Notes on the Genus Aceratoneuromyia Girault (Hymenoptera: Eulophidae)

Simple Summary Fruit flies in the family Tephritidae are economically important pests of edible fruits, with some known hymenopteran parasitoids. Although Aceratoneuromyia indica (Silvestri) is one of the most common parasitoids and has been used in biological control of fruit flies, its establishment in biocontrol is impeded by the difficulties of proper identity. The genus Aceratoneuromyia Girault (Hymenoptera: Eulophidae) is studied here, and the generic name Trjapitzinichus Kostjukov and Kosheleva is proposed as a new synonym under Aceratoneuromyia as well. Based on morphology and DNA barcodes, A. bilinis Huangfu and Cao sp. nov., A. carinata Cao and Zhu sp. nov. and A. trilinus Cao and Zhu sp. nov. are described and illustrated. The well-known parasitoid of fruit flies, A. indica, is treated here with the diagnosis and illustrations. In addition, this study provided a morphological diagnosis for Aceratoneuromyia as well as a key to world species of this genus. We also briefly discussed the relationship of Aceratoneuromyia with other possibly close groups based on available morphological data. Together with DNA barcodes of A. bilinis sp. nov. and A. indica generated here, this study provided essential and useful information for the species identity of Aceratoneuromyia. Abstract Fruit flies in the family Tephritidae are well known as economically important pests of edible fruits and can often cause serious damage and losses to both agriculture and the economy. One of the common parasitoids of fruit flies, Aceratoneuromyia indica (Silvestri), has been used in biological programs. However, the biocontrol utilities of parasitoids are impeded by the difficulties of proper identification. Species of the genus Aceratoneuromyia Girault (Hymenoptera: Eulophidae), usually developed as parasitoids of fruit flies, are studied here. Trjapitzinichus Kostjukov and Kosheleva is proposed as a new synonym under Aceratoneuromyia. Three new species of Aceratoneuromyia, A. bilinis Huangfu and Cao sp. nov., A. carinata Cao and Zhu sp. nov., and A. trilinus Cao and Zhu sp. nov., are described and illustrated from China. Aceratoneuromyia indica is also treated here with diagnosis and illustrations. DNA barcodes of A. bilinis and A. indica and a key to the world species of Aceratoneuromyia are provided. This study provided important identification information of parasitoids with morphology and molecular evidence, which is useful for imperative needs regarding the identity of parasitoids attacking fruit flies.


Introduction
Many members of dipteran flies are of great economic importance, and some of them are serious pests in agriculture. Among these pests, some species of the family multibracteatus Levl. et Vant. (Rosaceae), which was transplanted from the collection locality in Yunnan. Larvae and pupae of Bactrocera dorsalis were lab-reared and originally collected from field in Fujian, China. The host pupae were placed in plastic cups with a mesh cloth cover and moved to the laboratory of the Institute of Zoology, Chinese Academy of Sciences (IZCAS) to rear adults of fruit flies and parasitoid wasps. Emerged wasps were then preserved in 100% ethanol for further use after emergence.

Specimens Preparations
Specimens were examined with a Nikon SMZ 1500 stereomicroscope fitted with a 10 mm ocular grid having 100 divisions. Some ethanol-preserved specimens of parasitoid wasps were critical-point dried with a Leica EM CPD300 automated critical point dryer for morphological studies. Several critical-point dried specimens were dissected into head, mesosoma, and metasoma for scanning electron microscopy (SEM), which were subsequently sputter-coated with gold using a Leica EM SCD050 super cool sputter coater. Micrographs were produced using an FEI Quanta 450 environmental scanning electron microscope. Habitus pictures were taken with a Nikon D7000 digital camera connected to the Nikon SMZ 1500 stereomicroscope and then stacked in Helicon Focus software to generate single highly focused images. All images were processed with Adobe Photoshop CC 2019.

