A peculiar new species of gall-inducing, clearwing moth (Lepidoptera, Sesiidae) associated with Cayaponia in the Atlantic Forest

Abstract Larvae of most clearwing moths (Lepidoptera, Sesiidae) are endophagous borers of many angiosperms, including their fruits, stems, and roots. Their localized feeding may lead to swellings on those plant parts, but whether the structures produced should be considered true galls is still controversial. In this study we describe a peculiar sesiid moth, Neospheciacecidogenasp. nov. whose larvae induce unusual, external galls on Cayponiapilosa (Vell.) Cogn. (Cucurbitaceae) in the Atlantic Forest of southernmost Brazil. The adults, egg, larva, pupa and the gall are described and illustrated based on light and scanning electron microscopy. Galls are cylindrical and unilocular; they are induced individually on axillary buds of the C.pilosa stem. Unlike larvae of other sesiids, those of N.cecidogenasp. nov. lack abdominal pseudopodia, and show reduced stemata and chaetotaxy. Pupation occurs inside the gall, after having overwintered in the last larval instar. A maximum likelihood tree constructed based on mitochondrial DNA (COI) sequences showed that N.cecidogenasp. nov. is monophyletic and has an average distance of 13% to species of Melittia. The genera Neosphecia Le Cerf, 1916 stat. rev., Premelittia Le Cerf, 1916 stat rev., and Melittina Le Cerf, 1917 stat. rev. are restored from synonyms of Melittia Hübner, 1819 [“1816”].


Introduction
The Sesiidae is a worldwide family with ca 160 valid genera and 1,452 species. Of these, 268 species are found in the Neotropical region Kallies 2004, 2017) and 119 have been recorded in Brazil (Casagrande et al. 2018). These diurnal micromoths are well known by their morphological and behavioral modifications that have been associated with Batesian mimicry, particularly regarding bees and wasps (Duckworth and Eichlin 1974;Skowron et al. 2015). In the last decades, many of their species have been found by attracting the adults with the use of synthetic sex pheromones (Duckworth and Eichlin 1977b;Brown and Mizell 1993;Špatenka et al. 1999;Arita and Gorbunov 2000;Arita et al. 2009;Skowron et al. 2015;Gorbunov 2018).
In contrast to the adults, however, their immature stages are poorly known, especially in the tropical regions where the host plants for only a small proportion species have been documented (Heppner and Duckworth 1981;Eichlin 2003). The family is known to feed upon a wide variety of shrubs, trees, vines, and herbaceous plants (Heppner and Duckworth 1981;Brown and Mizell 1993;Špatenka et al. 1999). Most sesiid larvae are endophagous borers, and many are highly specific regarding host-plant use. They may be associated with various plant organs, including fruit (Harms and Aiello 1995;Lopes et al. 2003;Puchi 2005), trunks/stems (Bruch 1941;Eichlin and Passoa 1983;Brown and Mizell 1993;Eichlin 1995Eichlin , 2003, and roots (e.g. Lima 1945;Gorbunov 2015Gorbunov , 2019. Their feeding may lead occasionally to localized swellings on plants, but whether these structures should be considered true galls has been controversial (Eichlin 1995;Hanson et al. 2014). As far we know, none has been associated with the induction of external galls yet.
This study concerns a new species of clearwing moth that induces conspicuous, external galls on the stem of a vine, Cayaponia pilosa (Vell.) Cogn. (Cucurbitaceae) in the Atlantic Forest of southernmost Brazil. A preliminary comparison of genitalia structures suggested that it belongs to a genus closely related to Melittia Hübner, 1819 ["1816"], but does not conform to any species within it. We describe and illustrate the gall, the immature stages and adults of this moth under both light and scanning electron microscopy, and provide information on its natural history. By conducting a phylogenetic analysis of mitochondrial DNA (COI) sequences using representative members of Melittini (Melittia), we provide further support for the proposition of the new taxon, as well as its genus validation.
