A Taxonomic Review of Attevidae (Lepidoptera: Yponomeutoidea) from China with Descriptions of Two New Species and a Revised Identity of the Ailanthus Webworm Moth, Atteva fabriciella, from the Asian Tropics

This review describes four species of Atteva (Lepidoptera: Yponomeutoidea: Attevidae) from China, including two new species: A. wallengreni n. sp. and A. yanguifella n. sp. The taxonomic identity of the Ailanthus webworm moth from South and Southeast Asia is revised with a designation of neotype for Phalaena (Tinea) fabriciella Swederus. Lectotypes of Atteva brucea Moore and A. niviguttella Walker are designated. Atteva brucea is synonymized with A. fabriciella. Synonymy of Atteva niviguttella and A. fabriciella is reconfirmed. The previous Chinese records of A. fabriciella were based on confusions with A. wallengreni n. sp. Confirmed specimens of A. fabriciella from China are reported. A pair of confused species, A. fabriciella and A. wallengreni n. sp., are distinguished by the number of white dots on the forewings and the genital features. Another confused pair, A. niveigutta and A. yanguifella n. sp., are compared by external and genital features. All type specimens of the described species are illustrated and compared with the conspecific specimens from various countries of the Asian tropics. Keys to all the species from China are provided.


Introduction
The Atteva group (Lepidoptera: Yponomeutoidea) is a distinctive lineage of yponomeutoids with unique morphological features. It was once regarded as a subfamily of Yponomeutidae sensu Kyrki (1990), but currently it forms its own family, namely Attevidae (Heppner 1982;van Nieukerken et al. 2011). Kyrki (1984) suggested possible apomorphies for this group: i) the presence of chaetosema on head; ii) the reduced hindleg tibia and tarsus, particularly in males without tibial spurs; iii) larvae with two subventral setae on meso-and metathorax; and iv) the concealed labial palpi in the pupae. Dugdale et al. (1998) added the tegumen of male genitalia bearing an unscaled prong as another probable autapomorphy of the family. Attevids are also distinguished from other yponomeutoid groups by a combination of the characteristics, including their scape without pectens; 3-segmented maxillary palpi; lack of pterostigma; abdominal terga without spines; phallus without basal 'scape'; W-shaped gnathos; and larvae with cranial seta V1 short and P1 not below Af2-P2 line (Dugdale et al. 1998). They often have conspicuous coloration, which is supposedly involved in a warning signal and one of the antipredator adaptations. The oligophagous larvae, if known, feed on limited numbers of plant families, such as Fabaceae, Simaroubaceae, Lauraceae, and Araliaceae, and live in communal webs loosely spun with host leaves (Dugdale et al. 1998).
Atteva, the only genus of Attevidae, comprises 53 species that are distributed in the Pantropics (Heppner 2009). High species diversity of Atteva occurs in the Indo-Australian region (ca. 35 spp.). Contrasting to such high diversity in the region, only three species of Atteva, i.e., A. brucea Moore, A. fabriciella (Swederus), and A. niveigutta Walker, have been known from China, which encompasses a vast tropical zone (Hua 2005). The low number of species currently known suggests that concentrated inventories may reveal more new species of this group from China.
The aims of this paper are to review all the known Chinese species of Atteva, to rectify erroneous records, to verify synonymies, and to describe two new species. The identity of the Ailanthus webworm moths, Atteva fabriciella, from Asian tropics, including South China, is revised and clarified, and a neotype is designated. Species comparisons and keys to distinguish all four species of Atteva from China are provided.

Material and Methods
Pinned specimens from five institutional collections were examined. Besides materials from China, specimens from adjacent countries were considered for possible geographic variations.
Abbreviations for the specimen depositories are: CAUB -Department of Entomology, China Agricultural University, Beijing, China. IZCAS -Institute of Zoology, Chinese Academy of Sciences, Beijing, China. BMNH -Natural History Museum (formerly British Museum of Natural History), London, UK. OMNH -Oxford University Museum of Natural History, Oxford, UK. USNM -United States National Museum of Natural History, Washington DC, USA.
Selected specimens were dissected following Clarke (1941) to examine genitalia and abdominal structures, except that chlorazol black was used for staining. Dissected genitalia were mounted on a slide glass, using Euparal resin (BioQuip Inc., www.bioquip.com). Slide-mounted dissections were examined under Leica MZ APO stereoscope or Leica LETTZ-DMRX microscope (www.leicamicrosystems.com).
The genitalia slide numbers were given the suffix 'IOZ' for the specimens from IZCAS, 'USNM' from USNM, and 'BMNH' from BMNH. Unmounted genitalia and abdomens were stored in a glycerin-filled, transparent envelope attached to the corresponding pinned specimen. Terms for genitalia follow Klots (1970).

