On the identities of Neotropical Stegana species (Diptera, Drosophilidae). I. Redescription of Stegana magnifica Hendel, 1913 and Stegana fumipennis (Enderlein, 1922)

C T Two conspicuous Steganinae species, Stegana (Steganina) magnifica Hendel, 1913 from Amazonian Peru and Stegana (Ceratostylus) fumipennis (Enderlein, 1922) from southern Brazil, are redescribed based on holotypes, and their identities are clarified. Both species are exclusive to the Neotropical Region and the first, with a body length of about 5.5 mm, is the largest species of Stegana described so far in this region, while the latter displays a peculiar antenna bearing an unusual, forward-projected, comma-shaped flagellomere 1. The photomicrographs of the habitus and terminalia of each specimen are


Introduction
The genus Stegana Meigen, 1830 is easily recognized, and included species share several general and distinct characteristics (Bächli et al., 2004). However, external morphological features are highly variable with species overlapping and with determination of many species based on females remaining doubtful. Accurate identification normally depends on characters of the male terminalia. Thus, there is currently a lack of information about most of the 39 known species of Stegana described from the New World (Bächli, 2019;Pirani and Grimaldi, 2019). Consequently, we suspect that the identifications of numerous species in various collections are questionable.
Worldwide, there are presently five accepted subgenera within the genus Stegana and all but one of them (Oxyphortica Duda, 1923) have been recorded from the New World. The status and relationships of these subgenera remain open for interpretation (Pirani and Grimaldi, 2019, p. 453). While we treat for the moment the species we intend to study within their current subgeneric classification, we are convinced that the subgeneric division of Stegana needs revision and better characterization.
Chronologically, Stegana coleoptrata (Scopoli, 1763) and Stegana furta (Linnaeus, 1767) are the oldest Stegana species recorded in North America. Both of these described species were from Europe and their identities were questioned for a long time, even in Europe, and the latter has also been mentioned under the name Stegana curvipennis (Fallén, 1823). Since some authors have treated these two species as synonyms, it is almost certain that the older records of Stegana coleoptrata (Scopoli, 1763) and Stegana furta (Linnaeus, 1767) are inaccurate. S. furta has already been removed as a New World species.
Later, a third species, Stegana pallipes Wiedemann, 1830, was described from Brazil and again, the identity of this species is questioned because the type material is apparently lost and the short description does not unequivocally distinguish it from congeners. Towards the end of the 19 th Century, two more species were described from the Caribbean Islands of Saint Vincent and the Grenadines by Williston (1896): Stegana horae and Stegana tarsalis. Shortly afterwards, Stegana vittata (Coquillett, 1901) was described from the United States, which was referenced by Sturtevant (1921) as the species with straight wings, since the "standard" Stegana species have wings that bend downwards. Sturtevant treated S. vittata as a synonym of S. coleoptrata because he was unable to separate specimens of these species. Additional species were then described by Hendel (1913), Enderlein (1922), Malloch (1924aMalloch ( , 1924b, Duda (1925Duda ( , 1927 and Curran (1934). Although some of these species were sometimes assigned to different genera, this view was not always accepted by later authors.
Among the papers cited above, the largest contribution to the knowledge of New World Stegana species comes from Malloch (1924aMalloch ( , 1924b) who reported, described or redescribed 20 species, mostly collected by Pablo Schild in Costa Rica (collection currently in the USNM). Of these 20 species, 16 were new to science. Malloch also included four previously described species S. acutangula (Hendel, 1913), S. magnifica Hendel, 1913, S. curvipennis (Fallén, 1823 and S. tarsalis Williston, 1896. Malloch also compared S. horae Williston, 1896 with his S. tempifera. The six Nearctic species of Stegana were previously revised in detail (Laštovka & Máca, 1982); however, of the 33 Neotropical congeners only Stegana penicillata (Kertész, 1901), has recently been redescribed (Pirani and Grimaldi, 2019;Bächli and Vilela, 2020b). Moreover, based on their original descriptions and/or limited available keys, it has been virtually impossible to identify the remaining species.
To partially address these problems, we have analyzed, redescribed, photomicrographed, and dissected the holotypes of two conspicuous species, Stegana fumipennis (Enderlein, 1922) from Southern Brazil (state of Santa Catarina) and Stegana magnifica Hendel, 1913 from Amazonian Peru (probably near the border between the Cusco and Ucayali Regions). This is the first of a series of papers that are intended to clarify the identities of Neotropical species of Stegana. In this study we describe a cost-effective method for producing high-quality photomicrographs of the habitus and male terminalia, which are essential to identify these species.

