as I am aware, the world's holdings comprise three specimens—the male holotype and a female paratype, both at the MNHN in Paris, as well as a male in the Bernard Hermier Collection. Of these, I was able to examine the BHC male firsthand, using that specimen for photographs (pl. 28) and dissection (JSM-1775). After considerable study, I have assigned rectangularis to Phavaraea, but the species remains problematic.

Toulgoët and Navatte (1997) described rectangularis in the arctiid subfamily Pericopinae, but its membership in the Josiini is confirmed by a kettledrum thoracic tympanum and typical josiine male genitalia (fig. 309). Furthermore, the species belongs in the josiine subclade that includes Getta, Polyptychia and Phavaraea (Clade 19; fig. 7). Beyond that, its position is open to debate.

The only androconia in P. rectangularis are found in an elongate pouch, located along the anal margin of the HW (pl. 28). A similar HW system appears in Polyptychia (fig. 298E) as well as in other Phavaraea (fig. 306A, B), but the androconial organ of P. rectangularis—comprised of a relatively narrow fold enclosing a set of long, hairlike scales—is less elaborate than in these others. Furthermore, there are no abdominal androconia in P. rectangularis, either on the venter or pleuron.

Phavaraea rectangularis also exhibits a set of autapomorphies. For example, the abdominal dorsum is unusual in showing broad lateral patches of extremely short, appressed, light gray scales on A2–A4, and the dorsum of Tg8 exhibits a purplish cast. The antennae of P. rectangularis are unlike those of other Josiini, in that the rami along the mesal surface are much shorter than the lateral ones; elsewhere in the tribe, the pectinations on each side of the antennal shaft are roughly equal in length. The wing venation of P. rectangularis is not particularly helpful regarding generic placement; the species does not exhibit apomorphies characteristic of Getta (FW and HW discal cells short) or Polyptychia (FW Rs₄ branching below Rs₁). In the FW, vein M₁ arises from the anterolateral angle of the DC, touching the base of the radial sector, and M₃ is stalked with CuA₁. The HW venation is unremarkable, with Rs and M₁, as well as M₃ and CuA₁, being stalked. In lieu of additional character information, I have chosen Phavaraea as the most appropriate genus in which to place this species.

Distribution

French Guiana (BHC, MNHN).

Dissected

♂, /464/, Route forestière de Bélizon, Montagne Tortue pk 14.5, spk 2.5+(à droite)2, 1+(à droite)4.4, 20 May 2001, J.-L. Giuglaris, piège lumineux, BHC (JSM-1775).

Phavaraea rejecta (Geyer)

Figures 305A–D, 306A, 310; plate 28 [EX]

Centronia rejecta Geyer, 1832 (in 254Hübner, 1832) 4: 18, [pl. 115], figs. 663, 664.

Type Locality

“Java” (“patria falsa”; Bryk, 1930).

Type

Not seen.

Josia stygne Walker, 1854: 313.

Type Locality

Brazil, Pará.

Type

Syntype ♀, Bates Collection (BMNH).

Eucyane ortropea Druce, 1893: 286.

Type Locality

Interior of Colombia.

Type

Syntype ♀, “Wheeler, Mus. D”. (BMNH).

Discussion

From a taxonomic standpoint, the species described in the Zuträge zur Sammlung exotischer Schmetterlinge by Hübner and Geyer (1806–1838), have provided nothing but headaches. This taxon is no exception. As with all names published in those works, we are left with only a color plate showing the moth's dorsal and ventral views. The types are presumed lost, and no text accompanies the illustrations. Phavaraea rejecta is a particularly frustrating case because the type locality given, “Java”, is obviously in error. Subsequent authors (e.g., Bryk, 1930) have extrapolated its provenance as Brazil.

After attempting to establish the identity of P. rejecta, it has become apparent that two species co-occur in Brazil. One of them is large (♂ FW length  =  24.0–28.0) and broadly distributed across the Upper Amazon Basin from Brazil, Bolivia, and Peru north into Venezuela and French Guiana. The smaller one (♂ FW length  =  21.0–24.0) is confined to eastern Brazil, from Bahia south to Rio de Janeiro. The genitalia of the two are distinct, showing differences in the shape of male St8, and in the configuration of the vesica. The question is: Which one of these is rejecta?

Careful comparison of available museum material with the color plates in Hübner and Geyer leads me to believe that the smaller one from Rio, showing a wider and more orange-colored FW band, is P. rejecta. This moth is rare, known from six males and two females (MPM, OUMNH, USNM). The larger, possibly nameless, taxon is common; there are over 50 specimens in North American collections alone, as well as an even larger number at the BMNH. Unfortunately, all of this material is currently referred to a single name—rejecta. If specimens from southeastern Brazil are truly P. rejecta, then the name stygne Walker (type locality Pará), currently a synonym of rejecta, could be applied to the Amazonian taxon. Resolution will require dissection of Walker's type.

I defer to all previous authors (Prout, 1918; Hering, 1925; Bryk, 1930) in regarding ortropea Druce (1893), where the transverse FW band is deeply indented along its outer margin, as being a synonym of rejecta. However, otropea could potentially represent yet a third valid species within this complex.

Finally, a USNM female from Panama, collected by Gordon Small, represents the lone Central American record for Phavaraea. Small (1934–1989) is one of the most important modern collectors of neotropical Lepidoptera (see DeVries, 1997). His material, most of which resides at the USNM, has been hugely important throughout this paper. Small's specimen shows the same general size and wing pattern as P. rejecta. Its genitalia typify those of Phavaraea, but also show important autapomorphies. I have not described this taxon as new, choosing instead to wait until additional material, including males, is discovered. In the meantime, label data is as follows: Panama, Panama Prov., Distrito de El Llano, Cordillera de San Blas, North of El Llano, 330 m, 11 Jun 1978, leg. Gordon B. Small (genitalia slide no. JSM-1585).

Distribution

Brazil (BMNH, CMNH, CUIC, MNHN, MPM, OUMNH, PMNH, USNM, ZMH); Bolivia (ZMH); Peru (AMNH, BMNH, CUIC); Colombia (AMNH, BMNH, CMNH, MNHN, ZMH); Venezuela (AMNH, BMNH, CAS, CMNH, OUMNH, USNM); French Guiana (AMNH, BMNH, CAS, MNHN).

Dissected

♂, Peru, Río Marañon, 1918, BMNH (genitalia slide no. JSM-1527); ♂, [no data], USNM (genitalia slide no. JSM-1353, wing slide no. JSM-1354); ♀, Venezuela, Las Quiguas, Esteban Valley, Nov-Mar 1910, USNM (genitalia slide no. JSM-1355, wing slide no. JSM-1356); ♀, Brazil, Humayta, Rio Madeira, Jul–Sep 1906, leg. W. Hoffmanns, BMNH (genitalia slide no. JSM-1586).

EPHIALTIAS HÜBNER, 1816

Figures 311–Figure 312Figure 313Figure 314Figure 315Figure 316Figure 317Figure 318319; plates 28, 29, 39G

Figure 311

Morphology of Ephialtias (♂♂). A, head of E. dorsispilota, lateral view; B, head of E. dorsispilota, frontal view; C, head of E. dorsispilota, posterior view; D, head of E. monilis, lateral view; E, E. pseudena labial palpus; F, E. abrupta labial palpus; G, head of E. draconis, lateral view; H, E. draconis wings (illustration by J.S. Miller, except G by S. Goodman).

Figure 312

Scanning electron micrographs of Ephialtias draconis ♂. A, head in frontal view; B, proboscis in lateral view; C, sensilla styloconica near apex of proboscis; D, distal portion of antenna; E, junction of a single ramus with antennal shaft (ventral view); F, close-up of ramus, showing smooth area at ramus base.

Figure 313

Scanning electron micrographs of Ephialtias draconis ♂. A, large sensillum styloconicum of antenna (lateral view), located near apex of flagellomere; B, metathoracic tympanal opening, lateral view; C, close-up of tympanal opening; D, left epiphysis (mesal view); E, close-up of epiphysis surface; F, tibial spur, showing serrate apex.

Figure 314

Genitalia of Ephialtias abrupta (♂ JSM-1731, ♀ JSM-1732). A, ♂ genitalia; B, aedeagus; C, ♂ Tg8; D, ♀ genitalia; E, ♂ St8.

Figure 315

Genitalia of Ephialtias choba (♂ JSM-1343, ♀ JSM-1344). A, ♂ genitalia; B, aedeagus; C, ♂ Tg8; D, ♀ genitalia; E, ♂ St8.

Figure 316

Genitalia of Ephialtias dorsispilota (♂ JSM-125, ♀ JSM-126). A, ♂ genitalia; B, aedeagus; C, ♂ Tg8; D, ♂ St8; E, ♀ genitalia (illustration by A. Trabka).

Figure 317

Genitalia of Ephialtias monilis (♂ JSM-1347, ♀ JSM-1348). A, ♂ genitalia; B, ♂ Tg8; C, aedeagus; D, ♂ St8; E, ♀ genitalia (illustration by A. Trabka).

Figure 318

Genitalia of Ephialtias. A, ♂ of E. pseudena (JSM-301); B, ♀ of E. pseudena (JSM-860); C, aedeagus of E. pseudena; D, E. pseudena ♂ Tg8; E, E. pseudena ♂ St8; F, ♂ of E. tenuifascia (JSM-1541); G, ♀ of E. tenuifascia (JSM-1729); H, aedeagus of E. tenuifascia; I, E. tenuifascia ♂ Tg8; J, E. tenuifascia ♂ St8.

Figure 319

Genitalia of Ephialtias velutinum (♂ JSM-296, ♀ JSM-297). A, ♂ genitalia; B, ♂ Tg8; C, aedeagus; D, ♂ St8; E, ♀ genitalia (illustration by A. Trabka).

Ephialtias Hübner, 1819: 176. Type species: Hypocrita abrupta Hübner, 1806, vol. 1: pl. 184 (by subsequent designation by Grote, 1866).

Actea Walker, 1854: 307. Type species: Hipocrita monilis Hübner, 1811 (by monotypy); a junior homonym of Actea Germar (Insecta, Hemiptera).

Mitradaemon Butler, 1878: 60. Type species: Mitradaemon velutinum Butler, 1878 (by original designation). New synonymy.

Retila Boisduval, 1870: 94. Type species: Hypocrita abrupta Hübner, 1806 (by subsequent designation by Prout, 1918: 420); a junior objective synonym of Ephialtias Hübner, 1819.

Diagnosis

All Ephialtias species exhibit a dark brown to blackish-brown FW with some form of yellow-orange transverse band approximately halfway out, but HW coloration varies dramatically (pls. 28, 29). Members of the Bryce Group are unmistakable: the only Dioptinae with a rosy red HW band. The best way to recognize the majority of Ephialtias taxa is by the series of small, yellow or white abdominal spots along the dorsal midline on A2–A7. However, this trait is not universally reliable; both sexes of E. velutinum (pl. 28) and males of E. abrupta (pl. 29) lack spots, the dorsum completely black. These latter taxa could potentially be confused with certain Getta species (pl. 26). However, in Getta FW vein M₁ is fused with Rs₁–Rs₄ (fig. 289F, H), whereas in Ephialtias M₁ arises from the DC (fig. 311H). Males of the two genera can also be separated because the wings in most Getta bear androconial patches (fig. 289H), whereas these are absent in Ephialtias.

Male and female genitalia offer reliable means for recognizing Ephialtias. Males have a horn-shaped aedeagus base (e.g., fig. 314B). Females in the Bryce Group show an unusual, elongate sclerite—evenly spinose throughout its length—running from the DB to most of the length of the CB (fig. 318G). In the Abrupta Group, females exhibit a unique downward bend in the distal portion of the CB (fig. 318B).

Redescription

Male. FW length  =  14.5–19.0 mm. Head (fig. 311A–G, 312): Labial palpus moderately long, porrect, held away from face; Lp1 wide and gently curved, with a loose ridge of slightly longer scales below; Lp2 wide, variable in length, equal in length to Lp1 (e.g., E. dorsispilota), shorter than Lp1 (e.g., E. monilis), or longer (e.g., E. abrupta); Lp2 usually tightly scaled with slightly longer scales below, apex sometimes with a diffuse, ventral tuft of long, bristlelike scales (E. monilis, E. dorsispilota); Lp3 short, bullet shaped, in E. dorsispilota slightly elongate, Lp2 closely scaled, sometimes with an acute apical tuft of elongate scales (e.g., E. monilis); scales of front variable, usually pointing dorsomedially with dorsal scales forming a loose, triangular tuft between antennal bases, scales sometimes swooping down from below antennal bases and meeting at midline above clypeus (fig. 312A), in E. abrupta all scales converging from above and below toward center of front; eye large and bulging, gena scaleless, narrow, slightly wider anteriorly; postgena narrow; scales of vertex relatively short, tightly packed, pointing anteriorly; antenna bipectinate, rami sometimes long (e.g., E. dorsispilota, E. monilis), terminal 15–20 annulations simple; antennal shaft widened from base, then gradually tapering toward apex.