After careful examination of the habitus we came to the conclusion that this species is congeneric with Neosphecia combusta Le Cerf, 1916, which is the type species of the monotypic genus Neosphecia Le Cerf, 1916. This genus was formally synonymized with Melittia by Duckworth and Eichlin (1977a), but without concrete arguments. Also in this work, by the same reason, we formally restore this and two other names from synonymy with Melittia Hübner, 1819 ["1816"]: Premelittia Le Cerf, 1916, stat. rev., Neosphecia Le Cerf, 1916, stat. rev., and Melittina Le Cerf, 1917.

Material and methods
All specimens used in this study were either dissected or reared from galls in small plastic vials, which were maintained under controlled conditions (14 h light/10 h dark; 25 ± 2 °C) in the Laboratório de Morfologia e Comportamento de Insetos (LMCI), Departamento de Zoologia, Universidade Federal do Rio Grande do Sul (UFRGS). Galls were collected from C. pilosa during the autumn seasons of 2014-2018 in the Centro de Pesquisas e Conservação da Natureza (CPCN Pró-Mata/PUCRS; 29°28'36"S, 50°10'01"W), 900 m elevation, São Francisco de Paula Municipality, Rio Grande do Sul State, Brazil. Pupae were obtained later (September) by dissecting some galls under a stereomicroscope in the laboratory. Adults were pin-mounted and dried. Chorionated eggs were obtained by dissection of females used in genitalia preparations. Immature stages were fixed in Dietrich's fluid and preserved in 75% ethanol. Additional larvae used for DNA extraction were preserved in 100% ethanol at −20 °C.
For descriptions of genital morphology, larvae were cleared in a 10% potassium hydroxide (KOH) solution, stained with either Eosin or Chlorazol black E and slidemounted in either glycerin jelly or Canada balsam. Last instar larvae were prepared similarly for description of chaetotaxy. Observations were performed with the aid of a Leica M125 stereomicroscope. Structures selected to be drawn were previously photographed with a Sony Cyber-shot DSC-H10 digital camera attached to the stereomicroscope. Vectorized line drawings were then made with the software Corel Photo-Paint X7, using the corresponding digitalized images as a guide. Additional specimens were used for scanning electron microscope analyses. They were dehydrated in a Bal-tec CPD030 critical-point dryer, mounted with double-sided tape on metal stubs, coated with gold in a Bal-tec SCD050 sputter coater and examined and photographed in a JEOL JSM6060 scanning electron microscope at the Centro de Microscopia Eletrônica (CME) of UFRGS.

Molecular analysis
DNA was extracted from four larvae (specimens LMCI 263-33A, B, C, and D) of the new taxon using the PureLink Genomic DNA extraction kit (Invitrogen). Extracted DNA was resuspended in 50 mL Tris: EDTA (10 mm Tris-HCl, 1 mm EDTA, pH 5 8.0). Mitochondrial DNA PCR was conducted using primers LCO1490 and HCO2198 (Folmer et al. 1994), which amplified approximately 650 bp of the cytochrome oxidase I gene (COI), a fragment entitled DNA barcode. PCR reactions were conducted using 2 µL of the extracted DNA. The thermal cycler profile for CoI consisted of 35 cycles of 94 °C for 45 s, 48 °C for 45 s and 72 °C for 45 s. Excess dNTPs and primers were removed and the amplified DNA concentrated using exonuclease I and FastAP thermosensitive alkaline phosphatase (Thermo Fisher Scientific, Waltham, USA). Samples were sequenced using BigDye Terminator v. 3.1 Cycle Sequencing kit (Thermo Fisher Scientific) and analyzed in an ABI3730XL automatic sequencer. The new data were deposited in the GenBank and BOLD Systems (http:// www.boldsystems.org/) under the project MISA (Table 1). The COI sequences were initially aligned using Clustal W (Thompson et al. 1994) and subsequently refined by eye using CodonCode Aligner (CodonCode Corp, Massachusetts, USA).
