The male terminalia of seven American species of Drosophila ( Diptera , Drosophilidae )

The male terminalia of seven species of Drosophila endemic to the New World are described or redescribed and illustrated: one in the hydei subgroup (D. guayllabambae) and four in the mulleri subgroup (D. arizonae, D. navojoa, D. nigrodumosa, and D. sonorae) of the repleta group; one in the sticta group (D. sticta) and one so far unassigned to group (D. comosa). The D. guayllabambae terminalia redescription is based on a wild-caught fly. The redescriptions of the terminalia of the four species in the mulleri subgroup are based on strain specimens, while those of D. sticta and D. comosa terminalia are based on their holotypes. D. guayllabambae seems to be a strictly mountainous species of the Ecuadorian and Peruvian Andes. D. nigrodumosa is apparently endemic to Venezuela, occurring in the Andes as well as at lower altitudes. The remaining five occurs only at lower altitudes of the American continent. The detailed line drawings depicted in this paper aim to help interested taxonomists to tell those species apart. Their precise identification is of great importance to the knowledge of the American biodiversity increasingly threatened by human activities.


Introduction
Since early last century, the male terminalia have been the main feature used by entomologists to identify sibling species of insects, including flies of the genus Drosophila.Having this in mind, during the past forty years I have dedicated myself to clarify the identity of dozens of species of Drosophila mostly endemic to the American continents.Several line drawings of male terminalia prepared during this period have already been published (e.g.Vilela and Sene 1977, Vilela 1983, Vilela and Bächli 1990, Vilela et al. 2008).These were meant, in part, to complement those previously published by other authors (e.g.Duda, 1927, Wasserman 1962).Drawings of some species treated in the present paper were already published in the last 30 years by different authors (Rafael and Arcos 1988, Fontdevila et al. 1990, Ruiz et al. 1990, Heed and Castrezana 2008), although depictions of male sclerites in different views were offered.
In addition to specimens deposited in the Drosophila type and reference collection (DTRC) curated by Dr. Marshall R. Wheeler, all species strains belonging to the Drosophila repleta group formerly maintained at the University of Texas at Austin were examined in 1979.At that time, the author described the male terminalia of four undescribed species in his PhD thesis (Vilela 1981).However, for different reasons, they were not included in the resulting article entitled "A revision of the Drosophila repleta species group (Diptera, Drosophilidae)" by Vilela (1983).These four species are as follows: D. arizonae Ruiz, Heed & Wasserman, 1990, D. navojoa Ruiz, Heed & Wasserman, 1990, D. nigrodumosa Wasserman & Fontdevila in Fontdevila et al., 1990, and D. sonorae Heed & Castrezana, 2008.Yet in 1979, the author also examined the spot-thoraxed female holotype of Drosophila peruensis Wheeler, 1959 in the DTRC, collected in 1903 somewhere along the Urubamba River, Peru.Based on the wing pattern and the morphology of the arista, the author proposed to transfer it from the repleta species group to the guarani species group (Vilela 1983).In late February 1984, the author searched for a male specimen of Drosophila peruensis making a visit to the Urubamba Canyon at the end of the railroad track (Fig. 1) running from Cuzco to Machu Picchu (km 112).This final station is located just above the bridge crossing the Urubamba River (Reparaz 1984: 197, 201) at the beginning of the unpaved zigzag road (Hiram Bingham) going up to the Machu Picchu ruins (ca.2400 m altitude; Fig. 2A).One banana-baker's yeast-baited trap was placed on the rocks at 1 km northeast of the Machu Picchu station (13°09'S; 72°32'W) alongside the railroad (km 111, Fig. 2B).On March 2 nd seventeen specimens (6 males, 11 females) of an undescribed species were collected and later misidentified and redescribed (Vilela andPereira 1985, Vilela andBächli 1990: 78) as D. peruensis.Nine years later, the previously misidentified and redescribed Peruvian specimens were renamed by Vilela and Pereira (1993) as Drosophila urubamba, included in the guarani group, and D. peruensis was transferred to a new group of its own (Ratcov and Vilela 2007).Additionally, one Drosophila male belonging to an undescribed species of the hydei subgroup of the repleta group was collected in the same trap and date.The specimen was dissected and its terminalia illustrated.This latter species would be formally described four years later (Rafael and Arcos 1988) as Drosophila guayllabambae based on specimens sampled from an isofemale line established from a single gravid female collected at 2200 m altitude in Ecuador.
In December 1985 and February 1989, aiming to clarify the identity of some poorly known American Drosophila other than those belonging to the repleta group, the author received two type specimens from the National Museum of Natural History (NMNH), Smithsonian Institution, Washington, D.C., USA.These included a male holotype of Drosophila sticta Wheeler, 1957 (from Honduras), belonging to the sticta group, and a male holotype of Drosophila comosa Wheeler, 1968 (from Costa Rica), an ungrouped species.These were dissected and their terminalia described or redescribed and illustrated but never published.
The purpose of this paper is to clarify the identities of seven of the American Drosophila species cited above, five of them (members of repleta group) apparently associated with rotting cacti.