Thorax (fig. 313): Epiphysis short, gradually tapered, not extending beyond apex of tibia; tegula long and wide, over two-thirds as long as mesoscutum, distal portion sometimes slightly narrowed (e.g., E. abrupta), ventral portion acute at apex, the two areas divided by a strong transverse sulcus; metathoracic tympanum large, kettledrum shaped; tympanal membrane large and round, oriented horizontally.

Forewing (fig. 311H; pls. 28, 29): Elongate, sometimes broadly triangular with an acute apical angle (e.g., E. velutinum); vein Rs₁ arising from radial sector; Rs₂–Rs₄ in the pattern [2+3]+4; M₁ arising from anterolateral angle of DC near, but separate from, base of radial sector, UDC short, or sometimes (e.g., E. dorsispilota) fairly long and obliquely angled; DC much longer than one-half FW length; veins M₃ and CuA₁ long stalked; ground color dark brown to blackish brown; an oblique, orange-yellow transverse band crossing from behind costa near middle of wing to near tornus, either straddling distal margin of DC, or crossing immediately beyond DC, band varying in width and length; sometimes with a spot, or pair of spots, at position of transverse band (E. choba, E. monilis); ventral surface always showing a transverse band.

Hind wing (fig. 311H; pls. 28, 29): Usually broadly rounded, sometimes elongate (e.g., E. abrupta); M₃ and CuA₁ long stalked; central area highly variable, either uniformly blackish brown (e.g., E. velutinum), white (E. abrupta), with an orange-yellow transverse band (e.g., E. monilis), or sometimes rosy red (Bryce Group).

Abdomen: Wide at base, abruptly tapered toward apex; dorsum dark brown to blackish brown; A2–A7 usually with small white or yellow spots along midline (pls. 28, 29); almost always bearing white or yellow longitudinal pleural stripes, these faint in E. abrupta, absent in E. velutinum; venter usually white, occasionally dark brown (E. dorsispilota, E. velutinum).

Terminalia (figs. 314A–C, E; 315A–C, E; 316A–D; 317A–D; 318A, C–E; 318F, H–J; 319A–D): Tg8 short, widest anteriorly, tapered distally, anterior margin with a pair of extremely short, rounded apodemes on each side of a shallow mesal excavation, posterior margin lightly sclerotized, with a poorly defined, U-shaped or horseshoe-shaped mesal excavation; St8 much longer and wider than Tg8, widest anteriorly, gradually tapered distally, anterior margin with a moderately long, sometimes short, mesal apodeme, apodeme gradually tapered and blunt, or abruptly tapered and narrow, posterior margin with a small, poorly defined, U-shaped mesal excavation; socii/uncus complex relatively small, narrowly attached to tegumen; uncus narrow, curving gently downward, apex acute, occasionally truncate (Bryce Group); socii narrow, held close to uncus, apices acute, socii slightly shorter than uncus, with same curvature; tegumen slightly shorter than vinculum, extremely narrow at connection with valva below, becoming wider toward uncus/socii complex above; vinculum narrow; saccus absent, lower margin of genitalia transverse; valvae large, sometimes extremely so (e.g., E. velutinum), membranous, upright; BO extremely large, occupying almost entire valva, ventrolateral margin expanded, membrane fragile; inner surfaces of BO covered with fine, hairlike androconia, lateral surfaces bearing extremely long hairlike androconia in lower third, as well as short, petiolate androconia in upper two-thirds; costa of valva extremely short, narrow, simple, rarely with a small bump on dorsal margin (E. velutinum); apex of valva membranous, rounded, sometimes rugulose (e.g., E. monilis); arms of transtilla with a small curve where they arise from costa, oriented sharply downward, meeting at midline to form a wide, slightly concave sclerite, its upper margin broadly U-shaped; aedeagus wide, relatively short, knob shaped at base, dorsal opening of DES located posteriorly; apex of aedeagus gradually acute, forming a ventral point; vesica either broad, approximately as long as aedeagus, forming an upright sac (Abrupta Group), or long and relatively narrow, approximately twice as long as aedeagus, extending posteriorly (Bryce Group); vesica with short, thornlike or spinulose cornuti, as well as long, robust, spinelike cornuti.

Female. FW length  =  16.0–20.0 mm. Head and thorax similar to male, except: Lp2 without an apical tuft where one occurs (E. monilis, E. pseudena), instead with a short brush of distal scales; antenna ciliate; wings longer and broader, outer margins more rounded; frenulum comprising two bristles.

Abdomen: Shorter and much wider than male; dorsum with a row of ovoid, white or yellow-orange spots along midline on A2–A7, spots sometimes faint (e.g., E. consueta), rarely absent (E. velutinum).

Terminalia (314D, 315D, 316E, 317E, 318B, 318G, 319E): Tg7 long and wide, slightly narrower posteriorly, much longer than Tg6, anterior margin simple, posterior margin simple, rarely with a V-shaped mesal excavation (E. pseudena); St7 approximately equal in length to Tg7, somewhat narrower; St7 slightly narrower posteriorly, anterior and posterior margins simple; Tg8 absent, that area completely membranous; AA short and thin; A8 pleuron membranous; PP long, thin and straight; PA small, mostly membranous, posterior margin with a small dorsal lobe; PVP convex, extremely wide, lateral margins tapered, wrapping upward to touch AA, posterior margin of PVP simple, rarely with a wide, rounded mesal process (E. abrupta); DB short and wide, dorsoventrally compressed, membranous or occasionally lightly sclerotized (Bryce Group); CB large and wide, an elongate oval, anterior fourth often with a ventral fold (Abrupta Group), this portion bent downward; CB with a large, transverse, dorsal sclerite near base, lateral margins of sclerite often with short internal spines; Bryce Group with a broad, finely spiculate dorsal area near base of CB; signum in Abrupta Group elaborate, “bird-shaped”, located ventrally, comprising a triangular sclerite, each posterolateral angle bearing a group of greatly elongate internal spines, posterior margin of triangular sclerite with a large, knob-shaped process protruding from CB membrane; signum in Bryce Group apparently absent, ventral surface of CB instead with an extremely long, narrow, concave sclerite, its internal surface bearing long, robust internal spines along margins; DS arising from CB in basal third along dorsal midline; Abrupta Group with a thin, sclerotized strip at base of DS.

Distribution

Ephialtias is generally a genus of lowland forests. The Canal Zone in Panama is home to three species—draconis, dorsispilota, and pseudena (Forbes, 1939a)—but the majority occurs along the Amazon River and into the Upper Amazon Basin. Almost all Ephialtias have been collected at 400 meters or below, with only two taxa, E. pseudena (0–800 m) and E. dorsispilota (0–1200 m), occurring at midelevations. This lowland distribution is perhaps related to the habitat of its host plants, Lindackeria and Turnera (see below). Ephialtias species have been recorded as far north as Orosí, Costa Rica (E. dorsispilota), and as far south as Las Juntas, Bolivia (E. draconis).

Biology

The life histories of two Ephialtias species, both from Panama, have been discovered (table 6). Caterpillars of E. draconis feed on Turnera panamensis (Turneraceae). This is the second member of the Josiini known to be associated with Turneraceae; Josia megaera is associated with Turnera odorata (table 6). The growth forms of these two plants differ in that T. panamensis is a tree, whereas T. odorata is a weedy bush (pl. 41D). The other known host of Ephialtias is attributed to the work of Andres Orellana (Tachira, Venezuela). During a year of fieldwork in Panama, he documented an association between E. dorsispilota and Lindackeria laurina (Achariaceae). This is the first report of a dioptine on Achariaceae (see Discussion: Host Plants).

These findings set Ephialtias apart from other Josiini, the vast majority of which are restricted to Passifloraceae. It will be fascinating to learn the hosts of more Ephialtias species, since it appears that these may be pivotal in helping to unravel the evolution of host use for the Josiini.

Discussion

In previous synthetic works on the Dioptinae (Prout, 1918; Hering, 1925; Bryk, 1930), the species now in Ephialtias appeared in one of either two genera—Actea or Josia. Watson et al. (1980) noted that Actea is a homonym of a hemipteran genus (Germar, 1842). They further observed that its type species, monilis, is congeneric with velutinum Butler, the type species of Mitradaemon. The latter would thus be available as a replacement name for Actea. However, my research shows that abrupta Hübner, the type species of Ephialtias, and velutinum are congeneric. Their genitalia exhibit numerous synapomorphies (appendix 4), including presence of: a knob-shaped base of the aedeagus; a fan-shaped tuft of androconia at the base of the valvae; a ring-shaped sclerite at the base of the corpus bursae; and a large, bird-shaped signum with a triangular sclerite at its base. The oldest name is that of Hübner (1816), rather than Butler's Mitradaemon (1878), so the latter becomes a synonym of Ephialtias.

In summary, Actea of previous authors becomes synonymous with Ephialtias. To the five species formerly in Actea, I have added abrupta and consueta, both most recently in Josia (Bryk, 1930). Together these form the Abrupta Group. In addition, according to my analyses, draconis, also formerly in Josia, belongs at the base of the clade that includes abrupta, monilis, and dorsispilota (fig. 283). I have therefore included draconis and its relatives, as the Bryce Group (3 spp.), in a broadly conceived Ephialtias, bringing the number of included species to 10 (appendix 2).

As noted in the Diagnosis (above), Ephialtias shows considerable wing-pattern diversity. This genus provides a case strikingly similar to Proutiella, in which individual taxa seem to have evolved wings unlike those of their close relatives. For example, Ephialtias abrupta uniquely appears with white in the HW. Furthermore, the two Ephialtias with an orange-yellow HW central area, E. monilis and E. choba, are not sister species as one might assume, but instead show sister relationships with different Abrupta Group taxa. Such variation has confounded previous students of the Dioptinae. When morphological details are analyzed, the monophyly of Ephialtias is strongly supported (BS  =  12; fig. 2). Furthermore, each of the two subclades—the Abrupta and Bryce groups—shows evidence for monophyly.

The identities of the various Ephialtias species seem to have perplexed earlier authors. In certain cases, obviously distinctive taxa, such as E. dorsispilota, had been placed in synonymy but are here rightfully elevated to species status. Elsewhere in the genus, species status was recognized for subtle wing-pattern variants of a single species, and these are placed in synonymy; e.g., the numerous names now associated with E. abrupta. This confusion is somewhat surprising because most the characters needed to properly identify these taxa involve coloration traits easily observed on the head, wings, or abdomen. It is hoped that my taxonomic arrangement will provide a level of stability.

Type Locality

Not known.

Type

Not seen.

Ephialtias basalis Butler, 1878: 59.

Type Locality

Brazil, Pará.

Type

Syntype ♀, leg. 9 Mar 1875 (BMNH).

Josia dorsivitta Walker, 1854: 305. New synonymy.

Type Locality

Brazil, Ega.

Type

Syntype ♀, Bates Collection (BMNH).

Josia hyperia Walker, 1854: 306. New combination, revised synonymy.

Type Locality

Brazil, Pará.

Type

Syntype ♀, Bates Collection (BMNH).

Josia icca Prout, 1918: 421.

Type Locality

Brazil, Pará.

Type

Syntype ♀ (BMNH).

Josia pilarge Walker, 1854: 305. New combination, revised synonymy.

Type Locality

Brazil, Pará.

Type

Syntype ♀, Bates Collection (BMNH).

Discussion

Ephialtias abrupta has been a difficult taxon to characterize. After considerable study, a broad species concept has been applied; the name abrupta is used for all Ephialtias showing a white region on the HW dorsal surface (pl. 29). The white area varies greatly in size. In some specimens it occupies over half the wing, whereas in others it is a small, inconspicuous patch. The white HW area in the male and female moths figured by Hübner (1806; pl. 184) is of moderate size, but is considerably larger in the female.

Because of this pattern variation, previous authors recognized more than one species and described numerous forms and subspecies. In my arrangement, there are five synonyms of abrupta. Having made this move, it is interesting to note that the type locality for four of these five is the same—Pará, Brazil. The remaining locale, Ega, type locality for dorsivitta Walker, is nearby on the Amazon river. These observations add credence to the hypothesis that the names constitute a single species with highly variable wing pattern. Unfortunately, the Hübner type of abrupta has been lost, and its type locality will never be known.