The final dataset included 12 specimens: four individuals originally sequenced and six species of Melittia, the only representative of Melittini available, chosen based on morphologically established relationships with the new taxon. In addition, two species of Synanthedonini were used as outgroup to Melittini according to relationships proposed based on previous molecular phylogenies (e.g. McKern et al. 2008;Hansen et al. 2012;Lait and Hebert 2018).
The specific status of the new taxon was tested through a COI tree relying on monophyly. Tree reconstruction was based on the maximum likelihood (ML) algorithm performed in PHYML v. 3.0 (Guindon et al. 2010), using 1,000 replicates of heuristic search with random addition of sequences and TBR branch swapping. The substitution model GTR+I+G was selected based on the Akaike Information Criterion run in MEGA v. 7 (Kumar et al. 2016). Monophyly confidence limits were assessed with the bootstrap method at a 50% cutoff after 1000 iterations. Sequence divergences were quantified using the Kimura 2-parameter model in MEGA v. 7.

Museum collections
Abbreviations of the institutions from which specimens were examined are:

Molecular phylogeny
Sequencing of COI resulted in an average amplicon size of 633 bp. The aligned data matrix including all genera resulted in 668 characters. Of these, 189 (28%) were phylogenetically informative. Maximum likelihood analysis recovered an optimal ML tree = 5897 with nucleotide frequencies of A = 29.1%, C = 16.5%, G = 14.8%, and T = 39%. The four LMCI 263-33 specimens grouped in the COI tree as a single lineage, which was placed close to the genus Melittia (Fig. 1). Melittia grandis was the closest taxon (89% similarity) retrieved from online blast tools of Genbank and BOLD. Pairwise K2P distances estimated from the new taxon and Melittia ranged from 13% to 16% (M. grandis and M. calabaza, respectively) ( Table 2). Divergence to the outgroup varied from 18% to 22%.   Meyrick, 1931), but it can be differed by the practically undeveloped hair-like tuft of the hind leg (vs hair-like tuft of the hind leg well-developed in Agriomelissa) and undeveloped anal lobe of the hindwing (vs well-developed in the genus compared). Melittina can be easily distinguished from Premelittia and Neosphecia by the well-developed proboscis (proboscis undeveloped in both these compared genera) and shape of the antenna (antenna strongly clavate, short, slightly shorter than a half of forewing in Melittina, vs antenna fusiform, slightly longer than middle of forewing in Premelittia and Neosphecia). From all other genera in Melittiini, including species of New World so-called "Melittia" (Arita and Gorbunov 1996), Melittina differs by the shape of the antenna (antenna clavate with a hook apically in compared Melittiini) and by the undeveloped anal lobe of the hindwing (anal lobe of the hindwing well developed in all genera compared Melitiini).
Composition. Presently this genus contains only a single species: Melittina nigra Le Cerf, 1916, stat. rev. Remarks. The genus Melittina was described by Le Cerf (1917) based on a single female specimen of the type species, Melittina nigra Le Cerf, 1917. Fortunately, the original description is rather complete and supplemented with a fairly accurate figure of the type species, Melittina nigra (Le Cerf 1917: 239;pl. 477, fig. 3933). They contain important characters (venation of the hindwing), which clearly put this genus in the tribe Melittiini. As we have mentioned above, this genus was synonymized with Melittia by Duckworth and Eichlin (1977a) without concrete arguments. After a careful reexamination of the original description (Le Cerf 1917: 239) and the figure of the type species (Le Cerf 1917: pl. 477, fig. 3933), we conclude that Melittina is a distinct genus and restore it from synonymy with Melittia. Besides this, we also transfer Melittia nigra (Le Cerf, 1917)  Diagnosis (after Le Cerf 1916: pl. 375, fig. 3136, 1917. Head with antenna fusiform without a hook distally, slightly longer than half forewing; vertex smooth scaled; proboscis undeveloped. Legs with hind tibia and tarsus smoothly scaled. Forewing with transparent areas well-developed; veins R 4 and R 5 short stalked; hindwing transparent, anal lobe very small. Abdomen smoothly scaled with anal tuft poorly developed. Differential diagnosis. Without any doubt, by the venation of the hindwing this genus belongs to the tribe Melittiini. By the undeveloped proboscis and fusiform antenna this genus seems to be closely related to the genus Neosphecia Le Cerf, 1916, but it can be separated from it by the smoothly scaled vertex (vs vertex covered with hair-like scales in the compared genus) and venation of the forewing (veins R 4 and R 5 short stalked in Premelittia vs veins R 4 and R 5 separate basally in Neosphecia). It is easily distinguished from all other Melittiini, including species of New World so-called "Melittia" (Arita and Gorbunov 1996), by the smoothly scaled vertex (vs vertex with hair-like scales in all genera of Melittiini), form of the antenna (antenna clavate with a hooked apex in all other Melittiini), absence of the proboscis (proboscis well developed and functional in all other Melittiini), and form of the hind leg (with a tuft of hair-like scales in all other Melittiini).
