Eight new species of Dexosarcophaga Townsend, 1917 (Diptera, Sarcophagidae) from the Neotropical Region

. Eight new Neotropical species of Dexosarcophaga Townsend, 1917 are described, ﬁ ve from Brazil, Dexosarcophaga phoenix sp. nov., Dexosarcophaga jandainae sp. nov., Dexosarcophaga patiuorum sp. nov., Dexosarcophaga petra sp. nov., and Dexosarcophaga sphaera sp. nov., one from Costa Rica, Dexosarcophaga limon sp. nov., one from Ecuador, Dexosarcophaga napo sp. nov., and one from Colombia, Dexosarcophaga pallida sp. nov. Male and female morphology is documented with photographs and illustrations, including details of the male terminalia for all new species and female terminalia of Dexosarcophaga phoenix sp. nov. and Dexosarcophaga sphaera sp. nov. With the addition of these new species, 58 species of Dexosarcophaga are now known, with records from the American continent spanning from the southern United States to northern Argentina.


Introduction
Dexosarcophaga Townsend, 1917 is one of the 46 genera recognized in the subfamily Sarcophaginae (Buenaventura & Pape 2018), and it included 50 valid species prior to this study (Pape 1996;Mello-Patiu 2000;Mello-Patiu & Pape 2000;Silva & Mello-Patiu 2010;Carvalho-Filho & Esposito 2011;Carvalho-Filho et al. 2014De-Souza et al. 2021). The genus is endemic to the American continent and its geographic distribution is almost exclusively restricted to the Neotropics, with only D. transita Townsend, 1917 reaching into the southern part of the Nearctic and D. cistudinis (Aldrich, 1916) [if considered as a member of this genus, see below] being exclusively Nearctic (Pape 1996). The biology of the species of this genus is very poorly known. Dexosarcophaga termitaria (Lopes, 1939) was described from a single female bred from larvae recovered from a termite nest, and D. lenkoi Lopes, 1968 was described based on specimens bred from a nest of carpenter ants (Lopes 1939(Lopes , 1968. Lopes (1982b: 318) noted that he was unable to rear larvae of Dexosarcophaga spp. "in spite of the saprophagous characteristics", and he provided a personal observation of D. transita being attracted to damaged termite nests. The suggestion by Lopes (1968Lopes ( , 1982b) of a possible association with tunnels of wood-boring beetles or other insects was given without any supporting evidence. Méndez (2012) documented that D. megista (Hall, 1933) and D. varenna (Dodge, 1968) are attracted to damaged aerial termite nests for larviposition, and he referred to this as a "predator-prey association" although without specifi c details of the larval feeding. Adults of Dexosarcophaga have often been collected from decomposing vertebrate carrion, which has led authors to include this genus among necrophagous taxa or even imply a forensic importance (e.g., Buenaventura et al. 2009, Ledo et al. 2012, Paseto et al. 2014, Smania et al. 2014, Sousa et al. 2015. However, none of these studies reports any breeding records. Dexosarcophaga is one of the most species-rich genera of Neotropical Sarcophagidae, and several new Neotropical taxa have been added in recent decades (Mello-Patiu 2000;Mello-Patiu & Pape 2000;Silva & Mello-Patiu 2010;Carvalho-Filho & Esposito 2011;Carvalho-Filho et al. 2014De-Souza et al. 2021). Lopes (1975aLopes ( , 1982b included Bezzisca Lopes, 1975, Dexosarcophaga andFarrimyia Dodge, 1965 in the subtribe Dexosarcophagina Lopes, 1975, thereby proposing shared morphological similarities. Pape (1996) considered Bezzisca, Farrimyia and Ectomyia Dodge, 1968 as junior synonyms of Dexosarcophaga. Mello-Patiu & Pape (2000) followed Pape's (1996) proposal and defi ned species of Dexosarcophaga by the following combination of character states: most postgenal setae black (white setae restricted to posteriormost part); postalar wall setose; presence of a ctenidium composed of normal spines on the male mid femur; male sternite 5 broadly V-shaped; male terminalia black; juxta small; presence of three styli; occipital setae above occipital foramen black; phallus divided into basi-and distiphallus; and female tergite 8 divided, with broad and ventro-laterally truncated halves connected medially by a narrow strip. Mello-Patiu & Pape (2000) recognized three species-groups in this genus, recovering the genera recognized by Lopes (1975bLopes ( , 1982b in Dexosarcophagina. The authors did not mention where the monotypic genus Ectomyia could be tentatively placed in these groups. Buenaventura & Pape (2018) followed Pape's (1996) defi nition of Dexosarcophaga in their phylogenetic analysis based on morphological characters, in which Cistudinomyia Townsend, 1917 emerged as the sister group to Dexosarcophaga and these two groups were recognized as subgenera of a broader Dexosarcophaga. In the present paper, eight new species of Dexosarcophaga from the Neotropical Region are described, following the broad circumscription of Pape (1996) until a comprehensive phylogenetic analysis of Dexosarcophaga clarifi es relationships among species of this group of fl esh fl ies.