Details of the new synonymies are as follows: First, my research suggests that dorsivitta Walker, previously regarded as a species in Josia (Bryk, 1930), is a synonym of abrupta. The two differ only in the size of the white HW central area (smaller in abrupta). Their genitalia and other aspects of body coloring are identical. Secondly, study of BMNH types confirms that hyperia and pilarge, both described by Walker (1854), are synonyms of abrupta Hübner. Prout (1918: 421) and subsequent authors instead listed them as synonyms of E. consueta Walker. The types of hyperia and pilarge show white dorsal spots on the abdomen and a FW band typical of abrupta. However, hyperia has only a tiny amount of white in the HW, while pilarge exhibits a larger, but still faint, white central area (more clearly visible below). Finally, the names basalis Butler and icca Prout were treated as aberrations of abrupta by Prout (1918) and subsequent authors (Hering, 1925; Bryk, 1930). On those, I defer judgement.

No other Ephialtias species shows white in the HW. Other, similar-appearing Josiini, such as P. simplex and P. esoterica (pl. 26), belong in Proutiella. Certain Arctiidae, such as Ordishia klagesi (pl. 29), are mimetic with E. abrupta, but wing venation and tympanal morphology make these easily separable.

The abdomen of E. abrupta females exhibits a row of white spots along the midline. Interestingly, the abdominal dorsum in males is completely black. Abdominal spots are characteristic of Ephialtias, but only E. abrupta and E. consueta possess white ones.

Within the Abrupta Group, genital morphology suggests that E. abrupta (from the Amazon Basin) and E. pseudena (from Colombia and Panama) are sister species. In both, male St8 exhibits a large, wide anterior apodeme (figs. 314E, 318E), and the valva apex bears a tiny, thumblike, membranous appendix (figs. 314A, 318A). They can be distinguished because the FW band is consistently narrower in E. abrupta, and E. pseudena never shows white in the HW.

Distribution

Brazil (BMNH, CMNH, NMW, OUMNH, USNM); Peru (AMNH, CUIC, MUSM, USNM, ZMC); Ecuador (LACM); Venezuela (MNHN).

Dissected

♂, Brazil, ex Coll. Smith, 1844–5, BMNH (genitalia slide no. JSM-333); ♂, Brazil, Santarem, May 1919, leg. S.M. Klages, CMNH (genitalia slide no. JSM-1731); ♂, [no data], “comp. type B.M.”, Wm. Schaus Collection, USNM (genitalia slide no. JSM-1345, wing slide no. JSM-1346); ♀, Brazil, Pará, BMNH (genitalia slide no. JSM-334); ♀, Peru, Madre de Dios, Parque Manu, Pakitza, 11°53′S, 70°58′W, 400 m, 5 Oct 1990, leg. R. Robbins, USNM (genitalia slide no. JSM-1732).

Ephialtias choba (Druce), new combination

Figure 315; plate 29

Myonia choba 142Druce, 1899: 297.

Type Locality

Brazil, Amazonas, Santarem.

Type

Syntype ♂, leg. Leech (BMNH).

Actea transita Hering, 1925: 524. New synonymy.

Type Locality

Brazil, Pará (♂).

Type

Syntype ♂/♀ (ZMH).

Discussion

This species was formerly placed in the genus Actea Walker (Prout, 1918; Hering, 1925; Bryk, 1930). Here, it is newly combined with Ephialtias.

Typical specimens of Ephialtias choba differ from E. monilis (pl. 29) in possessing a single small yellow spot near the FW costal margin, rather than having a pair of large spots separated at M₂. Comparison of male and female genitalia (figs. 315, 317) confirms that the two species are distinct. They show strong differences in the shape of male St8, in the shape of the male vesica, and in the shape of the signum of the female CB. In fact, within the Abrupta Group, E. choba and E. monilis do not appear to be sister taxa. For example, they show different scale configurations on their labial palpi (see species key), and the frontal scales are oriented differently—vertically in monilis but horizontally in choba. Such traits indicate considerable divergence between the two. At this time, I am unable to suggest a precise placement of E. choba relative to other Ephialtias species.

Dissection of the ZMH syntypes of Actea transita, described as a species by Hering (1925) and retained as such by Bryk (1930), produced an interesting result. These specimens show a long yellow transverse FW band (pl. 29), rather than a single spot near the anterior margin. Their FW pattern thus approximates that of E. monilis. Nevertheless, their male and female genitalia (JSM-1764, 1765) are identical with those of E. choba, as are their palpus and head-scaling patterns. Here, transita Hering is placed in synonymy with E. choba (Druce). To stabilize the nomenclature of transita, I also designate the male as a lectotype and the female as a paratype. In addition to the ZMH type material, there is a male paratype of transita at the BMNH.

This synonymy demonstrates the existence of wing-pattern variation in E. choba, with most specimens showing a single small FW spot, but rare examples (the transita types) exhibiting a long, uneven band. The best wing-pattern trait for separating E. choba and E. monilis involves the FW ventral surface, where choba shows a diffuse basal streak, absent in E. monilis.

Ephialtias choba is not common in collections, but a fairly large series (6♂♂, 6♀♀) is housed at the USNM. The species is found on the lower Amazon of Brazil, in the vicinity of Santarém, occurring in sympatry with E. abrupta and E. consueta.

Distribution

Brazil (BMNH, CMNH, NMW, USNM, ZMH).

Dissected

♂, Brazil, Amazonas, Taperinha, Dognin Colln., USNM (genitalia slide no. JSM-1343); ♂ type of transita Hering, Brazil, Pará, A. Schulz S., 23/992, ZMH (genitalia slide no. JSM-1764); ♀, Brazil, Amazonas, Taperinha, Dognin Colln., USNM (genitalia slide no. JSM-1344); ♀ type of transita Hering, Brazil, Pará, “7738”, ZMH (genitalia slide no. JSM-1765).

Ephialtias consueta (Walker)

Plate 29

Josia consueta Walker, 1854: 304.

Type Locality

Brazil, Pará.

Type

Holotype ♀ (BMNH).

Josia lugens C. and R. Felder, 1874: pl. 105, fig. 22.

Type Locality

[Not known].

Type

Holotype ♀, “Bates”, ex Felder Collection (BMNH).

Discussion

Ephialtias consueta is known exclusively from two females—Walker's holotype (pl. 29) and the Felder holotype of lugens—both at the BMNH. Careful examintion reveals the presence in E. consueta of faint white spots along the abdominal dorsum, an important characteristic of Ephialtias. Furthermore, I dissected the female type of lugens C. and R. Felder. There is no locality data on that specimen, but a handwritten label gives Bates as the collector. It was therefore, in all probability, collected on the Amazon. The Felder type of lugens matches the type of consueta with precision; the two are undoubtedly synonyms, as Prout (1918) had proposed. My dissection (JSM-1545) confirms membership of consueta, formerly in Josia (Prout, 1918; Hering, 1925; Bryk, 1930), in Ephialtias, where Kirby (1892) had placed it long ago.

Ephialtias consueta is the only member of the genus in which the HW dorsal surface is completely dark, but the ventral surface shows a diffuse white region along the anal margin. Prout (1918) originally described cassa as a subspecies of consueta. It does not show this white ventral HW spot. I have instead moved cassa to its new position as a synonym of Ephialtias pseudena (appendix 2).

Distribution

Brazil (BMNH).

Dissected

♀ holotype of lugens, BMNH (genitalia slide no. JSM-1545).

Ephialtias dorsispilota Warren, revised status

Figures 311A–C, 316; plates 28, 39G [EX]

Ephialtias dorsispilota Warren, 1905: 313.

Type Locality

Colombia, Cananche, Cundinamarca.

Type

Syntype ♂, leg. de Mathan, Jul 1903 (BMNH).

Discussion

The identities of two taxa—E. dorsispilota Warren and E. pseudena (Boisduval)—are consistently confused in collections. They show similar wing patterns (pl. 28), and are usually treated as a single species, pseudena, with dorsispilota as its synonym (e.g., Bryk, 1930; Forbes, 1939a). They instead constitute two species that are easily separated. In E. dorsispilota, the abdomen is dark brown to blackish brown, with faint yellow spots along the dorsal midline and yellow lateral stripes on the pleuron. The venter is dark brown (pl. 28). Alternatively, the abdomen of E. pseudena is completely blackish brown dorsally and laterally, but the venter is immaculate white (pl. 28). Ephialtias pseudena is also whitish on the front and femora, is slightly smaller, and has narrower wings. Male and female genitalia in the two taxa differ significantly (figs. 316, 318A–E). These species appear to be roughly sympatric. The northernmost limit for Ephialtias is currently based on a female specimen of E. dorsispilota from Orosí, Costa Rica (BMNH), collected by Fassl.

Andres Orellana established the host plant of E. dorsispilota as Lindackeria laurina (Achariaceae). This is the first record of Achariaceae-feeding in the Dioptinae (table 6), but it could almost have been predicted. A broad association with Malphigiales mirrors the pattern known for butterflies in the Heliconiinae (see Discussion: Host Plants). The larvae of E. dorsispilota (pl. 39G) are unlike most Josiini; the yellowish brown head and reduced purple-black markings set them apart. They most closely resemble caterpillars of Josia megaera (pl. 39L).

Although described in Ephialtias (Warren), authors subsequent to Prout (1918) assigned this species to Actea.

Distribution

Costa Rica (BMNH); Panama (AMNH, BMNH, CAS, CMNH, FNHM, NMW, OUMNH, USNM); Colombia (AMNH, BMNH, USNM).

Dissected

♂, Colombia, Muzo, Nouv. Grenade, leg. Dr. O. Thieme, 10 Feb 1877, BMNH (genitalia slide no. JSM-857); ♂, Panama, Barro Colorado Island, AMNH (genitalia slide no. JSM-125); ♀, Panama, Empira, Canal Zone, AMNH (genitalia slide no. JSM-126, wing slide no. JSM-166); ♀, Costa Rica, Orosí, 1200 m, leg. Fassl, BMNH (genitalia slide no. JSM-858).

Ephialtias monilis (Hübner)

Figures 311D, 317; plate 29 [EX]

Hipocrita (Tineiformis) monilis Hübner, 1805: pl. 183, figs. 1–4.

Type Locality

Not known.

Type

Not seen.

Discussion

Ephialtias monilis is similar in wing pattern to typical examples of E. choba (pl. 29), differing in showing a pair of yellow spots, rather than a single spot, in the position of the FW cross band. These are formed by having the transverse band abruptly narrowed, sometimes broken, at the point beyond the DC where M₂ arises. However, as is discussed above, some rare wing pattern variants of E. choba (“transita”, pl. 29) show a broken transverse band nearly identical with that of E. monilis. More reliable wing pattern differences are the width of the yellow-orange HW band—wide in E. monilis, but narrow in E. choba—and, on the FW ventral surface, a basal streak in E. choba not found in E. monilis.

The labial palpi provide another method for separating males of E. monilis and E. choba: Lp2 and Lp3 in E. monilis bear long, bristlelike scales at their apices, forming a diffuse tuft. This tuft is absent in choba. Interestingly, a homologous palpus tuft occurs in E. dorsispilota, a species with an unbroken FW band and a uniformly blackish HW (pl. 28). Genital similarities further suggest that E. monilis and E. dorsispilota are sister species (figs. 316, 317).

Ephialtias monilis, widely distributed throughout the Upper Amazon Basin, also occurs in French Guiana. The species is common in museum collections, often being represented by large series. A single female, in the Naturhistorisches Museum, Vienna, bears labels indicating that it was captured in Chiriquí, Panama. That specimen, purchased by Staudinger, was apparently collected in August 1886, although the collector is not given. The veracity of this label seems doubtful (G. Lama, personal commun.), but were it to be correct, this is the first record of E. monilis for Central America.

During the course of its taxonomic history, this species has resided in Josia (Walker, 1854), Brachyglene (Kirby, 1892), and Actea (Prout, 1918; Hering, 1925; Bryk, 1930).

Distribution

Brazil (AMNH, BMNH, CMNH, LACM, NMW, OUMNH, USNM, ZMH); Peru (AMNH, MUSM, USNM, ZMH); Ecuador (BMNH); French Guiana (AMNH, BHC); Panama (NMW).

Dissected

♂, Peru, Madre de Dios, Parque Manu, Pakitza, 12°07′, 70°58′, 400 m, 21 Sept 1989, leg. R. Robbins, USNM (genitalia slide no. JSM-1543); ♂, [no data], colln. W. Schaus, USNM (genitalia slide no. JSM-1347); ♀, [no data], colln. W. Schaus, USNM (genitalia slide no. JSM-1348).

Ephialtias pseudena (Boisduval)

Figures 311E, 318A–E; plate 28

Retila pseudena Boisduval, 1870: 94.

Type Locality

“ Muzo”?.

Type

Syntype ♂ (BMNH).

Josia cassa Prout, 1918: 421. Revised synonymy.

Type Locality

Colombia, Muzo, Río Cantinero, 400 m.

Type

Holotype ♂, leg. A.H. Fassl (BMNH).