Results
During recent programs on the biology of Tephritidae in China, fresh materials of two species of Aceratoneuromyia were respectively collected and reared from the pupae of Bactrocera dorsalis (Hendel) and Pelmatops ichneumoneus (Westwood) in the family Tephritidae. The species reared from Bactrocera dorsali was identified as A. indica, and the other species reared from Pelmatops ichneumoneus was identified as a species new to science. By comparing with specimens deposited in the insect collection of IZCAS, two additional new species of Aceratoneuromyia were discovered as well. As a result, three new species were treated and illustrated in this study.
Fifteen COI sequences were successfully sequenced from 3 females of Aceratoneuromyia bilinis, and 6 females plus 6 males of A. indica. A COI matrix with a length of 657 base pairs was obtained based on these 15 COI sequences after alignment and trimming, which has no insertion or deletion. A NJ tree ( Figure 1) by K2P distances was generated based on this COI alignment.

Morphological Diagnosis and Species Treatments of Aceratoneuromyia
The genus Aceratoneuromyia Girault, 1917Aceratoneuromyia Girault, 1917 Diagnosis. Antenna with very long apical seta of terminal spine (about 2× as long as terminal spine, sometimes broken off) and with long, scattered erect or semi-erect setae (e.g., Figures 3d and 9b) as well as male antenna, with each funicular having a short transverse row of long setae on the dorsal surface (e.g., Figures 5e and 9h); female antenna with funiculars usually quadrate to transverse (e.g., Figure 9h), rarely distinctly longer than broad; male antennal scape usually with (e.g., Figure 3d) and rarely without (e.g., A. evanescens) ventral plaque. Malar sulcus present, from weak to groove-like. Frontofacial sutures (e.g., Figures 3b and 9a,g) not or only slightly divergent dorsally, sometimes absent. Scutellum with (e.g., Figure 9a) or without (e.g., Figure 9d,i) submedian lines, if present, then weak and incomplete, with sublateral lines distinct and laterally carinate (e.g., Figure 9a); scutellum with anterior pair of setae situated in the posterior half and close to the posterior pair of setae (e.g., Figure 9a,d,i). Gaster usually slender, rarely rounded, with ovipositor sheaths not visible even ventrally (e.g., Figures 4e and 9f,k); spiracles of penultimate segment of gaster rather large (relatively smaller in males), usually visible in dorsal view, rarely directed laterally and not visible in dorsal view (e.g., Figures 4f and 9e,j); female gaster usually tending to be convex dorsally (e.g., Figures 2a and 8a). Hypopygium extending distinctly more than half length of the gaster, nearly to the apex (e.g., Figures 4e and 9f,k). Remarks. The diagnosis of Aceratoneuromyia was listed here by referring to the three newly described species, descriptions of the other known species, as well as the diagnosis given by Bouček [19], Graham [7], LaSalle [8], and Ikeda [20]. Ikeda [20] has proposed a modified concept for Aceratoneuromyia with the addition of a new species from Japan and South Korea, A. kamijoi Ikeda, which disagrees in some characters with the concept previously given. This study follows those modifications to interpret the genus Aceratoneuromyia. Based on this genus concept of Aceratoneuromyia and re-descriptions of Trjapitzinichus by Kostjukov & Kosheleva [11], sometimes the placement of some intermediated species into one of them could be confusing. Furthermore, the separation of Trjapitzinichus from Aceratoneuromyia by Kostjukov & Kosheleva [10,11] obscures the differences between these two genera. It is hereby considered unreasonable to keep them as separate genera and Trjapitzinichus is synonymized with Aceratoneuromyia here.  (Figures 3a and 5a). Scutellum with somewhat distinct but incomplete submedian lines (Figures 3a and 5a). Dorsellum large, about 2× as broad as high. Metasoma slender (Figure 2a  Antenna (Figure 3d) with 3 funiculars and 3 clavomeres, setae of pedicel and flagellum standing out strongly; scape not reaching anterior ocellus, about 2.5× as long as pedicel; funicle proximally distinctly stouter than pedicel; pedicel distinctly longer than F1 (0.7:0.5); F1-F3 gradually decreasing in length (0.5:0.4:0.3) and increasing in width, F1 wider apically, length equal to apical width, F2 and especially F3 transverse; clava as long as F2 plus F3, clavomeres decreasing in length, C3 with a short but distinct terminal spine that has a distinctly long seta (about 2× length of terminal spine). Each flagellomere with longitudinal sensilla and apically with a circle of scattered, mushroom-shaped capitate peg sensilla; each flagellomere except C3 truncate apically.  