Material and methods
Descriptions were based on material from the Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil (MNRJ); the Museu de Zoologia, Universidade de São Paulo, São Paulo, Brazil (MZUSP); the Instituto Nacional de Biodiversidad, Costa Rica (INBio), and the Swedish Museum of Natural History, Stockholm, Sweden (NHRS). The fi rst species (Dexosarcophaga phoenix sp. nov.) is provided with a detailed description, and the following seven species are described by listing the most distinctive differences from this species plus, where necessary, any features needed for separation from already described species. The material includes 30 specimens (24 ♂♂, 6 ♀♀), designated as either holo-or paratypes, all of which were lost in the tragic fi re of the MNRJ in September 2018 (Kury et al. 2018), and they are indicated as "[lost]" in the material examined sections and in the captions to illustrations. A specimen may be designated as the name-bearing type even if it is known to have been lost (ICZN 1999: Article 73.1.4), and should future studies reveal one or more of the taxa described here to represent species-complexes, neotypes may be warranted . Geographic distributions are based on the labels of the type material and given by country, with provinces, states, or departments in parentheses.
Male terminalia were dissected, cleared in a heated 10% KOH solution, and neutralized in an acetic acid-ethanol solution. The dissected terminalia were temporarily mounted in glycerin on glass slides and observed and illustrated using a MOTIC K400® stereo microscope and a ZEISS M80® compound microscope, both equipped with a camera lucida. After examination and illustration, the terminalia were placed in a microvial fi lled with glycerin and pinned with their respective specimens. Photographs of the habitus and terminalia were taken with a Leica DFC450C digital camera coupled with a Leica M205 stereo microscope. Final composite images were digitally stacked by the software Leica Application Suite ver. 4.8.0. Images were edited using Adobe Photoshop ver. CS6® and Adobe Illustrator ver. CS6® (Adobe Systems, Inc., San Jose, CA). Label data were standardized to facilitate capture.

Diagnosis
Vein R 1 bare. Male: scutellum without apical setae; vesica funnel-shaped in ventral view (Fig. 1E); juxta like a cap or fl at helmet, spread evenly across the paraphallus (Fig. 1D). Female: the two halves of T8 separated by less than the width of the cercus, each half of T8 with a deep anteromedian incision ( Fig. 2A-B).

Etymology
The species epithet 'phoenix' should be treated as a noun in apposition. The 'Phoenix' is a mythological, resilient bird that arises from the ashes after having burned itself to death. This characteristic could symbolize a new beginning, like the start over of the

Remarks
Dexosarcophaga phoenix sp. nov. is morphologically similar to D. jandainae sp. nov. and D. angrensis (Lopes, 1975). Dexosarcophaga phoenix sp. nov. and D. jandainae sp. nov. can be differentiated from D. angrensis by the absence of setosity on vein R 1 and by the complete visibility of the juxta in lateral view (Figs 1D, 4D) as opposed to a setose vein R 1 and an indistinct juxta in D. angrensis (Silva & Mello-Patiu 2010: fi gs 46-47). Dexosarcophaga phoenix sp. nov. and D. jandainae sp. nov. share a helmetshaped juxta (Figs 1D-E, 4D-E), but they can be differentiated in particular by the shape of the vesica in ventral view, which is broadest proximally and looks like a funnel in D. phoenix sp. nov. (Fig. 1E), and is broadest distally and looks like an 'X' in D. jandainae sp. nov. (Fig. 4E).

Diagnosis
Vein R 1 bare. Male: scutellum without apical setae; vesica in ventral view X-shaped (Fig. 4E); juxta like a cap or helmet, spread across the paraphallus and strongly projected towards the ventral surface in lateral view, i.e., with the apical part recurving (Fig. 4D). [Female unknown.]

Etymology
The species epithet 'jandainae' ('jandaina' + 'ae'), a feminine genitive, is given in honor of Jandaína Rodrigues dos Santos de Oliveira, sister of the fi rst author.

Female
Unknown.

Female
Unknown.