Discussion

In his description of pseudena, Boisduval (1870: 94) listed Honduras and Mexico as localities. Subsequent authors thus mistakenly reported this species from Mexico south to Colombia (e.g., Hering, 1925). The locality label on Boisduval's pseudena type is extremely difficult to read. After considerable study of the handwriting, I propose “Muzo” as the type locality. Support for this hypothesis comes from other museum material. No verified examples of the moth have been collected north of the Canal Zone in Panama, whereas many specimens are known from Muzo, near the headwaters of the Río Magdalena in central Colombia.

Ephialtias dorsispilota, previously treated as a synonym of pseudena (Bryk, 1930), is here raised to species status following the original author (Warren, 1905). The two moths are readily separable (pl. 28) because the front and femora, as well as the venter of the abdomen, are white in E. pseudena, whereas these regions are dark brown to blackish brown in E. dorsispilota. Of the two species, E. pseudena is less common in collections.

Prout (1918: 421) described cassa as a subspecies of Josia consueta Walker, differing in having no white dorsal markings on the abdomen, in having the FW cross band straighter and narrower, and in lacking white scaling along the HW anal margin. Subsequent authors have followed this treatment (Hering, 1925; Bryk, 1930). My study of the relevant types suggests that cassa is instead a junior synonym of pseudena. Prout stated that cassa is known from Panama and from Bogotá, Colombia. This matches the known distribution of E. pseudena, but not that of E. consueta (Pará, Brazil).

Based on genital similarities, E. pseudena is an extremely close relative of E. abrupta from the Amazon Basin. Synapomorphies are noted in the discussion of the latter (above).

Distribution

Colombia (AMNH, BMNH, MNHN); Panama (AMNH, BMNH, NMW).

Dissected

♂, Colombia, Muzo, 400–800 m, leg. Fassl, BMNH (genitalia slide no. JSM-301); ♂, Colombia, Cundinamarca, Nouvelle Grenade, Cananche, 1 Sept 1900, leg. M. de Mathan, BMNH (genitalia slide no. JSM-294); ♂, Colombia, Cundinamarca, Cananche, 1 Sep 1900, leg. M. de Mathan, BMNH (genitalia slide no. JSM-859); ♀, Colombia, Muzo, 400–800 m, leg. Fassl, BMNH (genitalia slide no. JSM-302); ♀, Colombia, Muzo, R. Cantinero, 400 m, leg. A.H. Fassl, BMNH (genitalia slide no. JSM-860).

Ephialtias velutinum (Butler), new combination

Figure 319; plate 28

Mitradaemon velutinum Butler, 1878: 60, pl. 3, fig. 1.

Type Locality

Brazil, Amazonas, Barreiras de Jutahi.

Type

Syntype ♀, 18 Jan 1875 (BMNH).

Discussion

This species was described in Mitradaemon (Butler, 1878), retained there by Kirby (1892), and transferred to Actea by Prout (1918), where it has resided ever since. It is here assigned to Ephialtias for the first time.

Ephialtias velutinum, E. dorsispilota, and E. pseudena (pl. 28) are similar in showing a yellow transverse band in the FW and a completely dark HW. These three species can be differentiated by the pigmentation pattern on the abdomen: E. pseudena is completely dark above but white below; E. dorsispilota has small yellow spots along the dorsal midline and thin, longitudinal yellow stripes along the pleuron; while the abdomen of E. velutinum is completely dark brownish black, with no markings whatsoever. Of the three, E. velutinum has the longest wingspan (♂ FW length  =  17.0–20.0 mm).

Ephialtias velutinum occurs along the upper portion of the Amazon River, from Manaus west to eastern Ecuador and Peru. The species is represented in numerous collections around the world, but rarely by more than a few specimens in each.

Distribution

Brazil (BMNH, CMNH, LACM, ZMH); Peru (AMNH, BMNH, LACM, MPM, MUSM, ZMH); Ecuador (BMNH).

Dissected

♂, Ecuador, Sarayacu, leg. C. Buckley, BMNH (genitalia slide no. JSM-296); ♀, Peru, Iquitos, Upper Amazon, Jul 1932, leg. G. Klug, BMNH (genitalia slide no. JSM-297).

Type Locality

Brazil, Tapajós.

Type

Holotype ♂, Bates Collection (BMNH).

Josia rosea Hering, 1925: 526, fig. 70h. New synonymy.

Type Locality

Brazil, Itaituba.

Type

Syntype ♂/♀ (ZMH).

Josia subdraconis Bryk, 1953: 226.

Type Locality

Brazil, Amazonas, S. Gabriel.

Type

Holotype ♀ (not seen), leg. 29 Dec, ex Roman Collection.

Discussion

According to my taxonomic interpretation, E. bryce can be distinguished from other Bryce Group species because the transverse FW band shows reddish coloring at it anterior and posterior ends in bryce, whereas the band is red exclusively at its anterior end in the other two taxa. Ephialtias bryce is also slightly smaller than E. draconis and E. tenuifascia. The genitalia of E. bryce are also distinctive. For example, in females, the elongate sclerite of the corpus bursae shows less curvature than in the other two, and is also much more densely spined.

Comparison of the types of bryce Walker and rosea Hering suggests that these are synonyms; they have identical wing patterns, with the same amount of rosy red in the transverse FW band, and a similar peach-colored central area near the band's center (see Bryce Group species key). Both type localities are on the Rio Tapajós of Brazil—“Rio Tapajos” for bryce, and “Itaituba” for rosea. Hering's rosea is here made a junior synonym of E. bryce. The only specimen of E. bryce at the BMNH is the Walker type. There are, however, nine examples identified as rosea in the ZMH collection, including a paratype. The male dissected for this study (JSM-1421) is from that series.

By studying material with wings strictly adhering to this characterization of E. bryce, it appears that the species is endemic to a relatively small area in the vicinity of the Rio Tapajós, a large tributary that empties into the Amazon River near Santarém. The question becomes which name to apply to material from further west up the Amazon. Specimens from there do not show the pinkish-red FW band. I leave resolution of that issue to future revisionary work. I was unable to locate the type of subdraconis, described by Bryk (1953) as a subspecies of E. bryce.

Distribution

Brazil (AMNH, BMNH, CUIC, USNM, ZMH).

Dissected

♂, Brazil, Itaituba, ZMH (genitalia slide no. JSM-1421); ♂, Brazil, Pará, Itaituba, Rio Tapajós, Feb 1922, USNM (genitalia slide no. JSM-1728); ♀, Brazil, AMNH (genitalia slide no. JSM-134); ♀, Brazil, Rio Tapajós, leg. B. Pohl, CU Lot 819, Sub 309, CUIC (genitalia slide no. JSM-1730).

Ephialtias draconis (Druce), new combination

Figures 311G, 311H, 312, 313; plate 29 [EX]

Actea draconis Druce, 1885a: 145, pl. 14, fig. 6.

Type Locality

Panama, San Juan.

Type

Syntype ♂ (ZMH).

Discussion

Ephialtias draconis is common in the Canal Zone of Panama (Forbes, 1939a). I have observed it along the Pipeline Road (1993), flying during the day in considerable numbers. Because of the many entomologists who have worked in Panama over the years, E. draconis is well represented in most major collections. However, when attempting to delimit geographical boundaries for the species, serious problems arise. Aside from the vast amount of Panamanian material, only two examples of E. draconis are known from Colombia (USNM). None has been captured in Ecuador. The next occurrence of the species as one travels south along the Andes is southeastern Peru and northeastern Bolivia. One might presume that this material represents an unnamed taxon. The HW in these southern examples is more salmon colored than the rosy red HW typical of E. draconis. However, their genitalia (JSM-717, 718) are indistinguishable from those of Panamanian specimens. I have therefore listed Bolivian material as E. draconis. Clearly, more work is needed to refine the taxonomic and geographical boundaries of this taxon.

Interestingly, E. draconis has been listed as the only dioptine occurring in Jamaica (Prout, 1918), based on a single specimen in the BMNH collection. Todd (1981) questioned the veracity of that record. He discovered the specimen in question; it was purchased in 1888 from a “Mr. Mathew”. The relevant BMNH accession records mention Lepidoptera from Australia and Chile, but nothing from Jamaica. Todd concluded that the specimen is mislabeled. There have been no subsequent reports of E. draconis occurring on Jamaica, or on any other island in the Antilles.

The type of E. draconis, one of the few Druce types residing at the ZMH, originally came from the Staudinger Collection (Druce, 1885a).

This is one of only two members of the Josiini whose larvae are known to feed on Turnera (Turneraceae; see table 6). Ephialtias draconis male and female genitalia, as well as its caterpillars, are illustrated and discussed in Miller (1996).

Distribution

Panama (AMNH, BMNH, CAS, CUIC, MUSM, NMW, UCB, USNM, ZMC); Colombia (USNM); Peru (AMNH, BMNH); Bolivia (AMNH, CMNH).

Dissected

♂, Panama, Canal Zone, Barro Colorado Island, CAS (genitalia slide no. JSM-703); ♂, Panama, Canal Zone, Barro Colorado Island, AMNH (wing slide no. JSM-171; genitalia slide no. JSM-133); ♂, Panama, Canal Zone, Barro Colorado Island, AMNH (genitalia slide no. JSM-284); ♂, Bolivia, Las Juntas, leg. Steinbach, CMNH (genitalia slide no. JSM-718); ♀, Panama, Canal Zone, Barro Colorado Island, AMNH (genitalia slide no. JSM-285); ♀, Panama, Canal Zone, Barro Colorado Island, 22 Mar 1945, leg. C.D. Michener, AMNH (genitalia slide no. JSM-710); ♀, Bolivia, Las Juntas, leg. Steinbach, CMNH (genitalia slide no. JSM-717).

Ephialtias tenuifascia (Prout), new combination revised status

Figure 318F–J; plate 29

Josia Lyces tenuifascia Prout, 1918: 422.

Type Locality

Guyana, Essequibo River.

Type

Syntype ♂, leg. Whiteley (BMNH).

Discussion

Prout (1918) described tenuifascia as a subspecies of draconis Druce, noting its narrower transverse FW band. I here raise Prout's taxon to species status, based on genitalia differences between the two. The long, spiny sclerotized band in the female CB is much more angled in E. tenuifascia (fig. 318G), and the anterior apophysis on male St8 is longer and narrower (fig. 318J). The taxonomy of the Bryce Group needs attention. However, recognizing three species—bryce from Amazonian Brazil, draconis from Panama and eastern South America, and tenuifascia, endemic to northern South America and Trinidad—is a more accurate reflection of the group's systematics and geographical distribution than the hypotheses of previous authors.

Although E. tenuifascia is clearly endemic to Guyana, being well documented in the collections listed below, its geographical boundaries beyond there are unclear. Surprisingly, no specimens have surfaced from French Guiana. I list a BMNH specimen from Venezuela as “tenuifascia” pending confirmation. Finally, a female in the AMNH collection is labeled: “Lady Chancellors Rd., St. Anne, Trinidad, 4 Mar 1933, leg. A.S. Pinkus”. The latter appears to be correctly referred to E. tenuifascia, so the moth probably occurs broadly across northern South America, not only in Guyana.

Distribution

Guyana (BMNH, CMNH, USNM); Venezuela (BMNH); Trinidad (AMNH).

Dissected

♂, Guyana, leg. R. Doll, USNM (genitalia slide no. JSM-1541); ♀, Guyana, Acarai Mtns., Sipu R., 900 ft, 1°25.1′N, 58°57.2′W, 24 Oct–12 Nov 2000, leg. S. Fratello et al., USNM (genitalia slide no. JSM-1542); ♀, Guyana, Essequibo, Wineperu, 18–24 Mar 1969, Duckworth & Dietz, USNM (genitalia slide no. JSM-1729).

LYCES WALKER, 1854

Figures 320–Figure 321Figure 322Figure 323Figure 324325; plates 29–Plate 3031

Figure 320

Morphology of Lyces (♂♂). A, head of L. angulosa, lateral view; B, head of L. patula, lateral view; C, head of L. patula, frontal view; D, head of L. patula, posterior view; E, L. ena labial palpus; F, L. striata labial palpus; G, L. attenuata labial palpus; H, L. ariaca head, lateral view; I, L. angulosa wings (illustration by J.S. Miller).

Figure 321

Genitalia of Lyces. A, L. ariaca ♂ (JSM-315); B, L. ariaca aedeagus; C, L. ariaca ♀ (JSM-316); D, ♂ Tg8 of L. ariaca; E, ♂ St8 of L. ariaca; F, L. eterusialis ♂ (JSM-503); G, L. eterusialis aedeagus; H, L. eterusialis ♀ (JSM-1378); I, ♂ Tg8 of L. eterusialis; J, ♂ St8 of L. eterusialis.