Composition. Presently this genus contains the single species: Premelittia rufescens Le Cerf, 1916. Remarks. The generic name Premelittia was first introduced in the legends of figures as the binominal "Premelittia pufescens nov. sp." (Le Cerf 1916: 9). The original description of this genus was published a year later (Le Cerf 1917: 234). Unfortunately, it was based on a single female specimen of the type species, Premelittia pufescens Le Cerf, 1916, but it is rather complete and contains important characters, which clearly separate this genus from the genus Melittia. This genus was also synonymized with Melittia by Duckworth and Eichlin (1977a) without concrete arguments. After a careful re-examination of the description (Le Cerf 1917: 234) and the figure of the type species (Le Cerf 1916: pl. 375, fig. 3136) we conclude that Premelittia is a distinct genus of the tribe Melittiini and restore it from synonymy with Melittia. Besides this, we also transfer Melittia rufescens (Le Cerf, 1916)  Diagnosis (after Le Cerf 1916: pl. 375, fig. 3137, 1917. Head relatively broad; antenna fusiform without a hook distally, slightly longer than half forewing; vertex covered with hair-like scales; proboscis undeveloped. Legs with hind tibia and tarsus smoothly scaled. Forewing with transparent areas undeveloped or very small; veins R 4 and R 5 separate basally; hindwing transparent, anal lobe undeveloped or extremely small. Abdomen smooth scaled with anal tuft poorly developed. Differential diagnosis. Like the two previous taxa, on the basis of the venation of the hindwing this genus belongs to the tribe Melittiini. By the fusiform antenna and undeveloped proboscis this genus is closely related to the genus Premelittia Le Cerf, 1916, stat. rev., but it can be distinguished the forewing venation (veins R 4 and R 5 shortly stalked in Premelittia vs veins R 4 and R 5 separate basally in Neosphecia). Neosphecia differs from Melittina by the undeveloped proboscis (well developed and functional in the compared genus), shape of the antenna (strongly clavate and short, not reaching the middle of the forewing in Melittina vs fusiform, long, slightly longer than half forewing in Neosphecia), and the venation of the forewing (veins R 4 and R 5 long stalked in Melittina vs veins R 4 and R 5 separate basally in Neosphecia). From all other genera of the Melittiini, including New World so-called "Melittia" (Arita and Gorbunov 1996), it differs by the shape of the antenna (clavate with a hook apically in all other Melittiini vs fusiform in Neosphecia), completely undeveloped proboscis (vs well developed in all other Melittiini), and venation of the forewing (veins R 4 and R 5 separate basally in Neosphecia vs veins R 4 and R 5 stalked in all genera of Melittiini).
Composition. Currently, we include two species in this genus: Neosphecia combusta Le Cerf, 1916 (type species) and N. cecidogena sp. nov.