Head (Figure 3b,c) slightly broader than mesoscutum in dorsal view (4.0:3.5), about 3× as broad as long (4.0:1.3); vertex with long and erect setae, and a vertex suture at each side. OOL as long as POL (0.70:0.70). Ocelli arranged in a strong obtuse triangle. Head in frontal view about 1.14× as broad as high (4.0:3.5). Face with extremely fine delicately engraved reticulation, much weaker than that of thorax. Frons with a short and narrow Y-shaped frontofacial suture, with V-shaped diverging lines narrowly separated from each other, extending to vertex dorsally and sometimes connecting vertex suture, and with extremely short vertical line reaching scrobes (Figure 3b,c). Toruli inserted slightly above lower margin of eyes. Eyes with few of short and white hairs, diameter larger than malar space Scutellum flattened in profile, slightly broader than long (2.5:2.1), sculptured rather more finely than mesoscutum; submedian lines present, distinctly nearer to sublateral lines than to each other, enclosing a space about 2.0-2.2× as long as broad; however, submedian lines are often incomplete, superficially indicated basally and apically, with two pairs of setae subequal in length, anterior pair situated well behind the middle and close to posterior pair, and posterior pair situated near posterior margin; scutellum with a sloping frenum ( Figure 4a). Dorsellum about 1.86× as broad as high (1.3:0.7), about 0.54× length of propodeum (0.7:1.3), faintly engraved reticulate along outer margins (sculptures invisible under stereomicroscope), almost smooth; slightly incised in the middle of posterior margin; lateral panel of metanotum with more or less longitudinal but irregular carinae in the anterior 3/5 and smooth in the posterior 2/5 ( Figure 4a). Propodeum raised-reticulate, with reticulation much coarser than that of scutellum ( Figure 4a); with a strong medina carina having a short sulcus anteriorly, median carina distinctly raised, broadening caudad and then extending laterally; with rim of spiracle invisible, covered by the lobe of callus, while with a hole at the same position of spiracle and a broad sulcus from anterior margin to the spiracle position, with weak paraspiracular carina that is difficult to discern under certain lights of stereomicroscope; callus with raised reticulation as coarse as that of median area of propodeum, with 1 erect seta only. Lateral panel of pronotum weakly reticulate, prepectus, mesepimeron and mesepisternum nearly smooth; acropleuron smooth; metapleuron with raised reticulation like propodeum ( Figure 4b).
Petiole distinct, nearly as long as broad, but partially hidden under certain angles (Figures 2a and 4d). Gaster (Figure 4d,e) slender, about 2× as long as broad and nearly as long as head plus mesosoma, based on the median length, the largest Gt 1 1.8× Gt 2 , Gt 2 not clearly delimited from Gt 1 , Gt 2 2/3 length of Gt 3 , Gt 3 as long as Gt 4 ; sclerotized, convex dorsally, almost as hard as thorax and not easily collapsed; gastral tergites with superfacial engraved reticulation, almost polished under stereomicroscope even under SEM ( Figure 4d Host. This species, a gregarious endoparasitoid, parasitizes the larva of Tephritidae host Pelmatops ichneumoneus (Westwood) and emerges from the pupa of P. ichneumoneus.
Remarks. Before synonymizing Trjapitzinichus with Aceratoneuromyia, like A. kamijoi Ikeda, A. bilinis is another intermediated species by having submedian lines although incomplete on scutellum, torulli slightly above the lower margins of eyes, and distinct malar sulcus. This confirms the necessary modifications for the concept of Aceratoneuromyia by Ikeda [20] and suggests that it is reasonable to synonymize Trjapitzinichus with Aceratoneuromyia.  (Figures 6a,b and 7c). The longest cercal seta slightly longer than the other 2 setae equal in length, and very weakly curved apically (Figure 7c).