Figure 322

Genitalia of Lyces angulosa (♂ JSM-713, ♀ JSM-1485). A, ♂ genitalia; B, ♂ St8; C, aedeagus; D, ♂ Tg8; E, ♀ genitalia (illustration by A. Trabka).

Figure 323

Genitalia of Lyces. A, L. attenuata ♂ (JSM-329); B, L. attenuata aedeagus; C, L. attenuata ♀ (JSM-330); D, ♂ Tg8 of L. attenuata; E, ♂ St8 of L. attenuata; F, L. striata ♂ (JSM-1489); G, L. striata aedeagus; H, L. striata ♀ (JSM-1715); I, ♂ Tg8 of L. striata; J, ♂ St8 of L. striata.

Figure 324

Genitalia of Lyces. A, ♂ of L. minuta (JSM-1467); B, aedeagus of L. minuta; C, ♀ of L. minuta (JSM-721); D, ♂ Tg8 of L. minuta; E, ♂ St8 of L. minuta; F, ♂ of L. solaris (JSM-1391); G, aedeagus of L. solaris; H, ♀ of L. solaris (JSM-1392); I, ♂ Tg8 of L. solaris; J, ♂ St8 of L. solaris.

Figure 325

Genitalia of Lyces. A, ♂ of L. longistria (JSM-1489); B, ♀ of L. longistria (JSM-1715); C, aedeagus of L. longistria; D, ♂ Tg8 of L. longistria; E, ♂ St8 of L. longistria; F, ♂ of L. tamara (JSM-1674); G, ♀ of L. tamara (JSM-1675); H, aedeagus of L. tamara; I, ♂ Tg8 of L. tamara; J, ♂ St8 of L. tamara.

Lyces Walker 1854: 366. Type species: Lyces angulosa Walker, 1854 (by original designation).

Leptactea Prout, 1918: 397. Type species: Actea minuta Druce, 1885b (by original designation). New synonymy.

Diagnosis

Taken as a whole, Lyces shows remarkable variation in wing pattern. This variability includes: Species with a yellow or orange-yellow transverse FW band and a corresponding yellow or orange-yellow central area in the HW (e.g., L. angulosa and L. constricta, pl. 29); taxa with longitudinal yellow or orange FW and HW stripes (e.g., L. striata and L. cruciata, pl. 31); as well as species with a yellow transverse FW band and a completely dark HW (e.g., L. ignorata and L. ena, pl. 29). Anomalous wing patterns include those of L. banana, L. fluonia, and L. vulturata (pls. 29, 30)—all of which exhibit an oblique yellow FW stripe—as well as L. minuta (pl. 31), where the orange markings of the FW include an unusual longitudinal streak behind vein Sc, and a quadrate transverse band. Lyces solaris (pl. 30) possesses wings reminiscent of some Scea species (e.g., S. angustimargo, pl. 33). This spectrum of patterns notwithstanding, comprehensive analysis of adult morphology strongly supports the monophyly of a Lyces clade (fig. 283), separate from Josia.

Lyces species can be recognized by the following suite of characters: labial palpus porrect, barely ascending (figs. 320A, B, E–H); Lp2 almost straight, slightly shorter than Lp1; Lp3 bullet shaped, rarely elongate (L. attenuata); front and labial palpus uniformly dark brown to blackish brown, rarely (some Patula Group species) with white or orange markings; eye moderately large (fig. 320A–D), postgena relatively wide; FW vein M₁ arising from anterolateral angle of DC (fig. 320I), either touching or almost touching base of radial sector; male and female abdomen elongate, acute distally (especially in males). The male genitalia are variable, but always have an extremely large, membranous BO (e.g., figs. 322A, 323A), narrow, laterally compressed socii, and a relatively long vesica (e.g., fig. 323B, G). In females, the base of the CB usually forms a sclerotized, vaselike structure near its junction with the DB (figs. 323C, 323H, 324C, 324H, 325B, 325G). Internally, this structure is densely spinose.

In Lyces species with longitudinal FW stripes, such as members of the Patula Group (pl. 30), it is important to recognize that these stripes fall along the wing more anteriorly than the ones found in Josia (Forbes, 1931; Miller, 1996). In Lyces, the majority of the stripe lies within the DC, anterior to the cubitus. The stripe in Josia, on the other hand, invariably straddles the cubitus (fig. 336H), its trailing margin extending posteriorly well beyond the anal fold.

Redescription

Male. FW length  =  10.5–23.0 mm. Head (fig. 320A–H): Labial palpus extending almost horizontally, barely ascending; Lp1 relatively long, curved upward, with a wedge-shaped fringe of long scales below; Lp2 straight and fairly wide, shorter than Lp1, usually tightly scaled, without a fringe; Lp3 bullet-shaped; clypeus scaleless; frontal scales short, pointing ventrally from below antennal bases, then horizontally above clypeus; eye moderately large, postgena wide; scales of vertex relatively short, pointing anteriorly; antennal rami thin, variable in length, ranging from moderately long (e.g., L. striata) to short (L. gopala), approximately 12 terminal segments simple; antennal shaft widened from base, tapering distally.

Thorax: Epiphysis moderately wide, approximately two-thirds length of tibia, not extending beyond its apex; tegula moderately long, slightly less than two-thirds length of mesoscutum, ventral angle acute; metathoracic tympanum kettledrum shaped, drum extremely large, occupying almost entire metepimeron; tympanal membrane large, ovoid, oriented horizontally.

Forewing (fig. 320I; pls. 29–Plate 3031): Broad to slightly elongate; veins Rs₂–Rs₄ in the pattern [2+3]+4; M₁ arising from anterolateral angle of DC, either touching (e.g., L. aurimutua, L. flavissima) or almost touching (e.g., L. angulosa, L. ena) base of radial sector; DC much longer than one-half FW length; veins M₃ and CuA₁ long stalked; wing pattern variable, ground color dark brown to black; most species either with a yellow or orange-yellow transverse band crossing distal margin of DC (Angulosa Group), or with a longitudinal stripe (Patula Group).

Hind wing (fig. 320I; pls. 29–Plate 3031): Broad and full, rounded at apex; M₃ and CuA₁ long stalked; pattern highly variable; ground color dark brown to black, central area concolorous (e.g., L. fluonia, L. banana), with a yellow or orange central area of various sizes and shapes (e.g., L. attenuata), or with a longitudinal orange stripe (e.g., L. aurimutua).

Abdomen: Elongate, acute distally; often uniformly brownish gray to blackish gray above and below; some species with thin, longitudinal whitish (e.g., L. longistria) or orange (L. cruciata) pleural stripes; venter sometimes either with a faint white stripe along midline (e.g., L. longistria), or completely white (e.g., L. ena, L. cruciata).

Terminalia (figs. 321A, B, D–G, I, J; 322A–D; 323A, B, D–G, I, J; 324A, B, D, E, F, G, I, J): Tg8 short, roughly quadrate, occasionally tapered posteriorly (e.g., L. fluonia, L. vulturata), anterior margin simple, posterior margin either with a shallow, wide excavation (Angulosa Group), or with a deep, U-shaped mesal excavation (Patula, Eterusialis groups); St8 longer and wider than Tg8, narrowing posteriorly, sometimes markedly so (e.g., L. flavissima); anterior margin of St8 with a short, acute mesal apodeme (longer in L. flavissima), posterior margin either simple (Angulosa Group), or with a deep, often narrow, U-shaped or V-shaped mesal excavation; socii/uncus complex highly variable in shape, narrowly attached to tegumen; uncus usually long, thin and acute at apex, arching downward from base, sometimes with a large, bladelike dorsal flange (e.g., L. angulosa, L. ariaca, L. ena); socii variable in shape, usually laterally compressed, sometimes acutely tapered (e.g., L. striata, L. attenuata), occasionally with bristlelike dorsal setae (e.g., L. ariaca); tegumen wide (Angulosa Group), or moderately wide to narrow; tegumen taller than vinculum, or rarely shorter (Eterusialis Group); vinculum narrow; saccus apparently absent, ventral margin of genitalia broadly U-shaped; valva large, completely membranous except for costa and base, BO extremely large, occupying almost entire valva, expanded along outer margin; costa narrow, variable in length, sometimes extremely short (Eterusialis Group), frequently with an acute (e.g., L. patula, L. ariaca) or bladelike (e.g., L. cruciata) process approximately one-half way out; apex bearing a short, sclerotized process of variable shape, usually somewhat truncate (e.g., L. fluonia, L. solaris), rarely acute (L. angulosa); valva bases frequently joining below ventral margin of genitalia to form a W-shaped structure; arms of transtilla wide, pointing sharply downward, sometimes crenulate along dorsal margin (e.g., L. cruciata); transtillar arms meeting at midline to form a large, deeply concave medial plate, its lateral margins sometimes curled inward; most species with a triangular patch of spicules in manica above transtillar plate, spicules sometimes long and densely arranged (e.g., L. cruciata); aedeagus extremely wide and bulbous at base, abruptly tapered distally; apex of aedeagus forming a small point; vesica as long as, or longer than aedeagus, variable in shape, always extending upward; vesica wide and arising at a sharp angle from aedeagus (e.g., Angulosa and Eterusialis groups), or long and narrow, curled anteriorly (Patula Group); vesica with numerous spinelike cornuti, variable in size; often with a single, isolated cornutus at apex of vesica much larger than others (many Patula Group species).

Female. FW length  =  9.5–24.0 mm. Head: Similar to male, except antenna ciliate.

Thorax: Similar to male.

Forewing: Longer and broader than male (pls. 29–Plate 3031), yellow/orange-yellow maculations wider.

Hind wing: Longer and broader than male, outer margin more rounded, yellow/orange-yellow pattern wider; frenulum comprising two bristles.

Abdomen: Elongate as in male, but wider.

Terminalia (321C, H; 322E; 323C, H; 324C, H): Tg7 long and wide, slightly narrower posteriorly, anterior margin simple, posterior margin with a deep, wide, often poorly defined, U-shaped mesal excavation; posterior two-thirds of Tg7 more sclerotized than anterior third, a transverse depression dividing the two sections; St7 long and wide, equal in length to Tg7, anterior margin simple, posterior margin with a deep, narrow U-shaped mesal excavation; Tg8 completely membranous, membrane fragile; AA extremely thin, either short and delicate (Angulosa and Eterusialis groups) or moderately long, straight (Patula Group); A8 pleuron membranous, with a thin, straplike sclerite along anterior margin; PP moderately long, extremely thin; PA small, almost always with a dorsal lobe, sometimes (Angulosa Group) with ventrolateral margin heavily sclerotized, forming a bladelike lateral flange; area of PVP either completely membranous, or rarely lightly sclerotized; a short, half-moon–shaped antevaginal plate sometimes present (e.g., L. ena, L. auriaca); DB short, narrow or moderately wide, membranous; CB large, sometimes (e.g., L. flavissima, L. angulosa) extremely so, roughly ovoid, midsection broadly sclerotized above; base of CB variable, either with a complex series of folds and a large, laterally compressed dorsal appendix (Angulosa and Eterusialis groups), or narrowed and smoothly sclerotized to form a vaselike constriction (Patula Group); CB almost always with a large set of internal spines near base (absent in L. angulosa); signum located ventrally, bird shaped, comprising a wide, transverse sclerite with long, internal spines at each end, central portion of signum protruding from CB membrane; DS arising dorsally from a large, laterally compressed, often sclerotized, appendix, base of DS often curled, pigtail shaped.

Distribution

The Eterusialis Group (two species) is strictly South American, whereas the other two Lyces species groups occur from southern Mexico south to Bolivia and southeastern Brazil. The majority of Patula Group species are endemic to Andean cloud forests at elevations between 1500 and 3000 meters—a few have been recorded from 4000 meters. The remaining taxa, such a L. ena (Angulosa Group) and L. cruciata (Patula Group), live in lowland forests.

Biology

Immature stages have been discovered for 10 Lyces species (table 6)—more than a third of the genus. Their hosts seem to contrast with those of Josia. Almost all Lyces hosts belong in the Passiflora subgenus Granadilla, whereas Josia larvae are associated most commonly with Passiflora species currently assigned to the subgenus Decaloba. A similar dichotomy occurs within Heliconius (Smiley, 1985); the erato-charitonia species group specializes on Decaloba, whereas the numata-melpomene group is restricted to Granadilla. Plants in these two subgenera show basic differences. The leaves of most Granadilla species are glabrous and thick, whereas those in Plectostemma are pubescent and delicate (see, e.g., Holm-Nielsen et al., 1988). Whether this difference in host-plant utilization between Lyces and Josia is correlated with divergence in plant secondary chemicals within Passiflora, or whether other biological determinants are at play, remains to be investigated. Plant chemistry is only one of the factors determining Passiflora species usage in Heliconius (Smiley and Wisdom, 1985), other variables being plant nutritive value and plant-associated predators. It will be interesting to fill in the missing host-plant data for the remainder of the Josiini in search of general patterns.