Material and methods
The redescriptions are based on two holotypes: a male of Stegana magnifica, collected on 12.X.1902 by C. Schnuse and O. Garlepp from Amazonian Peru and described by Hendel (1913); and a female of Stegana fumipennis, collected in southern Brazil by H. Fruhstorfer and described by Enderlein (1922). The specimens were loaned to us from the Senckenberg Collection, Dresden, Germany (formerly known as the Staatliches Museum für Tierkunde Dresden), and from the Museum für Naturkunde, Leibniz-Institut, Berlin (formerly known as the Zoologisches Museum Berlin), respectively.
Label data attached to each type specimen are cited in full with a slash [/] indicating a line change and a double slash [//], indicating a label change. Our notes and/or interpretations are included within brackets throughout the text.
The habitus of the two holotypes was photomicrographed with the rear camera of a Samsung Galaxy S8 smartphone, which was attached to the right eyepiece of a Wild M3 stereomicroscope through a magnetic plate of a clothespin-shaped plastic adapter as shown in Figs. 1a-h and as described previously (Vilela and Prieto, 2018;Vilela and Bächli, 2019;Bächli and Vilela, 2020a). Objective lenses of 1.6×, and 4× were used, and the camera was set to default or optically zoomed to Figure 1 Plastic clothespin-shaped adapter used for connecting the smartphone to the stereomicroscope eyepiece, several views: a) swivel magnetic plate seen from below (blue rectangle), b) a movable metal plate placed over the magnetic plate seen from above, c) swivel magnetic plate (metal plate removed) with four circles in the center of velvet cover, oblique view, d) adapter attached to the stereomicroscope with the magnetic rectangle plate partially placed over the right eyepiece but not touching the left eyepiece, posterior view, e) idem, oblique view, f) idem, anterior view of the uncovered right eyepiece lens to where the rear camera should be positioned, g) a movable metal plate placed inside the smartphone case, h) smartphone obliquely positioned over the right eyepiece as seen by the observer. 2×, 3×, or 4× with the autofocus mode disabled. A series of pictures were taken at consecutively increasing depths of focus for each selected view.
The image series were mostly stacked using the "All Methods" algorithm of the open-source software CombineZP (Hadley, 2010) to create an in focus composite. In the case of the photomicrographs of the terminalia preparations, the "all methods" algorithm did not always provide the best results and is likely due to the transparency of this anatomical structure. Better results were obtained by opening the output folder, which was automatically created by the software, and comparing the quality of the images obtained using a specific stacking method. During this process, one can eliminate individual files with poor quality and only restack the best files from the 6 methods.
Microscope slide preparations were done according to Wheeler and Kambysellis (1966) and Kaneshiro (1969). The abdominal sclerites, including the disarticulated male terminalia, are preserved in glass microvials filled with glycerin and attached to the stopper with the pin holding the respective specimen. Further details can also be found in Vilela and Bächli (2000) and Bächli et al. (2004).
The composite photomicrograph images taken with either the stereomicroscope or compound microscope were edited with Adobe Photoshop software. Unless otherwise indicated, all of the photomicrographs on the same plate were taken and enlarged to the same magnification.
For morphological terminology, measurements and indices see Vilela and Bächli (1990, 2000)-and Bächli et al. (2004. We refer to the overview of Brake and Bächli (2008) and Bächli (2019) for the subgenera and included species and we also direct the reader to Pirani and Grimaldi (2019).