Diagnoses
This species is similar to other species of Atteva from Asian tropics that have orangecolored forewings with white markings, especially A. impariguttata Zeller and A. balanota Meyrick. The white patches are denser in A. fabriciella than in A. impariguttata. This species is also different from A. balanota in the shape of the white markings on forewings: rounds or rhomboids in A. fabriciella, longitudinal bars in A. balanota. Atteva fabriciella is often indistinguishable from A. wallengreni, a new species described here, in external appearance. See diagnoses of A. wallengreni for the differences between those two species.

Redescription
Head. Vertex and frons white; in some specimens (especially from China), a triangular, black marking on vertex. Antennae filiform, 2/3 as long as forewing costa; scape white; first 4-5 flagellomeres entirely silvery white, the resting flagellomeres gray in basal half, white in distal half. Labial palpi slender, ascending at basal 1/3, white, obtuse apically; first segment 1/2 as long as second or third segment.

Types
Phalaena (Tinea) fabriciella. Neotype: female, "Neotype" [a white, rectangular label with red borders and letters], "SOUTH INDIA | COIMBATORE | 6. XII. 12. | FLETCHER. COLL.", "231", "Genitalia slide ♀ | By J.C. Sohn | USNM 96367" [a green label]. See below for details of this designation. Of them, only two female specimens are still extant in BMNH. In the original description, collecting data of the lectotype were given as "Ceylon [= Sri Lanka]. Presented by Dr. Templeton". Walsingham and Durrant (1900) suggested that a specimen from Ceylon here designated as lectotype is actually A. impariguttata Zeller, 1877. This turned out to be incorrect because JCS's type examination showed that this specimen is clearly distinct from A. impariguttata.  Walsingham and Durrant (1900) suggested that this syntype could be a distinct species. This specimen no longer exists in the type series of C. niviguttella. It is very likely that the specimen was confused with A. niphocosma Turner 1903or A. megalastra Meyrick 1907. In fact, A. fabriciella (or its synonym, A. brucea) has never been recorded again from Australia (Edwards 1996) since then.

Host plant
Simaroubaceae. Ailanthus excelsa Roxb. (Fletcher 1914;Mishra and Pandey 1966); Ailanthus triphysa (Dennst.) Alston. (Verma 1992); Brucea sumatrana Roxb. (Horsfield's record cited in Moore [1859]); Quassia indica (Gaertn.) Nooteboom (Mohanadas and Verma 1984). Host records of A. fabriciella from Boswellia serrata Triana et Planch. (Burseraceae) and Santalum album L. (Santalaceae) by Beeson (1941) and Browne (1968) need further verification. Robinson et al. (1994) mentioned that some specimens of this species from India in the BMNH collection have rearing records from Acacia (Leguminosae). This seems to be due to misidentification of host plants, since no additional records of the larvae feeding on Acacia have been made to date. Swederus (1787) described Phalaena (Tinea) fabriciella, currently A. fabriciella, from India. The original description of the species was very short and did not include any illustration. This inadequate description can apparently be applied to more than one valid species, making the true identity of Phalaena (Tinea) fabriciella dubious. Wallengren (1861) redescribed Phalaena (Tinea) fabriciella with misspelling in specific name, i.e. fabricella [sic]. Wallengren's redescription included an illustration of fabriciella, based on a specimen from China, not from the original type locality, India. This has caused substantial confusion about the identity of A. fabriciella and obscured the actual distribution and host range of the species.