In contrast to Josia caterpillars (pl. 39L–O), Lyces larvae generally show simpler abdominal coloration—less reticulate and more geometric (pl. 39J, K). Other features of their morphology are the same (Miller and Otero, 1994; Miller, 1996).

Discussion

In previous treatments of the Dioptinae, the taxa now residing in Lyces were placed in Josia, which then included many more species than are recognized here. Earlier authors (Prout, 1918; Hering, 1925; Forbes, 1931) were obviously confused by the diversity of wing patterns occurring in the Josiini, and with few other character systems available to them, were unsure how to proceed regarding the group's taxonomy. Their solution was to create a large diverse genus called Josia, which we now know, was polyphyletic (Miller, 1996; Miller, Brower and DeSalle, 1997).

The genus name Lyces, prior to now not widely used, is attributed to Walker (1854), who described two species in the genus, angulosa and flavissima. Kirby (1892) later designated angulosa as the type, listing six species of Lyces, only four of which—angulosa, eterusialis, flavissima, and fornax—currently reside in the Josiini. According to the classification proposed here (appendix 2), all four of these are correctly assigned to Lyces. Schaus (1892) added a fifth, describing maera from Brazil, a species that Hering (1925) subsequently moved to its proper position in Erbessa. Prout (1918) placed Lyces in synonymy with Josia, applying the name to his “Section III”. Prout recognized 13 species in “Section III”, almost all of which belong in Lyces as conceived here. Both Hering (1925) and Bryk (1930) abandoned the use of Lyces in any capacity. According to my concept, Lyces includes 25 species in three species groups.

Unlike some genera of the Dioptinae, where there are numerous undescribed species and few names, I discovered fewer species in Lyces than names. In this paper, seven new junior synonyms are proposed, and one new species is described. Additional study will, in all likelihood, reduce the number of valid species even further. I know of only a handful of undescribed Lyces in collections, most of these represented by a single specimen collected at an exotic locality.

The three species groups defined here are built largely on characters from genitalia. Wing patterns vary dramatically within Lyces, but other features, such as the labial palpi and male antennae, are relatively uniform throughout (fig. 320). By and large, I believe these species groups are robust. At the very least they can act as hypotheses for future testing.

Type Locality

Colombia, “Bogotá”.

Type

Syntype ♂, ex Stevens Collection (BMNH).

Discussion

Lyces eterusialis is similar in all respects to L. flavissima, the only difference being that the markings of L. eterusialis are a dark, rich yellow, whereas those of L. flavissima are lighter lemon yellow to almost greenish yellow. When genital morphology for the two taxa is compared, they are impossible to separate. It thus appears that L. eterusialis and L. flavissima may represent color forms of a single species. I leave resolution of that question for future research. Material identified as Lyces eterusialis, uncommon in museum collections, is restricted to eastern Colombia in the vicinity of Bogotá.

Distribution

Colombia (AMNH, BMNH, MNHN, NMW, OUMNH, USNM, ZMH).

Dissected

♂, Colombia, Bogotá, leg. Lindig, MNHN (genitalia slide no. JSM-503); ♂, Colombia, Susumico, USNM (genitalia slide no. JSM-1377); ♀, Colombia, Villavicencio, nr. Bogotá, Apr 1917, leg. Apollinaire, USNM (genitalia slide no. JSM-1378).

Lyces flavissima Walker, revised status

Plates 30, 39J [EX]

Lyces flavissima Walker, 1854: 367.

Type Locality

Venezuela.

Type

Syntype ♀, ex Dyson Collection (BMNH).

Discussion

As noted above, L. flavissima is extremely similar to L. eterusialis, with one observable difference—the color of the FW and HW markings (pl. 30). The two taxa exhibit seemingly identical genital morphology. Furthermore, in both species the rami of the male antennae are extremely short. This characteristic can also be found in some members of the Patula Group, such as L. longistria.

My dissections suggest that specimens from eastern Ecuador, which exhibit lemon-yellow markings, are conspecific with material from Venezuela, the provenance of the L. flavissima type. A single known example from Guyana (Mt. Ayanganna; 05°24.1N, 59°57.4W; 4500–5500 ft; USNM), matches these. Assuming the two are even distinct species, the range of L. flavissima thus appears to be more extensive than that of L. eterusialis, so far recorded exclusively from Colombia. Lyces flavissima, discussed in Miller (1996), was reared by L.D. Otero (1992) in Venezuela on Passiflora oerstedi and P. ambigua (table 6).

An undescribed species belonging in the Eterusialis Group, closely related to but markedly distinct from L. flavissima, occurs in Venezuela. I know this taxon from a single specimen: ♂, Venezuela, Sucre, Elvecia, near Mt. Turumquire, leg. G. Netting (CMNH). Its wing pattern generally resembles that of L. flavissima, but the lemon-yellow FW band is displaced as it crosses the stem of M₃+CuA₁, and the yellow central area of the HW is scalloped along its outer margin. The male genitalia of this specimen (JSM-433) show all the characteristics of the Eterusialis Group, but differ markedly from the other two species, especially in the configuration of the vesica.

Distribution

Venezuela (AMNH, BMNH, IZA, OUMNH, USNM, ZMH); Guyana (USNM); Ecuador (AMNH, CMNH, NMW, USNM, ZMH).

Dissected

♂, Venezuela, Aragua, Rancho Grande, 1100 m, 27 Jul 1992, leg. L.D. Otero, ex ova on Passiflora oerstedi (genitalia slide no. JSM-652); ♂, Venezuela, Cucuta, USNM (genitalia slide no. JSM-624); ♂, Venezuela, ZMH (genitalia slide no. JSM-638); ♀, Venezuela, Aragua, Rancho Grande, 1100 m, 27 Jul 1992, leg. L.D. Otero, ex ova on Passiflora oerstedi (genitalia slide no. JSM-653); ♀, Venezuela, Aragua, Rancho Grande nr. Maracay, 30 Apr 1946, Tropical Research Station, N.Y. Zoological Society (genitalia slide no. JSM-623); ♀, Venezuela, Cucuta, USNM (genitalia slide no. JSM-625); ♀, Ecuador, Pastaza, Río Pastaza, ZMH (genitalia slide no. JSM-636); ♀, “Amaz.”, ZMH (genitalia slide no. JSM-638).

Type Locality

“Brazil”.

Type

Syntype ♀, ex Stevens Collection (BMNH).

Ephialtias morena Warren, 1906: 411. New synonymy.

Type Locality

Brazil, Rio Grande do Sul.

Type

Holotype ♂ (USNM type no. 9166).

Discussion

Because angulosa is the type species of Lyces, a genus reinstated in the current work, considerable effort was spent attempting to resolve its identity. There appears to be a significant amount of intraspecific variation in the width and shape of the yellow-orange transverse FW band, and in the shape and size of the yellow-orange HW area, all of which contributed to making this a particularly difficult identification problem.

Lyces angulosa can be characterized by its relatively large size (FW length  =  16.5–21.5 mm), by having a slightly curved, light yellow to lemon-yellow transverse band in the FW, and by having a mostly rounded central area of the HW (pl. 29). The majority of material in museum collections from Pelotas, Brazil, identified by C. Biezanko as “L. vittula”, a species now in the new genus Proutiella, is instead L. angulosa. A female from Teresopolis, Brazil, somewhat further north than the majority of material examined, matches the type precisely. Its genitalia (JSM-1492) are indistinguishable from more southerly females. Lyces angulosa occurs exclusively in the Atlantic forests of central and southeastern Brazil.

Dissection of the USNM type of Ephialtias morena Warren (JSM-1379), from Rio Grande do Sul, previously regarded as a distinct species (Bryk, 1930), shows that its male genitalia are identical with those of L. angulosa. Its wing pattern is also the same. For these reasons I propose morena as a new synonym of L. angulosa.

The following is an undescribed Lyces species, perhaps related to L. angulosa: ♂, Brazil, Rio de Janeiro, 7 Aug 1932, leg. P. Gagarin, MPM (genitalia slide no. JSM-556); ♀, Brazil, Rio de Janeiro, MPM (genitalia slide no. JSM-557).

Distribution

Brazil (AMNH, BMNH, FML, MCZ, MNHN, MPM, NMW, OUMNH, PMNH, USNM, VOB, ZMH).

Dissected

♂, Brazil, Pelotas, Rio Grande do Sul, 27 Jan 1952, leg. C. Biezanko, AMNH (genitalia slide no. JSM-426); ♂, Brazil, Pelotas, Rio Grande do Sul, 12 Jan 1957, leg. C. Biezanko, AMNH (genitalia slide no. JSM-1503); ♂ type of morena Warren, USNM (genitalia slide no. JSM-1379); ♂, [no data], USNM (genitalia slide no. JSM-713); ♂, [no data], H.G. Dyar collection, USNM (genitalia slide no. JSM-719); ♂, Brazil, Pelotas, Rio Grande do Sul, 12 Dec 1954, leg. C. Biezanko, AMNH (wing slide no. JSM-1665); ♀, Brazil, Rio de Janeiro, Teresopolis, 22°27′S, 42°59′W, 1000 m, 29 Sept 1994, leg. A. Caldas, USNM (genitalia slide no. JSM-1492); ♀, Brazil, Pelotas, Rio Grande do Sul, 16 Jan 1955, leg. C. Biezanko, AMNH (genitalia slide no. JSM-427); ♀, Brazil, H.G. Dyar collection, USNM (genitalia slide no. JSM-720); ♀, [no data], ex Colln. Brooklyn Museum, AMNH (genitalia slide no. JSM-1485).

Lyces ariaca (Druce), new combination

Figure 320H, 321A–E; plate 29 [EX]

Ephialtias ariaca Druce, 1885a: 147, pl. 13, fig. 27.

Type Locality

Mexico, Yucatan, Valladolid.

Type

Syntype ♀, leg. Gaumer (BMNH).

Ephialtias coatepeca Schaus, 1889: 192. New synonymy.

Type Locality

Mexico, Coatepec, Paso de San Juan.

Type

Syntype ♂/♀ (USNM type no. 18706).

Discussion

Lyces ariaca appears to be widely distributed throughout the northern countries of Central America. As occurs in L. angulosa (fig. 322A) and L. constricta, the uncus of L. ariaca bears a dorsal process (fig. 321A), in this species more spatulate than bladelike. Precisely the same spatulate process occurs in L. ena and L. banana. The taxon most similar in size and wing pattern to ariaca is L. latistriga Hering (pl. 29), from Peru, so far known exclusively from the ZMH type. Its relationship to L. ariaca is unknown.

There is no question regarding this moth's identity. Schaus (1889: 192), in his original description of Ephialtias coatepeca, noted that it is “Closely allied to Ephialtias ariaca Druce”. My examination of the L. coatepeca type material shows Schaus' taxon to be a newly recognized synonym of L. ariaca; no characters distinguish the two.

Distribution

Mexico (AMNH, BMNH, FNHM, LACM, MNHN, UCB, USNM, ZMH); Honduras (BMNH, USNM); Belize (AMNH); Guatemala (BMNH, CMNH, USNM).

Dissected

♂, Mexico, Veracruz, Atoyac, BMNH (genitalia slide no. JSM-315); ♀, Mexico, Veracruz, Jalapa, BMNH (genitalia slide no. JSM-316).

Lyces banana (Warren), new combination

Plate 29

Josia banana Warren, 1901: 441.

Type Locality

[Not known].

Type

Syntype ♂/♀ (BMNH).

Discussion

Although the type of L. banana bears no locality data, the identity of this species is clear—its wing pattern is distinctive. Lyces banana, endemic to Brazil, is one of the rarest species in the Josiini. It is known from only seven specimens: four at the BMNH (2 ♂♂, 2 ♀♀), including the types; 1 male at the MCZ; 1 female at the NMW; and 1 female at the ZMH. All of this material was collected over 100 years ago.

Traditionally (Prout, 1918; Hering, 1925; Bryk, 1930), L. banana (pl. 29) was thought to be a close relative of L. fluonia (pl. 30). That theory was apparently based on a general similarity of the unusual, obtuse FW markings in the two species. However, the male genitalia of L. banana exhibit a dorsal crest on the uncus, identical to the one found in L. ena and L. ariaca (fig. 321A). In fact, the male genitalia of these species are similar in every way. The implied relationship between L. banana and L. ariaca, a species with FW cross-bands typical of those found throughout Lyces, is fascinating. Wing-pattern similarities shared by L. banana and L. fluonia thus appear to have evolved through convergence.

Distribution

Brazil (BMNH, MCZ, NMW, ZMH).

Dissected

♂, “Ex. Musaeo Ach. Guénée”, BMNH (genitalia slide no. JSM-324).