Etymology
The species epithet 'limon' should be treated as a noun in apposition. The name refers to the province where the type locality of the new species is located. Frons about 0.25 × head width at level of ocellar triangle; 10 well-developed frontal setae reaching level of apex of pedicel; fi rst fl agellomere approximately 4 × as long as pedicel; thorax with slightly yellowish-silver pollinosity (Fig. 3F); intra-alars 2 + 2, postpronotals 2; meral setae 5-7; scutellum with pair of reduced apical setae; abdomen dark brown, with golden pollinosity; T5 with row of 12 marginal setae; cercus with pointed apex (Fig. 6B-C); surstylus trapezium-shaped in lateral view (Fig. 6B); pregonite almost straight, 1.5 × length of postgonite and with broad base (Fig. 6D); vesica consisting of membranous and sclerotized areas and with ventral dark lobe with 2-3 spines (Fig. 6D-E); juxta equipped with numerous spines (Fig. 6D-E); median stylus with enlarged base and narrow apex, slightly shorter than lateral stylus, base curved towards ventral and dorsal margins of paraphallus and with apical spines (Fig. 6D-E); lateral stylus with base slightly curved towards ventral margin of paraphallus and with apical spines (Fig. 6D-E).

Female
Unknown.

Remarks
Dexosarcophaga limon sp. nov. is morphologically similar to D. petra sp. nov. and D. paulistana (Lopes, 1982) by having a vesica with a rounded lobe in lateral view (Figs 6D, 7D; Lopes 1982a: fi g. 58). Dexosarcophaga limon sp. nov. and D. petra sp. nov. can be differentiated from D. paulistana by the distally completely rounded vesica (Figs 6D, 7D) and with a digitiform projection distally in D. paulistana (Lopes 1982a: fi g. 58). Males of D. limon sp. nov. can be differentiated from those of D. petra sp. nov. by the following features: ST5 with arm-like processes almost parallel-sides in middle part, apically rounded, and almost half as long as sternite (Fig. 6A), cercus almost straight in lateral view (Fig. 6B), and juxta with small spines apically (Fig. 6D-E). In D. petra sp. nov., these features appear as follows: ST5 with arm-like processes gradually tapering, apically pointed and distinctly shorter than half of sternite, and with two small but distinct lobes projecting near the midline (Fig. 7A); apical half of cercus curved anteriorly (Fig. 7B), and juxta with small spines only at base (Fig. 7D-E).

Diagnosis
Vein R 1 bare. Male: scutellum with a pair of reduced apical setae; vesica rounded with an apical projection in lateral view and rectangular with two projections in inferior part and two thorny projections in superior part in ventral view (Fig. 7D-E); juxta with spines at base (Fig. 7D-E).

Etymology
The species epithet 'petra' should be treated as a noun in apposition. The name is a Latin noun 'petra', meaning 'stone' (= 'pedra' in Portuguese) and refers to the type locality of the new species (Trilha Cidade de Pedra). Frons about 0.18 × head width at level of ocellar triangle; 9-12 well-developed frontal setae reaching level of apex of pedicel; gena and genal groove with slightly yellowish-silver pollinosity; intra-alars 2 + 2; meral setae 7-9; scutellum with pair of reduced apical setae; T4 with 1-2 pairs of lateral marginal setae; row of about 10 marginal setae on T5; arms of ST5 about half of basal length and inner margin with distinctly projected lobes (Fig. 7A); cercus with apical half curved anteriorly (Fig. 7B); cercal prong rounded (Fig. 7B-C); surstylus narrow, with long setae in apical half (Fig. 7B); vesica with membranous and sclerotized areas, rounded in lateral view with apical projection, rectangular in ventral view with two projections in proximal part and two thorny projections in distal part (Fig. 7D-E); juxta membranous and with spines proximally (Fig. 7D-E); median stylus narrow, with apical spines, shorter than lateral stylus and with base expanded towards ventral and dorsal margins of paraphallus (Fig. 7D-E); lateral stylus narrow, base twisted towards ventral margin of paraphallus and with apical spines (Fig. 7D-E).

Female
Unknown.

Remarks
Dexosarcophaga petra sp. nov. is morphologically similar to D. limon sp. nov. (see remarks under that species). There are also similarities with D. malaisei Dodge, 1968and D. salgada De-Souza, Souza, Soares & Carvalho-Filho, 2020(De-Souza et al. 2021). Males of these species have cercus curved and vesica mostly rounded. Lopes (1975c) reviewed the morphology of D. malaisei and commented that the terminalia of the holotype were lost apart from syntergosternite 7 + 8 (= "fi rst genital segment"). However, since the species does not have paratypes and the illustrations of the phallus provided by Dodge (1968: fi gs 17-18) are sketchy and insuffi cient for recognizing D. malaisei, we cannot make a detailed morphological comparison between these three nominal species. Consequently, either D. petra or D. salgada could be a junior synonym of D. malaisei, but we will here rely particularly on the shape of the male pregonite and cercus as illustrated by Dodge (1968: fi g. 17). The evenly curved or C-shaped pregonite and the cercus that is straight over most of its length and curved in the distal fourth appear suffi ciently different from both D. petra (pregonite with middle part straight; cercus bent rather than curved) and from D. salgada (pregonite with only distal part curved; cercus evenly curved) to reject conspecifi city. Therefore, we consider D. petra and D. salgada as valid nominal species until future studies produce material from the type locality of D. malaisei (i.e., Barro Colorado, Panama), allowing a better resolution of this question. Dexosarcophaga petra and D. salgada can be separated by the following differences in the shape of the vesica in particular: broadest proximally (lateral view), with broad median fi ssure, without median knob-like projections and with an apical projection in lateral view (Fig. 7D)