Neotype designation of Phalaena (Tinea) fabriciella
Berg (1880) treated the specimen described by Wallengren as Phalaena (Tinea) fabriciella and stated the type locality of the species as China. Probably from this, Wallengren's specimen has been erroneously regarded as the type specimen of Phalaena (Tinea) fabriciella from the database of the Swedish Museum of Natural History (www2.nrm.se/en/lep_nrm/f/phalaena_fabrici ella.html). Berg's view was also followed by Robinson et al. (1994), who provided a photo of A. fabriciella that was the same as what Wallengren (1861) illustrated. In a different study, Walsingham and Durrant (1900) suggested that Swederus and Wallengren described two different species of Atteva, followed by Meyrick (1914). The type specimen of Phalaena (Tinea) fabriciella has not been found or traced yet. Therefore, it is hard to prove which of those two views is right. Swederus (1787) stated that the specimen was from Mr. J. Lee among the several collectors. Considering a letter sent by F. W. Hope to J. O. Westwood (see Smith 1986: 11 and 133), the Lee collection had most likely been housed in the OUMN. The first author (JCS) searched for this specimen at the museum and also at other major European collections, but could not find it. In fact, none of lepidopteran type specimens described by Swederus have been found (see Poole 1989). It is very likely that the type specimens of fabriciella have been lost. To make the identity of Phalaena (Tinea) fabriciella clear, designation of a neotype is necessary.
The name, A. fabriciella, also known as the Ailanthus webworm moth from India, has appeared in the voluminous forestry literature (e.g., Beeson 1941;Browne 1968;Mathur et al. 1970;Singh and Sharma 1979;Mohanadas and Verma 1984;Verma 1992;Nair 2007). Especially, Fletcher (1914: 462) and Mishra and Pandey (1966: 465) illustrated A. fabriciella. Those illustrations showed that the species is apparently different from Wallengren's fabriciella and conspecific with A. brucea. Examination of Atteva specimens from various regions of the Asian tropics revealed that Wallengren's fabriciella does not occur in India. This contradicts Berg (1880). Swederus (1787) stated that Phalaena (Tinea) fabriciella has more than 30 white dots on the forewings. Wallengren's fabriciella (= Atteva wallengreni n. sp.) frequently has more than 40 dots (Figure 17). A. fabriciella illustrated by Fletcher (1914) and Mishra and Pandey (1966) have 35 dots and 32 dots respectively. This indicates that the specimens of A. fabriciella stated from the Indian forestry literatures better fit with the original description of A. fabriciella. To clarify the identity of Phalaena (Tinea) fabriciella, it is proposed to designate as neotype a female specimen from Coimbatore, India. This specimen is consistent with the original description in the locality and characteristics as recommended by the ICZN (Article 75.3.5 and 75.3.6). This designation also meets the Article 75.6 of the Code, conserving the prevailing usage of Phalaena (Tinea) fabriciella in the voluminous Indian literature.

Synonymic notes
A. brucea is here proposed to be a junior synonym of A. fabriciella according to designation of a neotype for Phalaena (Tinea) fabriciella. The female genitalia of A. niviguttella from the type series was examined, and it was found to be identical with A. brucea (= A. fabriciella), as already suggested by Walsingham and Durrant (1900). Becker (2009) stated that A. niveigutta is a synonym of A. fabriciella. This seems to be due to a confusion of the two similar names, niveigutta and niviguttella. Atteva niveigutta and A. fabriciella are two separate species, and are easily distinguished by external appearance.

Faunistic notes
A. fabriciella was first recorded in China by Wallengren (1861), followed by Meyrick (1914), Fletcher (1929) and Wu (1938). This record, however, turned out to be due to a misidentification. Wallengren's (1861) specimen of fabriciella is in fact a distinct species from A. fabriciella. Therefore, the specimens reported from our study are actually the first record of A. fabriciella from China.

Diagnoses
This new species has long been confused as A. fabriciella. Both species are so similar that they cannot be securely separated by external appearance alone. A. wallengreni usually have more white dots on the forewings (frequently > 40) than A. fabriciella (rarely > 40). When fused, those dots appear transverse-elongate in A. wallengreni, whereas rhomboids or larger dots in A. fabriciella. For reliable identification, examination of genitalia is necessary for both species as it reveals unambiguous differences: a comb-like spinose area on socii and a pair of processes on the top of uncus larger and more slender respectively in A. wallengreni than in A. fabriciella; costa of valva without protrusion in A. wallengreni; corpus bursae with no apparent pleats in A. wallengreni; signum broader and denticules on signum smaller in A. wallengreni than in A. fabriciella.

