Five additions to the list of Sepsidae Diptera for Vietnam: Perochaeta cuirassa sp. n., Perochaeta lobo sp. n., Sepsis spura sp. n., Sepsis sepsi Ozerov, 2003 and Sepsis monostigma Thompson, 1869

Abstract A recent collecting trip to Vietnam yielded three new species and two new records of Sepsidae (Diptera) for the country. Here we describe two new species in the species-poor genus Perochaeta (Perochaeta cuirassa sp. n. andPerochaeta lobo sp. n.) and one to the largest sepsid genus Sepsis (Sepsis spura sp. n.) which is also found in Sumatra and Sulawesi. Two additional Sepsis species are new records for Vietnam (Sepsis sepsi Ozerov, 2003; Sepsis monostigma Thompson, 1869). We conclude with a discussion of the distribution of Perochaeta and the three Sepsis species.


Introduction
Th e Sepsidae are a moderately large, cosmopolitan family of saprophagous fl ies, with over 300 extant species recorded from all zoogeographic regions (Ozerov 2005). Most species are attracted to dung, carrion, and other malodorous, decaying organic substrates (Pont and Meier 2002); i.e., by using diff erent substrates in diff erent microhabitats, the sepsid fauna from a specifi c locality can be quickly explored. Separating sepsids from the remaining saprophagous insects is also relatively straightforward because most sepsids can be easily recognized based on the constriction of the fi rst two abdominal segments which gives the fl ies a wasp-or ant-like habitus.
Here we update an existing species list for Vietnam by adding fi ve species: three are new to science while two others are new records. Th e relatively large number of additions is due to the fact that the Vietnamese sepsid fauna remains poorly studied (e.g. Ozerov 1993, Iwasa andTh inh 2008). Th e current species list comprises 21 species in six genera and is based on the sepsid world catalogue (Ozerov 2005) and subsequent taxonomic research by Iwasa and Th inh (2008). We complement this list by adding fi ve species that were collected during a brief collecting trip in July 2010: Perochaeta cuirassa sp. n., P. lobo sp. n., Sepsis spura sp. n., S. sepsi and S. monostigma.

Materials and methods
All fi ve species were collected between 11-16 July 2010 from Northern Vietnam (Ba Vi National Park and Sa Pa Valley). Cow dung was placed in various habitats for at least a few hours to attract sepsids, which were then caught by sweep-netting. Additional material for Sepsis sepsi and S. spura were also collected previously in Indonesia (Sulawesi and Sumatra) in 2007 and 2009. Specimens were photographed using a Leica Z16 APO-A stereomicroscope fi tted with a DFC425 digital microscope camera, and then digitally traced to illustrations using a Wacom© PTZ 630 tablet. We also amplifi ed and sequenced a 544-bp fragment of cytochrome oxidase c subunit I (COI) within the DNA barcoding region for the two new Perochaeta species based on the methods described in Tan et al. (2010). All type specimens and additional material are vouchered in 95% ethanol and kept in the Cryogenic Collection of the Raffl es Museum of Biodiversity and Research (RMBR), National University of Singapore, Singapore. We adopt the taxonomic terminology as described by Merz and Haenni (2000) for adult morphology (excluding terminalia) and Sinclair (2000) for male genitalia.