Female (Figures 6 and 7). Body length 1.5-1.7 mm. Body dark brown without metallic tinge (Figure 6a,b), with tegula yellow, with pale setae. Antenna brown, with scape and pedicel yellowish ventrally, and flagellum paler ventrally. Legs with coxa, femora and claw brown, with remainder parts as well as trochanter, base and tip of femora yellow (Figure 6a,b). Wings hyaline, with brown veins.
Legs with coxa, femora, and tibia-engraved reticulate; hind coxae compressed with distinct and complete dorsal carina. Fore wing (Figure 7f)  Host. This species was recorded from Pegomyia phyllostachys Fan (Diptera: Anthomyiidae), possibly from the host pupa as an endoparasitoid.
Distribution. Oriental (China: Jiangsu). Remarks. Aceratoneuromyia carinata, along with the other two species, A. claridgei Graham and A. granularis Domenichini, has pronotum with a row of more than 4 long setae near posterior margin, while other species usually have a row of 4 long setae. However, A. carinata is readily differentiated from the other two species by the presence of submedian lines on the scutellum which seems unclear under SEM because old specimens were not sputter-coated with gold. Additionally, A. carinata has hind coxa with a complete dorsal carina and ventrally paler antenna. Host. This species can widely parasitize the larva of fruit flies in the family Tephritidae of Diptera [9].
Distribution. Aceratoneuromyia indica was originally described in India and now has been thought to be widely introduced as a biological control agent to all zoological regions [7][8][9]. However, currently, it is difficult to define its original distribution pattern based on occurrence records only. Female. Body length 1.8-2.2 mm. Body dark brown without metallic tinge (Figure 10a,b), with pale setae. Antenna brown, with scape and pedicel paler dorsally and brownish yellow ventrally. Legs yellow, except for brown coxa and claw (Figure 10a). Wings hyaline, with brown veins.
Antenna with 3 funiculars and 3 clavomeres (Figure 11h), pale setae of pedicel and flagellum standing out strongly; scape somewhat reaching anterior ocellus, about 3× as long as pedicel (1.5:0.5), 3.78× as long as broad (1.5:0.4); funicle as stout as pedicel; pedicel, F1 and F2 equal in length (0.6:0.6:0.6) and longer than broad (0.6:0.5), F3 subquadrate (0.5:0.5); clava slightly shorter than F2 plus F3 (1.0:1.1), C1 longer than broad (0.7:0.5) and longer than F3 (0.7:0.5), C3 short, nearly as long as terminal spine. Each flagellomere with longitudinal sensilla and apically with a circle of scattered, mushroom-shaped capitate peg sensilla; each flagellomere except C3 truncate apically.   (Figure 11c,d). Scutellum more or less flattened in profile, distinctly broader than long (2.5:2.0), sculptured rather more finely than mesoscutum; with traces of incomplete submedian lines, superficially indicated basally and apically, distinctly nearer to sublateral lines than to each other, enclosing a space about 2.3-2.5× as long as broad (2.5:1.0), as well as with a trace of median line in the anterior half; with two pairs of setae subequal in length, anterior pair situated well behind the middle and close to posterior pair, and posterior pair situated near posterior margin; scutellum with an extremely short frenum (Figure 11b,e). Dorsellum about 3× as broad as high (1.5:0.5), about 0.5× length of propodeum (0.5:1.0), faintly engraved-reticulate; slightly incised in the middle of posterior margin; lateral panel of metanotum with more or less longitudinal but irregular carinae in the anterior 4/5 and smooth in the posterior 1/5 (Figure 11c). Propodeum raised-reticulate, with reticulation slightly coarser than that of scutellum (Figure 11c,f); with a strong medina carina distinctly raised, broadening caudad and then extending laterally; with rim of spiracle visible, propodeal spiracle separated by 0.5-0.8 its diameter from metanotum; callus with raised reticulation as coarse as that of median area of propodeum, with 3 long erect setae and 1 long seta at post-lateral corner, not at the same plane with median area because of elevated median area (Figure 11f). Lateral panel of pronotum weakly reticultate, prepectus, mesepimeron and mesepisternum nearly smooth; acropleuron smooth; metapleuron with raised reticulation like propodeum.  (Figure 11c). Gaster slender, about 2.64× as long as broad (9.5:3.6) and slightly longer than head plus mesosoma (9.5:7.5) (Figure 10a), based on the median length, the largest Gt 1 1.25× Gt 2 (2.5:2.0), Gt 2 , Gt 3 and G 4 equal in length; Gt 1 smooth, the other tergites with superfacial engraved reticulation; dorsolateral posterior margin of Gt 5 circled along outer rim of distinctly large spiracle of Gt 6 , large spiracle visible and distinct in dorsal view ( Figure 11a); Gt 7 with 4 cercal setae, the longest seta weakly curved apically and about 1.5× as long as the other 3 setae that are subequal in length. Tip of hypopygium almost reaching the apex; anterior margin of hypopygium truncate, posterior margin bidentate.