Etymology
This species is named after Hans Daniel Johan Wallengren (1823Wallengren ( -1894, the Swedish entomologist who described this species under the name Amblothridia fabricella [sic].

Diagnoses
This species is similar to Atteva aleatrix Meyrick, described from Fiji Island, but differs from the latter by having denser white dots on the forewings. It is also easily distinguished from A. fabriciella and A. wallengreni n. sp. by the labial palpi and the legs, which are mostly dark brown (white in the latter two).

Description
Head. Vertex white with a triangular, black marking; frons black, tinged with white laterally. Antennae filiform, 1/2 as long as forewing costa; scape white in basal half, dark brown in distal half; flagellum dark brown, first 3-4 flagellomeres with lustrous dark gray scale-covering dorsally on basal half. Labial palpi slender, obtuse apically, ascending after basal 1/4; first segment white, 1/2 as long as second or third segment; second and third segments dark brown, sparsely intermixed with white.

Diagnoses
This new species is very similar to A. niveigutta, but is distinguished from the latter by having rather reduced white dots on the forewings. The two species also differ in male and female genitalia (Table. 1).

Description
Head. Vertex dark brown, surrounded with white marginally; frons white. Antennae filiform, 1/2 as long as forewing; scape white on basal half, dark brown on distal half; flagellomeres dark gray. Labial palpi slender, upcurved; first segment as long as antennal scape, white; second segment 2× longer than first; third segment as long as second; second and third segment dark brown, sparsely intermixed with white.
Thorax and abdomen. Patagium orange, tinged with white posteriorly; tegula orange, with white marking posteroventrally; mesonotum orange, edged with white along scutoscutellar suture. Foreleg with coxa dark brown, intermixed with orange on lower surface, white on upper surface; femur dark brown dorsally, white ventrally; tibia dark brown, with white marking on distal 1/3; tarsi dark brownish-gray. Midleg with coxa orange, tinged with orange-white terminally; femur dark brown ventrally orange-white dorsally; tibia dark brownish-gray, with a white band on the middle and the distal end; tarsi dark brownish-gray; first tarsomere with a white ring on the distal end. Hindleg with coxa orange, mottled with white; femur dark brown, tinged to orange distally; tibia with dense hair tufts orange dorsally, yellowish-white ventrally; tarsi very slender, with gradual change from orange on base to white-orange to dark gray on distal end. Forewing 13.5 to 17.5mm (average 15.3mm, n = 3), elongate, with obtuse apex and straight termen, orange to orange-brown; 25 to 35 (average 30, n = 3) white markings scattered throughout the forewing. Hindwing as narrow as forewing, with broadly round termen, orange. Abdomen orange, with white marking on distal end of each sternite. (Figure 33-37). Uncus rectangular, bifid, each process 1/2 as long as socius; socius narrower to apex, long-hairy, with a comb-like spinose zone near to apex; Tegumen as long as socius; gnathos starting from broad arm and then curved perpendicularly at 1/3 and 2/3 respectively, with a densely ciliate, digitate medial plate. Valva elongate, elliptical, 2.5× longer than socius, densely setose except on basal area; costa slightly emarginated at basal 1/3 and convex at middle; sacculus lobate, setose at base. Saccus slender, 1/3 as long as valva. Phallus slightly bent at basal 1/4; a spinulate zone of cornuti 2/5 as long as phallus, with 4 to 5 of stout spines terminally. (Figure 47-49). Lamellae postvaginales as a pair of shallow, semicircular, setose lobes. Apophyses posteriores as long as anteriores. Ostium bursae surrounded by semicircular emargination. Ductus bursae bowl-shaped near to ostium bursae, with antrum at anterior 1/4; antrum cylindrical, with membraneous slit on a side. Corpus bursae oval, 1.4× longer than ductus bursae, with protruding cervical area; signum elongate, inverted pentagonal, narrowly less sclerotized medially, with denticules denser to margin.

Etymology
The species epithet is derived from 'Yanguifei,' one of the four legendary beauties in Chinese history, and refers to the beautiful coloration of the new species.