Taxonomy
Describing new species in genera that have not been revised recently requires extra care and justifi cation, because the risk of creating new synonyms based on overlooked or misinterpreted species in the literature is high. Fortunately, this is not the case for Pero-chaeta, which has only three described species (see Ozerov 2005 andAng et al. 2008) and no synonyms. In addition, the descriptions and illustrations for the described species are of good quality. Furthermore, molecular data are consistent with distinct species: the two new Perochaeta species are separated by 3.3% for the barcoding gene COI (uncorrected pairwise distances); while the distances of either to Perochaeta dikowi are 11.4% and 11.8% (see Table 1 for variable base pairs).
Describing a new Oriental Sepsis species in the absence of a generic revision is more problematic given that the genus is the largest in Sepsidae (ca. 80 described, valid species). Of the 23 Sepsis recorded in the Oriental region (Ozerov 2005), the Sepsis species described here closely resembles the widespread S. nitens Wiedemann, 1824 which has two synonyms (S. brevicosta Brunetti, 1910 andS. tuberculata Duda, 1926). Duda's (1926) description and fore leg illustration of S. tuberculata are suffi ciently detailed to confi rm that it is indeed a synonym of S. nitens. However, Brunetti's (1910) rather vague description of S. brevicosta based on one male from Calcutta and a few females from localities in other Indian localities (Calcutta, Shencotta, Tinpahar and Pusa), is more diffi cult to interpret. He describes S. brevicosta's fore femur as having "a small bump [on the ventromedial region] with three or four strong short spines". Th is description of a "bump" is in agreement with S. nitens (Fig. 21) while the ventromedial protrusion of the new species of Sepsis is shaped more like a spur (Figs 24,26). In addition, Brunetti's S. brevicosta is known from India, while the new Sepsis species described here is only known from Vietnam (Lào Cai) and Indonesia (Sumatra and Sulawesi).
As argued elsewhere, new species are hypotheses that are dependent on species concepts; it is therefore desirable that authors who describe species are explicit about which species concept was used and whether other species concepts would come to diff erent conclusions (Laamanen et al. 2003, Tan et al. 2008, 2010. Here we apply the Hennigian species concept (Meier and Willmann 2000) and use morphology and DNA sequence data (for Perochaeta) to estimate the species boundaries. Th e two new Perochaeta species are sympatric and the new Sepsis species is parapatric with S. nitens. In both cases we have not seen any intermediate specimens so that there is no evidence for hybridization. Th is supports our species hypotheses. However, both Perochaeta species are so rare that this test is relatively weak. As pointed out by Lim  (Table 1). When the distributional and morphological data are applied to the remaining species concepts in Wheeler and Meier (2000), most support the same species boundaries. Th e only exception is Mishler and Th eriot's (2000) phylogenetic species concept that requires a phylogenetic analysis before species can be delimited. However, such analyses are currently unavailable. We now describe the new Perochaeta and Sepsis species and state the new records for the two Sepsis species. Etymology. Th e specifi c epithet refers to the shape of the main scleral plate for the 4 th sternite, which resembles a cuirass or breastplate armor.

Perochaeta cuirassa
Diagnosis. Adult male Perochaeta cuirassa is very similar to Perochaeta lobo and can only be reliably distinguished from the latter based on the 4 th sternite [cf. P. cuirassa ( Fig.  1) and P. lobo ( Fig. 6)]: Th e sternite in P. cuirassa lacks distinct lobes on the posterior end of the 4 th sternite, while the sternite brush is thick and squat (as opposed to long and thin in P. lobo), and the main scleral plate is much broader (long as wide) than in P. lobo (twice long as wide). Th e hypopygium [cf. P. cuirassa (Figs 2-4) and P. lobo (Figs 7-9)] is also distinct, with P. cuirassa bearing a large median, decussating protrusion on the dorsal side of the surstylus, while P. lobo has a sub-median protrusion on the ventral side of the surstylus. Perochaeta cuirassa is also readily distinguished from all other Perochaeta species based on the morphology of the 4 th sternite and hypopygium: Th e sternites brush of P. cuirassa ( Fig. 1) has signifi cantly more bristles (>40 per brush) than either P. hennigi Ozerov, 1992 (Fig. 10) or P. dikowi (Fig. 12), both of which have only 5-6 large bristles in addition to a few weaker bristles. Perochaeta cuirassa also has strong bristles lining the distal margin of the sternite, which are not found in P. dikowi or P. hennigi. Th e surstylus of P. cuirassa (Fig. 2) resembles that of P. hennigi (Fig. 11), but can be distinguished by the large median surstylus projection, which is long and curved in P. cuirassa but short and broadly triangular in P. hennigi. Both P. dikowi (Fig. 13) and P. orientalis (De Meijere, 1913) (Fig. 14) lack large median projections. Perochaeta cuirassa can further be distinguished by the radial-medial cross-vein dividing the discal-medial cell which is in a ratio of 3 : 1 in P. cuirassa, 2.5 : 1 in P. dikowi, 2 : 1 in P. hennigi and 1 : 1 in P. orientalis.
Legs. Forelegs unmodifi ed in males; all femora and tibiae without posteriodorsal or anteriodorsal setae. Mid tibia with row of short setae on anterior apex. Rear tibia with barely-visible osomoterial patch on medial posteriodorsal side. Rear basitarsus with three ventral dark spines basally.
Hypopygium (Figs 2-4). Cercal plate with two very weak lobes; distal margin of each lobe covered with numerous setae. Hypopygium triangular with a two tooth-like projections on the inner side basal to where the surstylus branches off (Fig. 4). Surstylus itself fused to hypopygium, angled dorsally, and branches off subterminally (Fig. 3). Each surstylus has a large hook-like median projection that curves dorsally and decussates. Terminal section of surstylus shaped like a scapula, with cuticular "teeth" and setulae on distal margin, and a small inward-facing "tooth" on both the ventral and dorsal region subterminally pointing towards the median.
Distribution. Vietnam (Lào Cai). Etymology. Th e specifi c epithet is a phonetic translation of Greek "λοβό", which refers to the large, distinct lobe found on each lateral half on the posterior margin of the 4 th sternite.