Results and discussion
Diagnosis (modified from Bächli et al., 2004 andZhang et al., 2012). Head setae generally thick and long, but postocellar setae thinner and in some species minute or absent. Interfrontals may be numerous and large. Tibia of middle leg basally with a posterodorsal row of 3-5 erect setae, followed downwards by a row of shorter, less erect setae (see Figs Zhang et al., 2012). Wing (see Figs. 1C-E in Zhang et al., 2012) in most species conspicuously bent down at rest, generally dark brown, particularly along the Costa, but almost transparent in basal 1/4 and in cell m 4 , where (outside posterior crossvein) a roundish, fully transparent "window" may appear; vein R 2+3 usually almost straight, at most slightly curved upwards just before terminating; vein R 4+5 straight in apical half; vein M softly and more or less equally curved upwards towards vein R 4+5 ; costal section III with about 8 warts on the lower side; costal section IV weak and relatively short, acrocostal index large (ac index about 8-17 in the available descriptions).
Diagnosis. Eye longish, with almost horizontal main axis, red, medially bearing two parallel dark brown horizontal stripes, divided by a narrow paler band (Figs. 7a,c,d). Large body length (ca. 5.5 mm), distance between fore and middle legs unusually long. Wing with particular pattern: anterior margin brown, getting darker towards wing tip, extended backwards over vein R 4+5 ; both crossveins brown, vein M with a narrow brown band along its whole length, a pale (transparent) area in cell r 4+5 containing two isolated dark spots; pale (transparent) areas are also visible in cells m 3 and m 4 . Surstylus roundish, protruded, deeply concave, bearing one peg-shaped prensiseta medioventrally; outer margin square-lobed medially. Dorsal arch anteromedially densely scaled. Handle of ejaculatory apodeme twice as long as aedeagal apodeme. Ventrodistal margin of aedeagus conspicuously ornate with a fringe of five large, irregular, somewhat rectangular-shaped strips (with the median one being the shortest), resembling an upside-down strip curtain.     Material examined. Holotype male (Fig. 2)  white, index about 4.50. Eye longish, main axis nearly horizontal, red, medially bearing two parallel dark brown horizontal stripes which are divided by a narrow paler band (Figs. 7a, c, d), index about 1.34. Occiput brownish-yellow in the upper half, white below, with a dark brown band in between. Pedicel yellow. Flagellomere 1 longish, yellow, becoming brown towards the tip. Arista (Figs. 7c, d) with 11 long dorsal, 7 long ventral and about 12 distinct inner branches, plus a terminal fork. Proboscis, clypeus and palpus yellow. Thorax (Fig. 3) generally yellowish, prolonged: the distance between fore and middle legs is unusually long. Length about 3.20 mm.
Scutum yellowish (Fig. 3b), with 6 brown longitudinal stripes: the inner pair do not reach the anterior border, becoming broader towards the scutellum, a shorter pair laterally behind the suture, and a pair above the postpronotum alongside the margin of mesonotum, 16 rows of acrostichal setae. h index about 0.66, dc index about 0.47. Prescutellars strong, on each side with two short setae between prescutellars and posterior dorsocentrals. Scutellum (Figs. 3b, 5b) slightly pointed, brownish with a narrow, central whitish-yellow stripe which becomes broader towards the tip, the distance between apical scutellars is about 70% of that of the apical to basal distance; basal ones divergent; scut index about 1.44. Pleura (Figs. 3a, 4a) yellowish, with two dark brown longitudinal stripes, katepisternum brown below the lower stripe. Sterno index about 0.93. Halter yellow. Legs generally brownish, fe2 yellow on basal ¼, ti1 and ti2 apically paler yellowish, ti3 yellowish on basal half. Short preapical setae on all tibiae, short ventral apical setae on the middle tibia. Middle tibia with a distinct row of posterodorsal erect setae, becoming shorter downwards. Middle tarsal joints slightly broadened.
Wing (Figs. 5b, 6) with a particular brown pattern not seen in other Stegana species: anterior margin brown, getting darker towards the tip of the wing, extended backwards over vein R 4+5 ; both crossveins brown, crossvein m-cu somewhat sigmoid, vein M with a narrow brown band along its whole length, a pale (transparent) area in cell r 4+5 containing two isolated dark spots; pale ( Abdomen (Fig. 6b) dark brown, without a pattern but paler basally. Sternites 5 and 6 ( Fig. 9) wider than long, somewhat round-cornered rectangles, posterior margin straight and shallowly concave, respectively.
Comments. Hendel (1913, p. 631) wrote "1 female". The specimen we have analyzed, however, is a male, an observation that Hendel could have overlooked. To our knowledge, no other specimens are extant. It should also be stressed that the general characters of S. magnifica are very different from those of the other species included in the subgenus Steganina, where it has been considered. Notably, this species exhibits an almost horizontal main axis of the eyes and should be, in our opinion, placed in a different subgenus. Stegana (Ceratostylus) Enderlein, 1922 Type species: Ceratostylus fumipennis Enderlein, 1922:297 (original designation). Ceratostylus Enderlein, 1922: 296 Stegana (Ceratostylus) (Enderlein, 1922): Wheeler, 1981: 30 (affiliation, as a monotypic subgenus of Stegana); Brake and Bächli, 2008: 293 (affiliation).
Comments. The genus name Ceratostylus refers to the abnormal, comma-shaped flagellomere 1 of the antenna. Having checked the type specimen with the left flagellomere 1 largely missing, we were unable to conclusively determine whether this is the result of a developmental malformation or an artifact of shrinking.
However, based on our knowledge, no other Stegana species exhibit a comparable shrinking effect.

Distribution.
Brazil. Santa Catarina: Águas Mornas municipality, rural district of Teresópolis (known from the type locality only). Comments.
This species shows all of the generally accepted characteristics of Stegana species. At present, based mainly on the conspicuous shape of flagellomere 1 of the antenna, we have not decided whether placement of this specimen in a separate subgenus is justified.