Lyces constricta (Warren), new combination

Plate 29

Ephialtias constricta Warren, 1901: 440.

Type Locality

Brazil, Bahia.

Type

Syntype ♂ (BMNH).

Discussion

Warren (1901), in his original description of constricta, distinguished it from L. angulosa by being slightly smaller and by having a narrower, more curved stripe in the HW. The BMNH specimens of constricta I dissected match the type precisely and are from the state of Bahia in Central Brazil, where the type was collected. Their genitalia indicate an extremely close relationship with L. angulosa; the two appear to be sister species. They can be separated by wing size, as well as by several features of their male and female genitalia. Distinguishing characters of the genitalia include: thinner socii in L constricta males, lacking a dorsal flange as in L. angulosa (fig. 322A); a blunt (constricta) rather than acute (angulosa) valva apex; external process of signum rounded in L. constricta, but broadly truncate in L. angulosa (fig. 322E). The two species are apparently sympatric throughout their ranges.

Distribution

Brazil (AMNH, BMNH, NMW, VOB).

Dissected

♂, Brazil, Bahia, BMNH (genitalia slide no. JSM-206); ♂, Brazil, Rio Grande do Sul, Sao Leopoldo, 4 Oct 1958, MPM (genitalia slide no. JSM-554); ♀, Brazil, Bahia, leg. Fruhstorfer, BMNH (genitalia slide no. JSM-207); ♀, Brazil, Rio Grande do Sul, Sao Leopoldo, 4 Oct 1958, MPM (genitalia slide no. JSM-555).

Lyces ena (Boisduval), new combination

Figure 320E; plate 29

Retila ena Boisduval, 1870: 95.

Type Locality

French Guiana, Cayenne.

Type

Syntype ♀ (BMNH).

Ephialtias tryma Schaus, 1896: 154.

Type Locality

Trinidad.

Type

Syntype ♂ (USNM type no. 11574).

Discussion

As Hering (1925) noted, L. ena is perhaps the most common species of Josiini in collections. The taxon is widespread, occurring from Panama south to Brazil and Peru, and is one of only three josiines found on Trinidad. Based on genital similarities, L. ena belongs in a subclade of the Angulosa Group along with L. ariaca, L. banana, and L. enoides. All of these are separable by their wing patterns (pl. 29).

However, Lyces ena (pl. 29) and Proutiella infans (pl. 26) are essentially indistinguishable, sharing nearly identical wing pattern, body coloration, and size. Both show variation in the width and shape of the yellow transverse FW band. It has long been assumed by curators at collections around the world that all josiines of this size and appearance represent a single species—L. ena. Not until August 2005, when the type of infans was studied, did it become clear to me that two species were actually involved. Upon closer inspection, wing venation and abdominal coloration provide easy means for separation. As is true of other Lyces species (fig. 320I), FW vein M₁ in L. ena arises from the DC, almost touching—but separate from—the base of the radial sector; the abdominal venter is completely white or whitish gray; the thoracic scales are short, but they are not ovoid and appressed. In Proutiella infans, on the other hand, M₁ is stalked with Rs₁–Rs₄, arising basal to Rs₁ (see fig. 284C); the abdominal venter is white with a wide gray stripe along the midline; and the thoracic scales are extremely short, ovoid, and appressed.

Lyces ena has been recorded from two Passiflora species (table 6): in French Guiana, caterpillars were found on Passiflora laurifolia, whereas in Panama they feed on P. vitifolia. These plants belong in different Passiflora subgenera according to Killip (1938).

The type of tryma Schaus shows identical characteristics, including male genital morphology (JSM-1533), with those of L. ena, thus confirming Prout's (1918) synonymy.

Distribution

Panama (AMNH, BMNH, FNHM, USNM); Venezuela (AMNH, MNHN, NMW, USNM); Trinidad (AMNH, BMNH, CMNH, CUIC, MNHN, OUMNH, USNM); Guyana (USNM); French Guiana (AMNH, BMNH, CUIC, MNHN, OUMNH, USNM, ZMH); Brazil (AMNH, BMNH, CMNH, CUIC, MNHN, NMW, USNM, ZMC, ZMH); Colombia (MNHN, USNM); Peru (AMNH, LACM, MUSM, ZMH).

Dissected

♂ syntype of tryma Schaus, USNM (genitalia slide no. JSM-1533); ♂, French Guiana, Cayenne, Mar 1917, CMNH (genitalia slide no. JSM-708); ♂, Brazil, AMNH (genitalia slide no. JSM-139; wing slide no. JSM-176); ♀, French Guiana, Mana River, May 1917, CMNH (genitalia slide no. JSM-709); ♀, Panama, Pipeline Road, Canal Area, 9 May 1992, leg. C. Penz, ex egg on Passiflora vitifolia, USNM (genitalia slide no. JSM-504); ♀, Peru, Upper Río Huallaga, 20 Jul 1928, AMNH (genitalia slide no. JSM-140); ♀, Trinidad, BWI, Arima Valley, 25 Feb 1957, AMNH (genitalia slide no. JSM-505).

Lyces enoides (Boisduval), new combination

Plate 29

Retila enoides Boisduval, 1870: 94.

Type Locality

“Honduras and Mexico”.

Type Locality

Syntype ♂ (BMNH).

Discussion

The type of L. enoides (pl. 29), which I did not dissect, is similar to L. ena except that the FW cross-band is lighter yellow. The L. enoides type may have simply faded with time. Based on wing pattern, the only other species with which it might be associated is Proutiella infans (pl. 26). However, the wing of L. enoides shows vein M₁ arising at the base of Rs₁–Rs₄, thus excluding it from Proutiella.

Boisduval, in his original description, listed Honduras and Mexico as localities for this taxon. My search of museum collections produced no specimens similar in appearance to either L. enoides or L. ena from Honduras or Mexico—Panama is the furthest north that L. ena has been recorded. According to many Lepidopterists, the provenance of Boisduval types is often questionable. These observations place the status of enoides in doubt. In all probability, L. enoides and L. ena, described by Boisduval in the same publication, are conspecific. Both Boisduval types should be dissected to resolve the issue.

Distribution

Mexico (BMNH).

Dissected

None.

Lyces ignorata (Hering), new combination

Plate 29

Josia ignorata Hering, 1925: 525, fig. 71f.

Type Locality

[Not known].

Type

Holotype ♀, ex Staudinger Collection (ZMH).

Discussion

Lyces ignorata is represented in collections solely by the female holotype (pl. 29). That specimen bears no locality data, but in his original description, Hering (1925: 525) stated that it might have originated from Colombia. Lyces ignorata is closely related to L. ena, and the two could be easily confused. However, it differs in significant ways. The type of L. ignorata, with a FW length of 15.5 mm, is slightly larger than L. ena, the largest specimens of which have a FW length of 14.5 mm. The yellow FW band of L. ignorata is also much wider. Finally, the female genitalia of the ignorata type (JSM-1761) show structural differences distinguishing them from L. ena, especially with regard to the configuration of the folds at the base of the CB and the shape of the signum. I conclude that this taxon represents a valid species, and await the discovery of additional material.

Distribution

Not known.

Dissected

Holotype ♀ (JSM-1761).

Lyces latistriga (Hering), new combination

Plate 29

Josia latistriga Hering, 1925: 526, fig. 70i.

Type Locality

Peru, Marcapata.

Type

Holotype ♀ (ZMH).

Discussion

The only specimen of L. latistriga known to me is the female holotype (pl. 29), which I did not dissect. This species is extremely similar in size and wing pattern to L. ariaca (pl. 29), from Central America. The type of L. latistriga has a FW length of 13.0 mm, and females of L. ariaca exhibit FW lengths ranging between 11.5 and 15.0 mm. Lyces latistriga also resembles L. constricta (pl. 29), a slightly larger moth from eastern Brazil. Based on its size and wing pattern, I have assigned L. latistriga to the Angulosa Group, thus associating it with L. ariaca and L. constricta. However, it is not included in the key to Angulosa Group species for lack of morphological characters.

The type locality for L. latistriga, Marcapata Peru, is located less than 100 air km due east of Cuzco (see fig. 6) at an elevation of approximately 1900 meters.

Distribution

Peru (ZMH).

Dissected

None.

Type Locality

Colombia, Siato, Río Siato, Slopes of Choco, 5200 ft.

Type

Syntype ♀, Sep 1909 (BMNH).

Discussion

This species was intitally described in Josiomorpha, now in the Arctiidae, but was then moved by Prout (1918) to Josia. Here it is newly combined with Lyces. The abdomen of the andosa syntype is missing, but two females (AMNH, USNM) match its wing pattern precisely. These are the only examples of L. andosa known to me. Males have not been found. Examination of female genitalia (JSM-1711) supports the theory that L. andosa should be recognized as a distinct species. It belongs in a tight clade of six Andean Patula Group species exhibiting yellow to yellow-orange longitudinal FW stripes, such as L. gopala (pl. 30). These are also characterized by a completely black head and thorax, without markings. Here, the only pattern on the abdomen is a thin, light gray pleural stripe, as well as a thin, whitish gray stripe along the ventral midline. The FW stripe of L. andosa almost reaches the outer margin, as in L. attenuata and L. striata, but is lighter yellow and much wider. The female genitalia of L. andosa show numerous similarities to females of L. longistria, from eastern Ecuador. These six species are so closely related that resolving their interrelationships will require a detailed analysis, one that integrates characters from adults, larvae, and DNA.

Like L. striata and L. attenuata (pl. 31), L. andosa (pl. 30) is endemic to the western slope of the Andes. Interestingly, the remaining three species of this subclade—gopala, longistria, and patula—those with a short FW stripe (pl. 30) and a prominent process on the male costa, occur on the Amazonian side.

Distribution

Colombia (AMNH, BMNH, USNM).

Dissected

♀, Colombia, Dept. Antioquia, Mesopotamia, 5000 ft, Ac. 3977, AMNH (genitalia slide no. JSM-1711).

Lyces annulata (Dognin), new combination

Plate 31

Josia annulata Dognin, 1909: 223.

Type Locality

Colombia, San Antonio, 2000 m.

Type

Syntype ♂, leg. Fassl, 22 Jul 1908 (USNM).

Josia punonis Strand, 1920: 133–134.

Type Locality

Peru, Puno, Titicaca.

Type

Holotype ♀ (not seen).

Discussion

This species is here transferred from Josia Hübner. Prout (1918) hinted that annulata Dognin and cruciata Butler might be conspecific, but left them as separate taxa. Hering (1925) concurred, suggesting that the only wing-pattern difference is a slightly broader HW stripe in L. annulata, but again retained both as species. A series of genitalia dissections from specimens spanning a wide geographical distribution, provides further support; annulata and cruciata are distinct species. Differences between them are subtle. The strongest involves the shape of the socii. These are wide in L. annulata, but narrow with an elbowlike bend in L. cruciata. The female genitalia differ in the configuration at the base of the CB, and in signum shape. No other means for separating the two taxa were found. Width of the wing stripes offers no aid.

Lyces annulata and L. cruciata are closely related to two additional taxa: L. tamara from northern Central America and L. aurimutua from Brazil. This subclade also includes the Ecuadorian endemic, L. minuta. All of these are unique among Lyces in showing an orange-yellow longitudinal band on the propleuron, as well as a transverse band of orange-yellow scales on the A1 dorsum.

I have not seen the type of punonis Strand, and so cannot vouch for its placement as a synonym of annulata. I simply follow Bryk (1930).

Caterpillars of Lyces annulata were reared in Colombia and Venezuela on three different Passiflora species (table 6), all belonging in the subgenus Granadilla (Killip, 1938).

Distribution

Venezuela (AMNH, BMNH, CMNH, FNHM, IZA, LACM, MNHN, OUMNH, USNM, ZMH); Colombia (AMNH, BMNH, USNM, ZMC, ZMH); Ecuador (AMNH, LACM, BMNH, USNM); Peru (AMNH, BMNH, CAS, CUIC, FNHM, FML, LACM, MNHN, MPM, MUSM, ZMH); Bolivia (BMNH); Brazil (AMNH, BMNH, CAS, LACM, MNHN, OUMNH, USNM, VOB, ZMC, ZMH).