Perochaeta lobo
Diagnosis. Th e adult male is very similar to P. cuirassa and can only be reliably distinguished based on the 4 th sternite and hypopygium. Th e 4 th sternite [cf. P. lobo (Fig. 6) and P. cuirassa (Fig. 1)] can be distinguished to species by the presence of distinct lobes on the posterior end, the long, thin sternite brush (as opposed to short and squat in P. cuirassa), and the sternite itself being narrower (half wide as long). Th e structure of the hypopygium (Figs 7-9) is also diagnostic given that it is the only Perochaeta with a surstylus that has a dorsal fl ap along its length and a long distal-pointing projection sub-basally. Other diagnostic characters that distinguish P. lobo from P. dikowi, P. hennigi and P. orientalis as described in diagnosis for P. cuirassa.
Description (male). Colour. As described in P. cuirassa except for fore and rear basitarsi, which are brown with yellow base, mid basitarsus yellow with slight brown region apically.
Head. As described in P. cuirassa. Th orax. As described in P. cuirassa; pleural pruinosity pattern as in Fig. 5.
Abdomen. Tergites and sternites 1-3 as described in P. cuirassa. Sternite 4 heavily modifi ed (Fig. 6); almost twice long as wide and raised from the abdomen. Posterior edge of 4 th sternite deeply invaginated and raised to form two large distinct lobes densely populated with strong bristles, mainly lining the outer discal margin. Two long, thin moveable appendages (= sternite brushes) branch off laterally on the posterior end of the sclerite, each with large, thick bristles on the outer region and some shorter, thinner bristles closer to the inside. A Y-shaped apodeme extends underneath and anterior to the 4 th sternite.
Hypopygium (Figs 7-9). Cercal plate with two very weak lobes; distal margin covered with numerous setae. Hypopygium triangular with bicuspid projection before the base of the surstylus. Surstylus itself fused to hypopygium and branches off terminally, with a dorsal fl ap along its length and a longish distal-pointing projection sub-basally. Terminal section of surstylus shaped like a scapula, with distal-pointing cuticular "teeth" and setulae on distal margin, and a very small inward-facing "tooth" on both the ventral and dorsal region subterminally pointing towards the median.
Distribution. Vietnam (Lào Cai). Taxonomic remarks. Sepsis monostigma is an Oriental species that resembles Sepsis pseudomonostigma Ursu, 1969 but is geographically exclusive from S. pseudomonostigma (which has only been recorded in South and South-east Europe and Central Asia). Sepsis monostigma can be diff erentiated by the two long medioventral spines (one short spine in S. pseudomonostigma) on the fore femur (Fig. 15), lack of ventromedial spinules on the fore tibia (Fig. 16) and surstylus (Fig. 17) being much thinner than that in S. pseudomonostimga.
Distribution. Indonesia (Sumatra, Sumbawa), Vietnam (Ha Tay). Etymology. Th e specifi c epithet old English for "spur", and refers to the distinct spur-like medioventral tubercle found on the male fore femur.

Sepsis spura
Diagnosis. Adult males of Sepsis spura closely resemble Sepsis nitens but can be distinguished by the following characters: (1) Medioventral tubercle on male fore femur of S. spura is spur-like and bent at a forward angle with two smaller adjacent spines dorsally positioned at the end of the tubercle and one larger spine at the ventral end (Figs 24,26), while the tubercle in S. nitens is thicker on the base and has its three spines positioned more in a anterio-posterior fashion (Fig. 21). (2) Th e basal laminalike projection on the fore tibia of S. spura (Figs 25,27) merges back with the tibia gently, but ends off with a distinct lobe in S. nitens (Fig. 22). Th e short spines found posteriorly on the projection are also much weaker than those found in S. nitens. (3) Sepsis spura (Figs 25, 27) has only one anterior lamina-like protrusion on the distal portion of the fore tibia, while S. nitens (Fig. 22) has such protrusions on both anterior and posterior sides. (4) Th e surstylus of S. nitens (Fig. 23) has a rather angular basal swelling and is relatively straight, curved only at the terminus, while the surstyli of S. spura (Figs 30, 31) has a rounded basal swelling and is medially curved for the entirety of the surstylus. Sepsis spura can be distinguished from other Sepsis, also based on the specifi c structure of the male fore leg ornamentation and the shape of the surstylus.
Description (male). Colour. Head capsule mostly brown with a thin light brown strip on gena; fascial margin black. Vertex dark brown. Facial carina and lunule light grey-brown. Pedicel dark brown, 1 st fl agellomere yellowish, arista brown. Proboscis whit-ish yellow. Forelegs wholly yellow. Mid and rear coxa yellow with brown base. Mid femur brown but yellow on basal and distal tips; mid tibia brown on basal half and diff uses to yellow on apical half. Rear femur yellow but brown on dorsal region, while rear tibia wholly brown. Fore tarsus with tarsomeres 3-5 brown, mid tarsus with tarsomeres 3-5 very lightly brown, rear tarsus with tarsomeres 4 and 5 brown. Wing clear except for basicostal cell and basal region of costal cell, which is light brown. Veins dark brown. Calypter clear, margin and fringe-hairs yellowish. Haltere white. Th orax mostly dark brown, but pronotopleuron is yellow. Abdominal tergites and sternites glossy dark brown.