Material and methods
Male specimens belonging to seven American species of Drosophila were analyzed and had their terminalia illustrated.The dissected specimens, in alphabetical order, are as follows.
The nine non-type specimens were either collected by the author in Peru (the fourth one) or obtained in 1979 from stocks maintained in the National Drosophila Species Resource Center (NDSRC), originally at the Department of Zoology, University of Texas (Austin).The NDS-RC is currently named Drosophila Species Stock Center of the University of California at San Diego, which is on the way to be moved to Cornell University in New York.
All double-mounted non-type specimens cited above are deposited in the Museu de Zoologia da Universidade de São Paulo, São Paulo, SP, Brazil (MZSP).
Additionally, four double-mounted specimens (1♂, 3♀♀) sampled from each of the three Mexican stocks and from the Venezuelan strain maintained in the NDSRC were originally deposited in the DTRC (at the University of Texas, Austin) and later transferred to the AMNH (American Museum of Natural History, New York).
The analyzed and dissected type material includes the male holotypes of the poorly known Drosophila comosa and Drosophila sticta, both housed in the collections of the National Museum of Natural History, Washington, D.C., USA (NMNH).

Discussion
The morphology of the male terminalia harbors highly informative and thus widely used characters for species identification and delineation.The detailed line drawings of the morphology of seven American species depicted here aim at facilitating the identification of these species.
In turn, I discuss the implications of these line drawings, the systematics positioning of certain species and give additional comments on the species' distribution range.

Drosophila guayllabambae Rafael & Arcos, 1988
The spot-thoraxed D. guayllabambae seems to be a true mountain dweller.This Andean species was described from specimens collected by Gabriela Arcos and Violeta Rafael at 2200 m altitude, 30 km NE of Quito on the right margin of the Guayllabamba River, Guayllabamba, Province of Pichincha, Ecuador.The Guayllabamba River is a tributary of the Esmeraldas drainage system that flows into the Pacific Ocean.The male terminalia of the specimen analyzed in the present paper differ slightly from those depicted in the original description.The differences may be due to both intraspecific variation and different interpretation of the sclerites given by different authors.Thus, it was considered to belong to the same species, which extends the distribution range of D. guayllabambae to a slightly lower altitude (ca.2050 m) of the Andean western slopes of southern Peru (Region or Department of Cuzco).This Peruvian specimen was collected as a byproduct of my incessant and so far unsuccessful quest for a male of the spot-thoraxed D. peruensis (see Vilela and Pereira 1985, 1993, Vilela and Bächli 1990, Ratcov and Vilela 2007).The author collected this cited specimen of D. guayllabambae on the right margin of the Urubamba River, a tributary of the Amazon drainage system that flows into the Atlantic Ocean.A virtual field trip using Google Earth and Street View software was made in August 2017 to the collection site at kilometer marker 111 of the railroad track.It revealed that the section of the railroad track that once connected Aguas Calientes (km 110) and Machu Picchu (km 112, ca.1900 m altitude) stations just above the bridge crossing the Urubamba River no longer exists.The 2 kilometers railroad section    Ruiz et al. (1990).

Drosophila navojoa Ruiz, Heed & Wasserman, 1990
This species is included in the Drosophila mojavensis cluster, together with Drosophila arizonae and Drosophila mojavensis (see Ruiz et al. 1990).The male terminalia line drawings depicted here complement those published by Ruiz et al. (1990).

Drosophila sonorae Heed & Castrezana, 2008
Being placed in the longicornis cluster of the longicornis complex, this species occurs in the western Mexican states of Sinaloa and Sonora (Heed and Castrezana (2008).The male terminalia line drawings depicted here complement those published by them.