Remarks. Like the genus Premelittia (see above), the genus name Neosphecia was firstly introduced in the legends of the figures as the binomen "Neosphecia combusta nov. sp." (Le Cerf 1916: 9). The original description of this genus was published a year later (Le Cerf 1917: 236). Unfortunately, it was based on a single female specimen of the type species, Neosphecia combusta Le Cerf, 1916, but it is rather complete and contains important characters, which clearly separate this genus from Melittia. This genus was also synonymized with Melittia by Duckworth and Eichlin (1977a) without concrete arguments, as with the two previous genera. After careful re-examination of the description (Le Cerf 1917: 236) and the figure of the type species (Le Cerf 1916: pl. 375, fig. 3137) we conclude that Neosphecia is a distinct genus of the tribe Melittiini and restore it from synonymy with Melittia. Besides this, we also transfer Mellittia combusta (Le Cerf, 1916) back to its original combination, Neosphecia combusta Le Cerf, 1916. Description. Male (holotype) ( Fig. 2A-D). Alar expanse 23.1 mm; body length 10.8 mm; forewing 10.5 mm; antenna 5.8 mm.

Neosphecia cecidogena
Head with antenna dark brown to black dorso-externally and yellow ventro-externally; scapus yellow and narrowly lined with dark brown dorsally; frons entirely dark brown with purple-blue sheen; vertex black with dark-blue sheen and an admixture of individual white and yellow hair-like scales; proboscis completely undeveloped; labial palpus dark brown to black with an admixture of yellow scales dorsally and white, long, hair-like scales ventrally in distal half; occipital black with a few white scales dorsally.
Thorax with patagia dark brown to black with a small, yellow, transverse spot anterior-ventrally; tegula dark brown to black with yellow, hair-like scales distally; mesothorax entirely dark brown to black; metathorax dark brown to black with two tufts of yellow, hair-like scales laterally; thorax laterally dark gray-brown with bronze-violet sheen. Legs with neck plate dark brown to black; fore coxa dark brown to black with a narrow, yellow exterior margin; hind tibia dark yellow with an admixture of black elongated scales on basal half; spurs dark yellow with golden sheen and a few black scales exterior-basally; hind tarsus dark yellow with a dense admixture of elongated black scales dorso-externally. Forewing: dorsally dark brown with darkviolet sheen and an admixture of individual yellow-orange scales, more dense distally and at anal margin; transparent areas present but very small: anterior and posterior ones very narrow, external one divided into two very short cells; cilia dark brown to black with dark violet-purple sheen. Hindwing transparent; veins broadly covered with dark brown and a few yellow-orange scales; discal spot undeveloped; outer margin about as broad as cilia, dark yellow and narrowly dark brown distally; cilia dark brown with dark violet sheen.
Abdomen including anal tuft black with dark blue sheen and a few yellow scales at base of second tergite.
Female (paratype) (Fig. 2E-H). Antenna with more broad yellow stripe ventroexternally; vertex with more numerous white hair-like scales; labial palpus with more numerous yellow scales dorsally; patagia with more yellow scales anteriorly; legs with more numerous yellow scales; both tergites 4 and 5 with a sparse, dark yellow stripe medially. Color patterns otherwise as in male.
Individual variability. The type series practically invariable in individual size and in the coloration of various parts of the body and wings.  Etymology. The species name, an adjective, is derived from a composition between the Portuguese "Cecidia" (a gall) and the suffix gena (derived from the Latin verb "gigno", be born). Thus, the epithet refers to the cecidogenous habit of the new described clearwing moth.
Description of immature stages. Eggs (Fig. 4): light brown, obovoid, with the anterior end slightly flattened; maximum length (average ± standard deviation) = 0.05 ± 0.01 mm, maximum width = 0.39 ± 0.01 mm, n = 6. Surface of chorion with faint carenae, delimiting irregular, mostly hexagonal cells and minutely pitted, forming a continuous meshwork-like plastron (sensu Hinton 1981), except for the anterior end where corresponding holes are sparse. Micropylar area on anterior pole, consisting of a subtrapezoidal indentation in the center, which is surrounded by a rosette of about 20 subpentagonal cells that increase in size centrifugally.