Notes on Perochaeta
Perochaeta cuirassa and P. lobo are the fi rst Perochaeta species from Vietnam. Th at they are found in the same locality is surprising, given that Perochaeta is a small Oriental genus with only three described species. As discussed in Ang et al. (2008), the species in this genus appear to be only found in mid-and high-elevation localities in Asia above 650m ASL: Perochaeta dikowi was found on Fraser's Hill, Malaysia (1300m ASL; Ang et al. 2008), P. hennigi is only known from Th awalamtenne, Sri Lanka (670m ASL; Ozerov 1992) and Tamil Nadu, India (1200m and 1400m ASL; Iwasa and Tewari 1994). Th e type locality of P. orientalis is the Jiji Township in Taiwan ("Chip-chip" as designated by de Meijere; Duda 1926) which is within the Chung Yang mountain range that has an average elevation >1000m ASL, and can also found in regions such as Indonesia (Seram Is., 750m; Iwasa 2001). Th e new species P. cuirassa and P. lobo were found at 2600m ASL. A male Perochaeta specimen from Flores Island, Indonesia, examined by Hennig (1941) was also collected from a mid elevation site (1200m ASL). Unfortunately, the latter was poorly preserved and could not be described, but based on Hennig's illustration, it likely constitutes a new species. Given that Perochaeta species are restricted to elevated sites, we predict that the number of species will rapidly grow as more mid and high elevation localities are sampled. Adding two new species to Perochaeta is also of interest, because this (small) genus is very atypical for sepsids. Males of most sepsids have modifi ed forelegs (e.g., cuticular tubercles and stout, enlarged spines; see [18][19][21][22][24][25][26][27] which are used to grab the base of female wings during mating (Pont and Meier 2002, Ingram et al. 2008). However, Perochaeta has secondarily reduced foreleg armature, and at least P. dikowi has evolved a novel mounting behavior that does not involve the foreleg grasp (Ang et al. 2008). Th is has made Perochaeta a model for testing the correlation between the evolution of behavior and morphology (see Puniamoorthy et al. 2008Puniamoorthy et al. , 2009).

Notes on Sepsis
Sepsis spura appears to be a relatively widespread species ranging between Indonesia and Vietnam, and it can be found across all elevations, ranging from highlands (Sa Pa Valley, Vietnam; 1250m ASL) through mid elevation areas (Kampung Jawa, Indonesia; 650m ASL) to sea level (Bandar Gadang Beach, Indonesia). Such widespread species of insects are relatively common in Southeast Asia and are increasingly attracting attention because they can give rise to new species (see Balke et al. 2009). At all localities the S. spura is found in low numbers on bovine dung while other Sepsis species can be very abundant. Th is, along with its morphological resemblance to S. nitens, may explain why S. spura is only described now. It is likely to belong to a clade of Sepsis species without wingspots (see Su et al. 2008).
Sepsis sepsi was fi rst described from Sumbawa Is., Indonesia at 450m ASL (Ozerov 2003), but has subsequently been collected at low elevation sites in Sumatra as well as now at mid-elevation sites (800m ASL) in Ba Vi, Vietnam. Th is suggests that Sepsis sepsi is a relatively widespread species able to live in low to mid elevation habitats. We predict that it is also likely to be found in other areas between Indonesia and Vietnam. Th e record of the widely distributed Sepsis monostigma in Vietnam is not unexpected given that it falls within the recorded range from India to the far east of Russia.