Redescription of Trochosa urbana (Araneae: Lycosidae) with notes on its distribution

Abstract A survey of specimens identified as Trochosa urbana O. Pickard-Cambridge, 1876, kept in the Zoological Museum, University of Turku, reveals that they belong to four species: T. urbana, T. dentichelis Buchar, 1997, T. ruricola (De Geer, 1778), and T. ruricoloides Schenkel, 1963. Trochosa urbana is redescribed based on syntypes, and a lectotype male is designated. Our study found that the actual range of T. urbana is restricted to eastern Africa (from Egypt to the Seychelles) and Israel, and not stretching from Africa to India as was thought before. Four species and one subspecies are synonymized with Trochosa urbana O. Pickard-Cambridge, 1876: Caporiaccosa arctosaeformis (Caporiacco, 1940) syn. n., Geolycosa gofensis (Strand, 1906) syn. n., Hogna sansibarensis (Strand, 1907) syn. n., Pirata molensis Strand, 1908 syn. n., and Trochosa urbana hova (Strand, 1907) syn. n. Two genera, Caporiaccosa Roewer, 1960 syn. n. and Trochosippa Roewer, 1960 syn. n. are synonymized with Trochosa C. L. Koch, 1847. For two species, we establish new combinations: Trochosa meruensis (Lessert, 1926) comb. n. ex Trochosippa, and Hogna taurirtensis (Schenkel, 1937) comb. n. ex Pirata. Trochosa dentichelis, previously known from Bhutan, is reported from India. Trochosa aquatica Tanaka, 1985 is likely a junior synonym of T. ruricoloides. The latter species, previously known from China, is reported for the first time from India, Indonesia, Malaysia, Papua New Guinea, and Thailand.


Introduction
Trochosa urbana O. Pickard-Cambridge, 1876 is considered to be a widespread species, occurring from Africa to India (Saaristo 2010;World Spider Catalog 2020). However, this species lacks diagnostic figures or a proper written diagnosis that allow it to be separated from congeners. Identification of both sexes of Trochosa species, even in wellstudied areas such as Europe, is difficult, and especially so for females. Therefore, it is unclear how arachnologists have previously been able to identify specimens as T. urbana considering that all existing figures of the epigyne are not similar to each other (compare figs. 7a-e in Nentwig et al. 2019). We found no reference of who reported T. urbana from India other than the information provided in Roewer's catalog (Roewer 1955a: 241). While working in the Zoological Museum in Turku (ZMUT), we found numerous specimens identified as T. urbana from China, India, Indonesia, Malaysia, Philippines, the Seychelles, Thailand, Turkey, and Vietnam. Examination of this material revealed four morphospecies. To determine the true T. urbana and to provide a detailed redescription of this species, we borrowed the syntypes from the Oxford University Museum of Natural History. Comparison of the types with the four morphospecies from the ZMUT revealed that only specimens from the Seychelles can be considered T. urbana.
The goals of this paper are to provide diagnostic figures of Trochosa urbana, discuss its diagnostic features, and comment on five subspecies assigned originally to T. urbana, some of which are considered to be in different subfamilies of Lycosidae (Lycosinae and Zoicinae).

Material and methods
Specimens were photographed using an Olympus Camedia E-520 camera attached to an Olympus SZX16 stereomicroscope or to the eye piece of an Olympus BH2 transmission microscope, and an SEM JEOL JSM-5200 scanning electron microscope at the Zoological Museum of University of Turku, Finland. Digital images were prepared using CombineZP image stacking software. Illustrations of internal genitalia were made after clearing them in a 10% KOH aqueous solution. Lengths of leg segments were measured on the lateral side. All measurements are given in millimeters.
Depositories: OUMNH = Oxford University Museum of Natural History, ZMMU = Zoological Museum of Moscow University, ZMUT = Zoological Museum of University of Turku.
We identified four species from the examined specimens at ZMUT, all previously identified as Trochosa urbana: T. urbana, T. dentichelis Buchar, 1997, T. ruricola (De Geer, 1778, and T. ruricoloides Schenkel, 1963 fig. 3 B-C,F), and the tegular apophysis lacks an anterior protrusion (v. with protrusion in T. dentichelis and T. ruricoloides, cf. Figs 2a-c and 2e,g-h). In addition, T. urbana differs from T. dentichelis by lacking spines on the tibia and cymbium (v. 1-2 spines on the prolateral side of the cymbium and 2 spines and 1 spine on the prolateral and dorsal sides of the tibia, respectively). Females of T. urbana are indistinguishable from those of T. dentichelis and T. ruricoloides.
Palp as in Figs 1a, 2a-d, 3a-c. Distal part of cymbium conical with straight edges; tip with claw. Anterior side of tegular apophysis flat. Tip of embolus curved toward the synembolus.
Epigyne as in Figs 4a-c,g. Anterior hoods deep, curved inward, their tips converging, widely separated by the approximate distance of the narrowest part of septum. Septum as long as wide, stalk widest anteriorly. Width of receptacles equal to width of septal stalk. Receptacle supplied with massive conical gland (Rg).
Remarks: Femur I spination varies and can be 1, 2, or 3 retrolateral spines. Therefore, species cannot be separated using spination. Carapace/femur I length ratio greatly varies. The specimens from Egypt (types) are larger than those from the Seychelles, although some insular specimens are similar in size (Fig. 5). We found no distinct differences in the body proportions between T. urbana and sibling species (Fig. 5).
Distribution: According to the material studied, this species is limited to eastern Africa, occurring from Egypt to the Seychelles, and seems to be present on Nosy Be Island north of Madagascar. In Asia, it has been properly documented in Israel only. A record of this species from Iran (Roewer 1955b) seems to refer to either T. ruricola or T. hispanica (see Marusik & Nadolny 2020). Buchar, 1997

