Review of the Lispe tentaculata-group (Diptera: Muscidae) in China, with one new synonym

ABSTRACT The Chinese fauna of Lispe tentaculata-group is reviewed. A new synonym for L. sericipalpis Stein, 1904 is established: L. fanjingshanensis Wei, 2006 n. syn. Taxonomic status of the suspected species L. alpinicola Wu & Fan, 1981 is discussed. The keys to males and females of all five species within the group in China are established. Morphological details of both sexes are illustrated using light microscopy photographs, scanning electronic microscopy images and line drawings, of which male and female terminalia are summarized as novel characteristics for diagnosis. The group is divided into two subgroups (L. tentaculata-subgroup and L. orientalis-subgroup) based on two characteristics (first fore tarsomere and phallus). To fill the gap of molecular research in Lispe, we construct a phylogenetic tree within the genus focusing on L. tentaculata-group, which verifies the division of two subgroups.

Well-defi ned as it is, L. tentaculata-group remains to be treated more explicitly and carefully since considerable morphological diff erences still exist between the species within this group.
Th ough work on Lispe morphology is quite abundant, rare molecular research is available for us to explore the true relationship within this genus. As mitochondrial DNA sequences (12S, 16S, COI and Cytb) and nuclear genome sequences (28S, Ef1a and a fragment of the carbamoly phosphate synthetase (CPS) region of the CAD gene) are becoming increasingly useful to reconstruct the relationships within the Muscoidea, especially the Muscidae (Kutty et al, 2008). Here we present the phylogeny of four species from L. tentaculata-group and species from other Lispe groups based on the seven sequences mentioned above.

TAXONOMY AND MORPHOLOGY
All specimens studied are deposited in MBFU except that the type specimens of L. aipinicola Zhong, Wu & Fan and L. tienmuensis Fan are deposited in SEMCAS. Some specimens of L. tentaculata-group are stored in MNHN, which are listed beneath.
Distribution data in China is taken from Fan (1992) with additional records of collection. For distribution information around the world, it refers to Vikhrev (2014). Wellacknowledged synonyms and quotations are omitted here as they are available in Hennig (1960), Pont (1977, 1986 and Vikhrev (2014).
Light microscopy photographs were taken with a Canon 550D camera mounted on an Olympus SZX16 stereomicroscope except that photos of L. tienmuensis with low quality were taken years ago with poor equipments. Scanning electronic Microscopy (SEM) photographs were taken with a HITACHI S34Q scanning electron microscope (Hitachi Corp., Tokyo, Japan) at the Microscopy Core Facility, Biological Technology Center, Beijing Forestry University.

Molecular phylogeny
Eight Lispe species were selected, including four species from L. tentaculata-group (L. consanguinea, L. orientalis, L. sericipalpis and L. tentaculata) which represent most characters of this species group. As Limnophora simulans Stein, 1913 andCoenosia semifumosa Stein, 1914 are close to Lispe (according to Kutty et al. 2014), they are selected to root the tree. Pinned specimens are used for molecular experiment in this paper (for collecting date and location of the specimens, see Table 3). DNA extraction for all species is carried out utilizing phenol-chloroform extraction as described in Zhang et al. (2016), except that genomic DNA of L. apicalis Mik, 1869 is extracted using QIAGEN DNeasy Blood and Tissue Kit following the manual. Seven genes are amplifi ed including four mitochondrial genes (12S, 16S, COI and Cytb) and three nuclear genes (28S, Ef1a and CAD) using primers (Table 2). Th e reaction system of PCR (polymerase chain reaction) follows Zhang et al. (2016) and the PCR procedure for amplifi cations are shown in Kutty et al. (2010). PCR products are sequenced bidirectionally by BGI in the same way as Zhang et al. (2016).

Lispe tentaculata-group
-Th e Lispe tentaculata-group is distinguished from other species-group of this genus by the following characters (modifi ed from Snyder 1954, Hennig 1960and Vikhrev 2011, 2014: arista plumose, palpi abruptly expanded in apical half; meron setulose above hind coxa and the following leg chaetotaxy: fore tibia mostly without p setae (only present in L. alpinicola and some aberrant L. tentaculata); mid tibia mostly with 1 p (3 p in L. alpinicola), hind tibia with 1 weak pd and 1 strong ad.
Herein, we update the diagnosis with more detailed genital characters of both sexes (some male genital characters have been mentioned by Snyder 1954) and bring up the distinctive characteristics of two subgroups: Male: fi fth sternite with a median, anteriorly directed process in addition to the short and apically semi-membranous lateral processes; cercal plate with an inverted U-to V-shaped incision on each side (Fig. 1); gonopod degenerated and fused with hypandrium into a projection (Figs 6; 7); phallus non-membranous. Female: abdominal tergites 6 and 7 sometimes divided medially (L. consanguinea and L. tentaculata), abdominal sternite 6 sometimes degraded into 2 small semi-transparent setulose sclerites (L. consanguinea and L. tentaculata), abdominal tergite 7 fused with sternite 7, abdominal tergite 8 partly fused with sternite 8, epiproct divided medially into 2 setulose sclertites, cerci with appendage (Figs 3; 4).