Pupa (Figs 7, 8). Body length (average ± standard deviation) = 11.52 ± 0.67 mm; n = 5. Yellowish brown, becoming dark brown near adult emergence (Fig. 7C). Head with stout, short, bow-shaped frontal gall-cutter process in dorsal view (Figs 7A, C), which is continued latero-caudally up to eye margin by slightly elevated ridges that limit depressions on frons under lateral view. Vertex with two pairs of small setae laterally. Clypeus little pronounced, with one pair of small setae laterally; labrum short, slightly bilobed (Fig. 8B). Antennae clubbed at the end, reaching anterior portion of third abdominal segment. Mandibles small, rounded, meso-anterior to the eyes. Maxillary palpi small, rounded, latero-posteriorly to the eyes. Proboscis shorter than and laterally margined by the prothoracic legs; galea converging mesally along the second half portion. Labial palpi contiguous on the center, extending to half length of the galea. Pronotum fairly developed, bearing a central ridge that extends caudally along the meso-and metanotum. Hindwings concealed by forewings, both extending to sixth abdominal segment. Protho-, meso-, and methatoracic legs reaching the second, fifth, and seventh abdominal segments, respectively. Thoracic and abdominal setae extremely reduced in size: one pair, latero-dorsally, on meso-and metathorax, and A2-A9; another pair, subspiracular, on A2-A7. Abdominal spiracles rounded, with slightly elevated peritreme (Fig. 8E), laterally on A2-A7; spiracle on A8 closed. Basal and caudal transverse rows of spines (Figs. 8D, F) present from abdominal segments two to seven on males; only one row of such spines is found on segment seven of females, and also on segments eight and nine on both sexes. Last abdominal segment with four pairs of stout, scaly spines on caudal margin: two pairs in latero-dorsal and two pairs in latero-ventral position (Fig. 8H, G).
Distribution. This new species is known only from the type locality, the humid forest portions of the CPCN Pró-Mata, São Francisco de Paula municipality, Rio Grande do Sul State, Brazil.
Host plant. Galls of N. cecidogena have been found only in association with Cayaponia pilosa (Vell.) Cogn. (Cucurbitaceae), which is distributed in the ombrophilous Atlantic Forest of southern Brazil (from Minas Gerais to Rio Grande do Sul State) (Gomes-Klein et al. 2015). Biology and natural history of this cucurbit are poorly known. It is a herbaceous, prehensile vine (Fig. 9A), which bears pairs of forked, axillary tendrils, simple, alternate leaves with lamina that may vary from entire, to three to five lobed; flowers are solitary, axillary, and with penduncles varying from 7 to 9 cm long; fruits are ellipsoid and ca 2 cm in length, which are initially green (Fig. 9B) but changing to wine-colored when mature (Porto 1974;Villagra and Romaniuc Neto 2011). At the type locality, C. pilosa plants are found scattered on forest borders, particularly along trails. Natural history. The unilocular, cylindrical galls of N. cecidogena measure on average (± standard deviation) 3.44 ± 2.51 cm (n = 9) in length when mature. They appear individually and from the beginning develop externally on axillary buds of C. pilosa vines. Contrary to the oval C. pilosa fruits (Fig. 9B), N. cecidogena galls are not pedunculate (Fig. 9C). They are green during development and later turn dark brown with the progress of senescence (Fig. 9I). The internal chamber is filled with a yellowish nutritive tissue (Fig. 9D, E) which is consumed by larvae during development. With the end of feeding, the last larval instar builds a blackish, circular operculum ( Fig. 9F-H) that splits the chamber into two sections, one distal, where the frass is deposited, and one basal, which has the distal portion of the wall lined with light-gray silk (Fig. 9E) and where pupation occurs. Achieving maturation, the wall of the gall hardens with the exception of the distal, pointed end, which remains thin and soft and through which adult emergence occurs (Fig. 9I). During emergence, with the action of the frontal process and body contortions, the pupa detaches the operculum and ruptures the distal, weaker portion of the wall. By continuing these movements and anchoring the body laterally with its abdominal spines, the pupa pushes itself partially out of the gall. During this process, the anterior portion of the exuviae is split, allowing adult emergence. In all cases of adult emergence under laboratory conditions, the anterior part of the pupal exuviae (head and thorax) was found protruding to the outside (Fig.  9J), while the posterior third remained in the chamber. A few C. pilosa plants have been found at the type locality bearing from one to five N. cecidogena galls per plant. Field collections carried out during five consecutive years at the type locality indicated that it is a univoltine species, larvae growing during the summer when young galls are seen on C. pilosa vines. Fully developed galls containing last instar larvae have been collected mainly during autumn. When brought to the laboratory, these remained larvae during the winter, apparently in a diapause state. Pupation in this case occurred during the first week of September and emergence a few days later during early spring. The absence of a proboscis suggests that adults of N. cecidogena are not active feeders. The appearance of a substantial number of corionated eggs in the abdomen of dissected females shortly after emergence in the laboratory indicates that reproduction occurs early in adult life, and thus, adults may not live long.