Drosophila sticta Wheeler, 1957
Drosophila sticta occurs from El Salvador to Brazil (Brake and Bächli 2008).This Neotropical species, seldom attracted to fruit-baited traps, was included in a non-diagnosed group of its own by Clayton and Wheeler (1975).Two males belonging to this species emerged from living flowers of Cestrum intermedium Sendtn.(Solanaceae) collected by the author in 1982 in the municipality of Santa Isabel, state of São Paulo, Brazil (Vilela 1984).However, at that time, the author was unable to identify those males to species level, and they were just mistakenly cited as "unidentified species of tripunctata group" (see also Santos and It is however possible that specimens belonging to this species are present in many collections from different Brazilian states but have not been identified to the species level.The male terminalia illustrations of the holotype of D. sticta depicted here (Fig. 8) will complement that of the aedeagus tip in ventral view published in the original description (Wheeler 1957: 91;fig. 13).

Drosophila comosa Wheeler, 1968
In the original description of the Drosophila comosa, Wheeler (1968) states that "males of this species have an unusual development of thick hairs on the carina [hence the Latin epithet comosa, meaning hairy]" and "females have not been identified as yet".This is a rare and ungrouped species only known from 10 male type specimens (1 holotype and 9 paratypes) collected in Costa Rica on July 4, 1959 by W.B. Heed and H.L. Carson.
In the present paper the male terminalia of this species, apparently endemic to Costa Rica, is illustrated for the first time.However, it was of no help in clarifying its relationships with other species, as the general aspect of the sclerites is not similar to any Neotropical species of Drosophila known to this author.

Conclusion
The detailed analyses of the male terminalia of selected seven American species of the genus Drosophila depicted here reinforce the idea that the knowledge of their component sclerites are essential to reliably tell closely related species apart.Although the terminalia of the Peruvian specimen of D. guayllabambae depicted in this paper slightly differ from those illustrated by Rafael and Arcos (1985) for the Ecuadorian holotype, the two specimens are considered to belong to the same species.The differences may be due to both intraspecific variation and different interpretation of the sclerites given by different authors.This species seems to be a strictly mountainous species of the Ecuadorian and Peruvian Andes, while the remaining six analyzed species occur at variable altitudes of the American continent.

Note
This paper is a tribute to my great master and friend Gerhard Bächli in recognition for his Herculean efforts to create and manage TAXODROS, the indispensable database for anyone working on Taxonomy of Drosophilidae, and for his remarkable devotion to the Mitteilungen der Schweizerischen Entomologischen Gesellschaft as Editor-in-Chief during the past 22 years.

Figure 1 .
Figure 1.Urubamba Canyon (as seen from Machu Picchu ruins), Department of Cuzco, Peru.III.1984, C.R. Vilela phot.Arrow points the final stop of the railroad at km 112 marker (Estación Ferrocarril de Machu Picchu, that no longer exists).
Bächli (2017) al. 1990n extension of the unpaved Hiram Bingham roadway, currently leading from Machu Picchu pueblo to the Inca city of Machu Picchu.Based on the appearance of the male terminalia sclerites, Drosophila guayllabambae is closer to D. novemaristataDobzhansky & Pavan, 1943 (illustrated by Vilela 1983: 36)than to the remaining species belonging to the hydei subgroup.This species is apparently endemic to Venezuela and has a conspicuous surstylus covered with ca.20 peg-like setae, hence its nickname "bushy tail" used in the NDS-RC, Austin.The description of this species was based on specimens sampled from a strain established from flies collected in the Venezuelan Andean mountains at 25 km S of Mérida to San Cristobal, state of Mérida.As it also has been recorded from valleys of lower altitudes of the eastern Cordillera de Mérida, it might not be considered a true mountain species.According toFontdevila et al. 1990, D. nigrodumosa belongs to the mulleri cluster of the mulleri complex.According toBächli (2017)it belongs to the Drosophila mulleri species complex that also includes Drosophila aldrichi Patterson inPatterson &  Crow, 1940, Drosophila huayla Suyo, Pilares & Vasquez,  1988, Drosophila mulleri Sturtevant, 1921, and Drosophila wheeleri Patterson & Alexander, 1952.
Garcia et al. (2012)ter having analyzed and dissected the holotype of D. sticta in 1984 the author was able to clarify their identity.Thus, he could also assign to this species a single male specimen collected in 1998 from a banana-baited trap set by Hermes Fonsêca de Medeiros (Medeiros and Klaczko 2004) at Barreiro Rico Farm [currently Bacury Farm], in the municipality of Anhembi, state of São Paulo, Brazil.Later, another single specimen (sex not stated) was collected and identified as Drosophila sticta byGarcia et al. (2012)from a banana-baited trap set at the Botanic Garden of Porto Alegre, state of Rio Grande do Sul, Brasil.