REMARKS
Vikhrev (2011) doubted that this species might be a synonym for L. tentaculata and later Vikhrev (2014) proposed this synonym for the reason that the presence of additional p on fore and mid tarsi of L. alpinicola also presents in several specimens of L. tentaculata in the collection of Zoological Museum of the Moscow University. However, no other characters were compared and discussed between these two species in his works. In order to clarify the status of L. alpinicola, we tried to fi nd the types in SEMCAS but failed, which is likely to be lost. Without examining the types, we can't draw a convincing conclusion about its validity. Here we present the original text (Zhong et al. 1981, in Chinese and with an English abstract, not cited in Vikhrev's works): "Th is species is closely related to the Holarctic species Lispe tentaculata, from which it differs in having the palpi black, the fore tibia with a median seta on posterior surface, the process of the fore metatarsus rather long, the mid tibia with 3 pd etc." Also, in Zhong, Wu & Fan's remarks, females are distinguished from the color of velutinous patches at postesutural part of scutum. According to the original description, we separate these two species in the keys above. Another evidence confi rms our opinion about the validity of L. alpinicola is its diff erence in male phallus compared with other related species. Th eir phallus images are placed together for comparison (Fig. 11) especially concentrated on: 1) relative length of aedeagal apodeme and hypandrium; 2) relative length of paramere and phallus; and 3) relative size of submedian membranous part which only exists in L. alpinicola.

REMARKS
Th e fi nger-like protuberances on fore tarsi of L. consanguinea and L. tentaculata look identical by eyes. However, SEM images of the structure reveal interspecifi c diff erence concerning the blunt end of the protuberances (Fig. 5): L. consanguinea possesses less thorns (Fig. 5C) while L. tentaculata has thicker and bended thorns (Fig. 5F). More to mention, all these thorns are bifurcate except the middle one in both species (Fig. 5C, F).   Fan (1974) recorded Lispe tienmuensis (Figs 9, 10) (though the specimens were labeled as type material in 1963, it was not until 1974 when the work was published) and noted, "Closely related to L. orientalis Wd. but with narrowed parafacialia, poorly haired mid-femur in both sexes, narrowly pointed anal He commented that it is closely related to L. quaerens but diff ers in its greyish blue abdominal pollinosity and developed distiphallus which is bended at a vertical angle (Fig. 10B). After going through original description (focus on leg chaetotaxy especially hind legs: femur with complete ad row and incomplete av row in basal half, tibia with a developed median ad) and illustrations of L. fanjingshanensis, we propose that it is conspecifi c to L. sericipalpis. DeGeer, 1776 (Figs 1D;2D;4B, D, F;5D, E, F;11C, D) Lispe tentaculata DeGeer, 1776: 86.

REMARKS
Hennig's drawing (1960: 408, fi g. 131, with a bended distal part) of the phallus of L. tentaculata is not the same as in our dissection (Fig. 11D, specimen from Xinjiang). In our dissection, the phallus of L. tentaculata is straight and without a bended distal part, identical to Snyder's drawing (1954: 15, fi g. 53) of the phallus of L. sociabilis, about which he stated "might serve equally well for patellata and Nearctic specimens of tentaculata" (Snyder 1954: 38). Snyder's statement was confi rmed by Vikhrev (2014: 154). Th ese images are placed in one plate for comparison. We speculate that Hennig's drawing was based on either an aberrant L. tentaculata or an underlying species that curiously have both modifi ed fi rst fore tarsomere and bended phallus, an intermediate taxon between L. tentaculata-subgroup and L. orientalis-subgroup (more information in Discussion below).

MOLECULAR PHYLOGENY
Analysis of maximum likelihood and Bayesian inference recovered largely congruent results and most of the species has a high leaf stability value (Fig. 12). Th e monophyly of Lispe is highly supported (maximum likelihood bootstrap: 98, Bayesian posterior probability: 1), which was already suggested by Kutty et al. (2014). Th e L. tentaculata-group is confi rmed to be monophyletic here as well (maximum likelihood bootstrap: 87, Bayesian posterior probability: 1).

A B
Based on two morphological features (fi rst fore tarsomere and phallus), the L. tentaculata-group could be separated as two subgroups (L. tentaculata-subgroup with modifi ed fi rst fore tarsomere and straight phallus while L. orientalis-subgroup is distinguished with regular fi srt fore tarsomere and bended phallus). Th e tree topologies from the diff ernt analyses, ML and Bayesian, are highly corroborated that the group is split into two clades (maximum likelihood bootstrap: 100, Bayesian posterior probability: 1). Th erefore, the proposal of two subgroups within L. tentaculata-group is verifi ed.

DISCUSSION
Our analyses of this group both in morphology and phylogeny reveal the relationships between four species within the L. tentaculata-group. Other holarctic species from the group all fall strictly into the defi nitions of two subgroups except the Nearctic L. sociabilis, which has straight phallus but unmodifi ed fi rst fore tarsomere. Another case with characters that contradict with the diagnoses of two subgroups is Hennig's drawing of L. tentaculata (1960: 408, fi g. 131). We can not be certain about whether Hennig's illustration was derived from an abnormal L. tentaculata or an independent species that has both modifi ed fi rst fore tarsomere and bended phallus. However, these two cases are clear evidences of the existence of intermediate taxa between L. tentaculata-subgroup and L. orientalis-subgroups. More species should be involved in both morphological and phylogenetic research to explore the diversifi cation patterns within this species-group.