Discussion
A few molecular phylogenies have been proposed for Sesiidae, particularly including Melittini (McKern et al. 2008;Hans et al. 2012;Lait and Hebert 2018), which present a controversial relationship within Sesiinae, placed in the basal position. Melittia is the main representative of the tribe so far sequenced, and in this study presented high internal divergence (6-11%),which suggests that distinct lineages have been lumped into this genus. This seems to hold true particularly for the "New World" lineages that were separated from Melittia based on genitalia (Arita and Gorbunov 1996). Molecular data presented here gave further support in the sense that the genus Neosphecia is closely related to Melittia, from which COI sequences of N. cecidogena diverged ca 13%. Corresponding divergences between the genera Synanthedon and Carmenta were 18 and 22%, respectively. Unfortunately, DNA sequences are not available for the genera Premelittia and Melittina, which precludes further comparison within the Neotropical Melitiini, whose phylogenetic relationships are still controversial, as already mentioned.
Premelittia and Melittina are poorly known monotypic genera, with data available restricted to the original taxonomic description, and observations on morphology made a posteriori based also on the adult female holotypes only by Arita and Gorbunov (1996). According to these authors, the Nearctic and Neotropical Melittini as a whole are in need of revision, and thus, the taxonomic positioning of these groups within the family is still pending. Results presented in the present study demonstrate that variation among the immature stages of Sesiidae, especially the larva and its feeding habits, is broader than previously thought and should be taken into account in future phylogenetic studies involving these lineages. We demonstrate clearly that not only internal damage related to mobile (borer) activity is to be found among clearwing moths, but also typical, external galls containing differentiated nutritive tissues as induced by N. cecidogena.
Sesiid eggs have been described as ventrally flat, being laid with the longitudinal axis parallel to the substrate and having the lower surface less ornamented than the upper one (Duckworth and Eichlin 1974;Eichlin and Passoa 1983;Eichlin 1995;Puchi 2005). This may not the case in N. cecidogena, a situation that should be better examined under natural conditions. Eggs dissected in this study were transversally nearly elliptical and uniformly covered by a continuous plastron, with the exception of the anterior pole where the micropyles are located, as already mentioned. We could not identify differentiated aeropyles on the chorion surface of N. cecidogena eggs, which are supposedly located underneath such a plastron.
Unlike larvae of other sesiids that have pseudopodia bearing crochets (Heppner and Duckworth 1981;Heppner 1987), those of N. cecidogena lack such structures. It is still uncertain whether this reduction is unique to Neosphecia species. It may represent an evolutionary loss, associated with the stationary way of life within a gall. This hypothesis should be tested further by comparing this species with other representatives including those having similar cecidogenous habit (sessile, confined life style) where corresponding convergences to N. cecidogena is expected, and to different life habits like borers (more mobile life style), where pseudopodia bearing crochets are expected to be found. This could be also the case of setae that are reduced in size and in number, and which are tentatively named here. For example, P and Md groups are well developed in the head of other sesiids (e.g. Eichlin and Passoa 1983) but absent in N. cecidogena, except for P1. Heppner and Duckworth (1981) and Heppner (1987) stated that the L group is trisetose on the protorax of sesiids in general, which is not the case in N. cecidogena where this group is bisetose, also on abdominal segments A1-8. Furthermore, we could not find any trace of SV2 setae on N. cecidogena, except for the prothoracic segment; the SV3 and MV setae were not found.