Dissected

Syntype ♂ (JSM-702); ♂, Venezuela, Mérida, Carr. Tovar-Zea, 2 km de El Amparo, 1100 m, 24 Jan 1993, leg. L.D. Otero, ex ovi on Passiflora cyanea, AMNH (genitalia slide no. JSM-657); ♂, Bolivia, Reyes, 7 Apr 1895, leg. Stuart, BMNH (genitalia slide no. JSM-307); ♂, Colombia, Meta, Río Negro, nr. Guayabetal, Pipiralito, 4°15′N, 73°45′W, 1000 m, May 1977, leg. J. Mallet, D. Jackson, ex larva on Passiflora nitida, AMNH (genitalia slide no. JSM-553); ♂, Peru, Río Pacaya–Río Ucayali, Jun–Sep 1912, AMNH (genitalia slide no. JSM-1385); ♂, Brazil, Bahia, leg. D. Davies, CAS (genitalia slide no. JSM-1386); ♀, Venezuela, Mérida, Carr. Tovar-Zea, 2 km de El Amparo, 1100 m, 24 Jan 1993, leg. L.D. Otero, ex ovi on Passiflora cyanea, AMNH (genitalia slide no. JSM-658).

Lyces attenuata, new species

Figures 320G, 323A–E; plate 31

Diagnosis

As mentioned above with reference to L. andosa, six Andean species form a tight subclade within the Patula Group: L. andosa, L. attenuata, L. gopala, L. longistria, L. patula, and L. striata (pls. 30, 31). The identities of these are so difficult that essentially every specimen must be carefully scrutinized. Even then, identifications are often open to question. Of the six, the taxon with which L. attenuata could most easily be confused is L. striata (pl. 31). Both exhibit particularly narrow FW stripes. They are distinguished because Lyces attenuata is generally larger. Furthermore, the HW of L. attenuata shows a deep indentation along the posterior margin of the yellow area, located at the fork of M₃+CuA₁. In most specimens of L. striata, the posterior margin of the yellow area is irregular, but does not show a deep indentation at that point. Another means for separating the two species involves the labial palpi: In L. attenuata, Lp3 is quite long (fig. 320G). On the other hand, Lp3 is short and acute at its apex in L. striata (fig. 320F).

The most reliable characters for separating L. attenuata and L. striata are found in their genitalia. In males of L. attenuata, the base of the valval costa is simple (fig. 323A), whereas in L. striata the costa bears a small flange at its base (fig. 323F). Females differ markedly in the shape of the sclerotized section at the base of the CB, which is irregular in shape and densely spinose in L. attenuata (fig. 323C), but is elongate, more regular in shape and less spinose in L. striata (fig. 323H).

Description

Male. FW length  =  20.5–22.0 mm. Head (fig. 320G): Lp3 somewhat elongate, Lp2 roughly equal in size to Lp1; all features covered with steely brown scales.

Thorax: Covered with steely brown, hairlike scales shorter than those on the head.

Abdomen: Steely brown, with a thin, buff-colored longitudinal stripe along pleuron and a prominent, cream to buff-colored stripe on venter along midline.

Terminalia (fig. 323A, B, D, E): Tg8 shorter and narrower than St8, slightly tapered posteriorly, anterior margin with a wide, shallow mesal excavation, posterior margin with a large V-shaped mesal excavation; St8 large, wide, strongly tapered distally, anterior margin bearing broadly tapered mesal apodeme, posterior margin bearing a large, deep V-shaped mesal excavation; socii/uncus complex robust; uncus beak shaped; socii strongly attenuated, curved; costa of valva wide, strongly concave, upper margin simple, apex forming a robust flange; BO extremely large, internal apodeme robust; membrane above transtillar plate bearing a wide, ovoid spiculate patch; aedeagus relatively wide; vesica curving gently upward; spinelike cornuti robust, distal club-shaped cornutus short, wide.

Female. FW length  =  21.0–22.5. Body characters similar to male, except: wings slightly longer; FW and HW stripes wider; indentation near apex of HW yellow area not as pronounced.

Terminalia (fig. 323C): PA relatively large, posterior margin rounded; DB irregular in shape, not elongate (compared to L. striata; fig. 323H); DB narrow near ostium, wider at junction with CB, internal spines robust; CB large, laterally compressed; internal spines of signum short; DS arising from a strongly pigtail-shaped appendix on dorsum of CB.

Distribution

Lyces attenuata, endemic to the Andean foothills of western Colombia (BMNH, NMW, USNM), roughly co-occurs with L. andosa and L. striata. However, it seems to be more northern in distribution than L. striata, which is most commonly collected in northwestern Ecuador. Based on available museum material, Lyces attenuata has never been captured in Ecuador.

Etymology

The name applied here, attenuata, was chosen by Warren when he described this taxon as an aberration of L. striata Druce. According to him (Warren, 1901: 442), attenuata differs from striata in that the FW stripe “is narrow and of nearly uniform width throughout, and thinning out gradually almost to a point at extremity”. While his wing-pattern characterization of L. attenuata does not seem particularly robust, since many examples of L. striata show equally narrow FW stripes (pl. 31), Warren's name is honored here because it was he who first noticed the uniqueness of L. attenuata.

Discussion

Warren named attenuata as an aberration of L. striata Druce, designating a lone female in the BMNH collection, lacking locality data, as the type. In recognizing Warren's phenotype as a species distinct from L. striata and all other Lyces, that taxon is here described as new. The male specimen of L. attenuata designated here as the holotype matches the wing pattern and body size of Warren's female with precision. The ZMH and BMNH collections contain L. attenuata in fairly long series (11♂♂, 2♀♀ at ZMH; 5♂♂, 8♀♀ at BMNH). In addition, there are five specimens at the USNM and three at the NMW. Except for the holotype, literally all the material of L. attenuata in museums was captured by Anton Fassl. Now that the identity of this species has been established, additional material will undoubtedly surface in collections around the world.

Holotype

Male (fig. 323A, B, D, E; pl. 31). Colombia: Cauca: Torne, Jan 1907, leg. Paine & Brinkley (genitalia slide no. JSM-329). The type is deposited at the BMNH.

OTHER MATERIAL EXAMINED: Colombia: Valle: 1♀, Río Aguacatal, 1800 m, Feb 1909, leg. Fassl (USNM); 1♂, 2♀♀, Río Aguacatal, 2000 m, leg. Fassl (NMW). 1♀, Alto de las Cruces, 2500 m, Nov 1908, leg. Fassl (USNM, genitalia slide no. JSM-330); 1♀, Alto de las Cruces, 2000 m, Oct 1908, leg. Fassl, Dognin Collection (USNM); 1♀, San Antonio, 2000 m, Dec 1908, leg. Fassl, Dognin Collection (USNM). 1♂, 2♀♀, “Colombie”, leg. Fassl, Dognin Collection (USNM).

Lyces aurimutua (Walker), new combination

Plate 31

Josia aurimutua Walker, 1854: 292.

Type Locality

[Not known].

Type

Neotype ♂, Hardwick Bequest (BMNH).

Josia fulvia Hübner, 1822: 19, figs. 289, 290.

Type Locality

Brazil.

Type

Not seen.

Bombyx jesuita Fabricius, 1787: 586.

Type Locality

[Not known].

Type

Not seen.

Discussion

All authors since Prout (1918) have recognized aurimutua Walker as a valid species, while the status of fulvia Hübner (1822)—described from Brazil—is listed as questionable. Writers earlier than Prout, including Walker (1854) and Druce (1885b), considered fulvia to be valid. After comparing the Walker syntypes of aurimutua with the moth figured in Hübner, I conclude that they are conspecific. Since Hübner's types have been lost or destroyed, and in the interest of stabilizing this name, I here select one of Walker's syntypes of aurimutua as the neotype. The Walker neotype, at the BMNH, bears a green, circular type label. Lyces aurimuta (Walker) is here referred from Josia as a new combination.

Comparison of available museum material with Walker's aurimutua suggests that this species is endemic to Brazil. The male and female genitalia of L. aurimutua indicate a close relationship with four other Patula Group species—L. annulata, L. cruciata, L. minuta, and L. tamara (pl. 31). These exhibit a conspicuous orange band on the dorsum of A1. This band appears to be lacking in L. aurimutua, but close inspection reveals a few scattered orange scales on A1.

Distribution

Brazil (AMNH, BMNH, CMNH, MNHN, MPM, NMW, USNM, ZMH).

Dissected

♂, Brazil, Espirito-Santo, BMNH (genitalia slide no. JSM-344); ♀, Brazil, Bahia, Ilha Grande, Oct 1900. leg. P. de la Garde, BMNH (genitalia slide no. JSM-345).

Lyces cruciata (Butler), new combination

Plate 31 [EX]

Josia cruciata Butler, 1875: 340.

Type Locality

Panama, Veragua.

Type

Syntype ♀, “75·28”, leg. Salvin (BMNH).

Discussion

This is another Patula Group species taken from Josia Hübner and placed in Lyces as a new combination. Based on my species concepts, L. cruciata shows a relatively restricted geographical distribution, occurring from the Osa Peninsula of southwestern Costa Rica (fig. 4) south to Veraguas, Panama. On the other hand, the species with which L. cruciata is essentially identical in size and wing pattern, L. annulata, is broadly distributed across northern South America, south to Bolivia and Brazil. The life history of Lyces cruciata was discovered by independent research groups in Costa Rica and Panama (table 6). The caterpillars feed on Passiflora menispermifolia.

According to Prout (1918), the type locality for cruciata is in Panama. The label data on the type and in the original description (Butler, 1875) give the locality as “Veragua”. The state of Veraguas, Panama, would be closest to this name. Butler also gives “Salvin” as the collector. Salvin collected extensively throughout Costa Rica and Panama.

Butler's description of cruciata (1875: 340) is misleading. He described the body as “black above, head and thorax spotted with orange, metathorax orange”. However, the prominent, orange transverse band straddles the base of A1 (pl. 31), not the metathorax. On the head, Lp1, Lp2, the front, and the occiput are orange. Overall, there appears to be a great deal of variation in the amount of orange scaling on the body of this species, including tremendous variation in the width of the longitudinal FW and HW stripes.

Distribution

Costa Rica (AMNH, CAS, INBio, USNM); Panama (AMNH, BMNH, CAS, CMNH, FNHM, LAC, MUSNM, NMW, ZMH).

Dissected

♂, Costa Rica, Osa Peninsula, Corcovado Park, Sirena Station 995, 15 Jan 1989, leg. C.N. Duckett, reared on Passiflora menispermifolia, AMNH (genitalia slide no. JSM-552); ♂, Panama, Barro Colorado, Canal Zone, 9 Feb 1936, leg. Gertsch, Lutz, Wood, AMNH (genitalia slide no. JSM-1709); ♂, Panama, Chiriquí, BMNH (genitalia slide no. JSM-308); ♀, Panama, Barro Colorado, Canal Zone, 11 Mar 1936, leg. F.E. Lutz, AMNH (genitalia slide no. JSM-1710).

Lyces fluonia (Druce), new combination

Plate 30

Scea fluonia Druce, 1885b: 525, pl. 32, fig. 12.

Type Locality

Ecuador, Chiguinda.

Type

Syntype ♂, leg. C. Buckely (BMNH).

Discussion

The oblique yellow FW marking of L. fluonia is distinctive. Although this species has historically been considered a close relative of L. banana because of roughly similar wing patterns (Hering, 1925), numerous male genital differences suggest that the two belong in separate species groups within Lyces. In an earlier paper using a relatively small set of josiines, but employing larval as well as adult morphology (Miller, 1996), I suggested that fluonia belongs in a clade with cruciata, gopala, and striata. Here, all these taxa are placed together in the Patula Group of Lyces. A species-level analysis for Lyces would be required to more precisely determine the position of L. fluonia. Such an analysis might pinpoint the evolutionary transition that gave rise to its anomalous wing pattern.

Rex Friesen (USDA) reared L. fluonia on Passiflora mollissima in Ecuador (table 6). Even though this species is not particularly well represented in museum collections, it is apparently quite common in the habitats where it occurs (R. Friesen, personal commun.). The AMNH holdings contain a recently collected series (8 ♂♂, 7 ♀♀).

Distribution

Ecuador (AMNH, BMNH, CMNH, USNM, ZMH); Peru (BMNH, MUSM, OUMNH).

Dissected

♂, Ecuador, Carchi, Hwy between Mira-El Angel, 8300 ft, 8 Jun 1992, leg, R. D. Friesen, ex larva on Passiflora mollissima, AMNH (genitalia slide no. JSM-714); ♂, Peru, Amazonas, Chachapoyas, 1889, leg. M. de Mathan, BMNH (genitalia slide no. JSM-309); ♀, Ecuador, Ibarra, May 1897, leg. Rosenberg, BMNH (genitalia slide no. JSM-310).

Lyces fornax Druce, revised status

Plates 30, 39K

Lyces fornax Druce, 1885b: 525, pl. 32, fig. 11.

Type Locality

Ecuador, Intaj.

Type

Syntype ♂, leg. Buckley (BMNH).

Ephialtias aperta Warren, 1905: 313. New synonymy.

Type Locality

Peru, Cuzco, Cajon.

Type

Holotype ♂, leg. Garlepp, Oct 1900 (BMNH).

Ephialtias latimargo Warren, 1904: 16.