Legs with coxa (Figure 11c), femora, and tibia engraved-reticulate; hind coxae compressed with complete dorsal carina. Fore wing (Figure 11g)  Hosts. Based on the label information of the above type material, two females were reared from the pupa of Dendrolimus superans (Butler) (Lepidoptera: Lasiocampidae), and one female was reared from the pupa of Stilpnotia candida Staudinger (Lepidoptera: Lymantriidae). The possibility that A. trilinus may be a hyperparasitoid of lepidopteran moths through other parasitoids is suggested by the host records of A. evanescens summarized in Noyes [9].
Distribution. Palaearctic (China: Hebei, Liaoning). Remarks. Aceratoneuromyia trilinus is uniquely differentiated from the other known species in this genus by the trace of a median line on the scutellum.

Discussion
Kostjukov and Kosheleva [10] established the genus Trjapitzinichus to include three known species originally placed in Aceratoneuromyia and only summarized some differences with Aceratoneuromyia without argumentation. Kostjukov and Kosheleva [11] provided re-descriptions for Trjapitzinichus, and they did not mention and discuss the variations in some of those differences between Aceratoneuromyia and Trjapitzinichus. However, some intermediate species (e.g., A. kamijoi and A. bilinis) show some characteristics shared by both two genera, such as distinct malar sulcus, submedian lines of scutellum, toruli above the lower margin of eyes, and lateral projections on pronotum. These variations obscure the borderline between these two genera and bring difficulties in separating them. Those three species assigned to Trjapitzinichus are possibly just aberrant members of Aceratoneuromyia despite their lateral conical tubercles on the pronotum. Therefore, the generic name Trjapitzinichus is treated as a junior synonym of Aceratoneuromyia here although generic delimitations of most Tetrastichinae genera require confirmation from molecular evidence.
The phylogenetic relationships between Aceratoneuromyia and the other Tetrastichinae genera remain unclear. Gahan [23] synonymized Aceratoneuromyia with Melittobia, which were subsequently considered to be two distinct genera by other Eulophidae taxonomists [7,8,24,25]. Nevertheless, the affinity between Aceratoneuromyia and Melittobia was not supported by courtship behavior [26] and morphology which was briefly discussed by Cao et al. [27]. Although species of both Tachinobia (one of three genera in the Melittobia complex) and Aceratoneuromyia are gregarious parasitoids of some Diptera hosts, between which have no intersection, the former mostly in Tachinidae and the latter mostly in Tephritidae [8,28]. In addition, the Melittobia complex shows notable sexual morphological dimorphism, such as in the antenna, wings and eyes (males of Koucoreckia unknown) [27], whereas Aceratoneuromyia exhibits sexual difference restricted to only the antenna (e.g., Figure 9b,h). Therefore, the similarity between Aceratoneuromyia and the Melittobia complex is likely to be due to convergence, which requires further confirmation by comprehensive molecular evidence. Currently, it is still extremely hard to discuss the systematic position of Aceratoneuromyia as little is known about the phylogeny of Tetrastichinae. In contrast, Aceratoneuromyia is hypothetically more close to Aprostocetus than to Tetrastichus, with these two genera regarded as the two main lineages of Tetrastichinae.