Pupae of N. cecidogena showed the general integument morphology usually found in the family (Patočka and Turčani 2005). Pads of galea are well differentiated externally on N. cecidogena pupa, but contrary to other buccal appendages that are filled with the corresponding adult buccal appendages, they are empty later in pupal development, thus suggesting that the proboscis is in fact being lost during that ontogenetic stage in this species. Additional differences exist in the pupal stage of N. cecidogena in relation to the absence of a differentiated frontal process on the head, which is well developed and bow-shaped, for example, in Melittia gilberti Eichlin (Eichlin 1992). It also slightly differs regarding the shape, size, and arrangement of A10 spines, compared for example to those described for Carmenta species (Eichlin and Passoa 1983;Puchi 2005). Most sesiids pupate internally, within the gallery opened by larval feeding (e.g. Bruch 1941;Duckworth and Eichlin 1974;Eichlin 1995;Harms and Aiello 1995;Newland and Sawyer 2014). In this case, the larva may either tunnel to the plant surface prior to pupation, leaving a thin exterior covering through which emergence occurs later (Duckworth and Eichlin 1974), or build a cocoon within the feeding site that is surrounded by the frass (e.g. Lopes et al. 2003). However, other species leave the feeding site later in development, weaving a silk cocoon in the soil within which pupation occurs (e.g. Heppner and Duckworth 1981;Gorbunov 2015). In N. cecidogena pupation happens inside the gall, which is used as the pupal chamber after being closed distally with a silk-made operculum and deprived of frasses, and thus without tunneling or building a typical cocoon. Variation in this life history trait among other sesiid lineages should be also evaluated further from a phylogenetic perspective. Finally, the galls described here appeared figured twice in association with Cayaponia spp. in a survey carried out at the type locality by Toma and Mendonça (2013). The one in early development stage that was illustrated by them (green color; Toma and Mendonça 2013: fig. 1P) had the induction attributed to an unidentified species of Coleoptera, which resulted from misidentification. The one that appears in a senescent stage (brown color; Toma and Mendonça 2013: fig. 1Q) had the induction correctly associated by these authors to an unidentified Lepidoptera. Similar galls are also found on C. pilosa plants in Parque Nacional do Itatiaia Nacional, Rio de Janeiro, as illustrated by Maia and Mascarenhas (2017: fig. 139). However, it is unlikely in this case that they are conspecific to N. cecidogena as the galls are differently shaped; they are pedunculate and present five conspicuous filiform projections distally. Bruch (1941) showed the feeding damage made by larvae of Melittia tayuyana Bruch, 1941, which appears externally as swellings on the stems of Cayaponia ficifolia (Vell.) Cogn. [= C. bonariensis (Mill.) Mart. Crov.] in Argentina, a plant that is sympatric to C. pilosa (Gomes-Klein et al. 2015). Possible tuberculate galls that are induced by an unidentified species of Sincara Walker, 1856 on the roots of Wilbrandria verticillata (Vell.) Cogn. (Cucurbitaceae) were illustrated by Lima (1945), also for the Atlantic Forest in Brazil. Recently, Gorbunov (2015) suggested the existence of a gregarious feeding habit on galls induced by Melittia ambo Gorbunov, 2015 at the root collar of Citrullus colocynthis (L.) Schrad. (Cucurbitaceae) in Ethiopia. These few records suggest cucurbit galls are associated with more than one lineage of sesiid moths, and may vary in general shape, host plant species, plant parts where they are induced, and the density of larvae they bear inside.