Diagnosis
Vein R 1 bare. Male: scutellum with a pair of reduced apical setae; vesica like a small sclerotized plate ( Fig. 9D-E); juxta like a narrow band (Fig. 9D-E). Female: the two halves of T8 widely separated by at least four times the width of cercus, each half pointed in its median part ( Fig. 10A-B).

Etymology
The species epithet 'sphaera' should be treated as a noun in apposition. The name is a Latin noun 'sphaera', meaning 'ball' and alluding to the spherical paraphallus.

Etymology
The species epithet 'napo' should be treated as a noun in apposition. The name refers to the province (named after the Río Napo) of the type locality of the new species.

Female
Unknown.

Etymology
The species epithet 'pallida' should be treated as a noun in apposition. The name is a Latin adjective 'pallida' (= Latin), meaning 'pale', and alluding to the light apex of the distiphallus.

Female
Unknown.

Remarks
Dexosarcophaga pallida sp. nov. is morphologically similar to Dexosarcophaga currani Dodge, 1968. Males of both species have a stick-like vesica in lateral view ( Fig. 12D; Mello-Patiu 2002: fi gs 53, 56) and a membranous projection from the ventral part of the paraphallus at the level of the acrophallus ( Fig. 12D; Mello-Patiu 2002: fi gs 53, 56). The latter projection may be formed by the paraphallus or be a composite of both the paraphallus and the juxta. These two species can be differentiated by the following features: vesica strongly bifurcated at apex (Fig. 12E) and lateral and median styli approximately ⅓ the length of paraphallus (Fig. 12D-E) in Dexosarcophaga pallida sp. nov., versus vesica slightly bifurcated at apex (Mello-Patiu 2002: fi g. 58) and lateral and median styli approximately ½ the length of paraphallus (Mello-Patiu 2002: fi gs 53, 56-57) in Dexosarcophaga currani.

Discussion
Sarcophagidae species, as those of many other Diptera families, are mostly identifi ed based on the male terminalia (Sinclair et al. 2013). The morphological defi nitions of Dexosarcophaga proposed by Lopes (1975bLopes ( , 1982b [as tribe Dexosarcophagina], Pape (1996), Mello-Patiu & Pape (2000), and Buenaventura & Pape (2018) show that there is not an exclusive male character state that can unambiguously identify species as belonging to this genus. Buenaventura & Pape (2018) tested the monophyly of this taxon. However, their sampling included just three species, one species for each of the species-groups proposed by Mello-Patiu & Pape (2000) plus the monotypic Cistudinomyia. The autapomorphy shared by species of Dexosarcophaga in this phylogenetic study was the C-shaped male pregonite (e.g., Figs 1D, 4D), and this character state was used by the authors to propose Cistudinomyia as a junior synonym under a more broadly defi ned Dexosarcophaga. Our study indicates that there are species of Dexosarcophaga with a mostly straight pregonite, which would a priori appear as plesiomorphic (e.g., Fig. 5C). Thus, we have abstained from assigning the new species in this paper to any of the groupings proposed by Mello-Patiu & Pape (2000) or Buenaventura & Pape (2018), awaiting a comprehensive phylogenetic analysis of these morphological variables to test the monophyly of this genus as well as the character state changes involved.
Therefore, we agree that the better approach to recognize a species as belonging to Dexosarcophaga is to use a combination of characters, as proposed by Lopes ( 1975bLopes ( , 1982b and Mello-Patiu & Pape (2000). Females are diffi cult to identify, and the female sex is still unknown for many species of Sarcophagidae; however, Lopes (1975bLopes ( , 1982b and Mello-Patiu & Pape (2000) observed that female tergite 8 with broad and ventro-laterally truncated halves connected medially by a narrow strip as a possible autapomorphy for Dexosarcophaga. Thus, female terminalia may add important diagnostic and phylogenetic information for a better understanding of the evolution of these fl ies.