Trochosa dentichelis
Trochosa dentichelis Buchar, 1997: 29, f. 30-34 (♂♀).  Remarks: This species is known from a single taxonomic entry and only from Bhutan. In the original description it was compared with T. ruricola. We are not absolutely sure that our specimens are conspecific with the holotype of T. dentichelis.
Diagnosis: Males of this species can be distinguished from other species with a tooth on the fang by the shape of the cymbium, tegular apophysis, and embolus. Cymbial length/width ratio in T. dentichelis is 2.4 (v. 1.8-2.2 in other species). The anterior side of the tegular apophysis has a protrusion (Pr), and the posterior side is rounded (Pt) (Fig.  2g-h) (v. without a protrusion and posterior side slightly curved or with well-developed, rounded protrusion in T. ruricoloides). Embolus and synembolus are parallel in T. dentichelis (Fig. 3d) (v. tip of embolus slightly curved toward the synembolus or coiled). In addition, this species differs from congeners by having 1-2 spines on the prolateral side of the cymbium (Fig. 1f). Females of T. dentichelis, T. urbana, and T. ruricoloides are indistinguishable. Females (n = 5). Total length 7.5-11.5; carapace 4.0-4.4 long, 2.9-3.2 wide; Fe I 2.7-3.1. For detailed description see Buchar (1997).
Distribution: This species was known only from Bhutan (type locality) and India. It is reported from India for the first time, and the new records extend the known range bỹ 1200 km to the northwest, indicating it may be found elsewhere. Schenkel, 1963 (Figs. 1d-e, 2e-f, 4d, 5) Trochosa ruricoloides Schenkel, 1963: 350, f. 202a-b (♂).
Remarks: Considering the published figures of T. aquatica Tanaka, 1985 from Japan, China, and Korea, it could be a junior synonym of T. ruricoloides.
Distribution: This species previously was known from mainland China and Taiwan (World Spider Catalog 2020), but our data indicate that the actual range is much broader. It is distributed across South-east Asia, occurring from eastern India to Papua New Guinea. This species is reported here for the first time from India, Indonesia, Malaysia, Papua New Guinea, and Thailand.  Diagnosis: Males of T. ruricola can be distinguished from other species with a tooth on the fang by the shape of the tegular apophysis and the cymbium: the anterior side of the tegular apophysis is sinuous in ventral view ( Fig. 3f-g) (v. straight or with a protrusion), the cymbial length/width ratio is 1.8 (v. ≥ 2). The ratio in T. hispanica is the same, but the tip of the embolus is coiled (Marusik & Nadolny 2020: fig. 5) (v. in T. ruricola the embolus is slightly curved toward the synembolus, 3e-g). Females of T. ruricola are most similar to those of T. hispanica. They are distinguished by the position of the epigynal hoods: in T. ruricola they are separated (Fig. 4f), while in T. hispanica they are touching each other (Marusik & Nadolny 2020: fig. 6).
Distribution: This species has a Transpalaearctic range and is known from Europe to Japan. According to the World Spider Catalog (2020), it has been introduced to North America, Cuba, Puerto Rico, and Bermuda.

Survey of former subspecies of Trochosa urbana
The catalog of Bonnet (1959) provides a list of six additional subspecies of T. urbana: Tarentula urbana molensis Strand, 1906, Tarentula urbana gofensis Strand, 1906, Tarentula urbana hova Strand, 1907, Tarentula urbana sansibarensis Strand, 1913, Lycosa urbana meruensis Lessert, 1926, and Lycosa urbana taurirtensis Schenkel, 1937 All subspecies of T. urbana were transferred to other genera by Roewer (1960). Roewer (1960) even described a new genus, Trochosippa, with a type species of Lycosa urbana meruensis Lessert, 1926. All of these subspecies, except Trochosa u. hova, were elevated to species level and placed in four different genera (World Spider Catalog 2020), some of which belong to different subfamilies: Geolycosa Montgomery, 1904, Hogna Simon, 1885, Pirata Sundevall, 1833, and Trochosippa Roewer, 1960 While searching for different figures of T. urbana subspecies, we noticed that the monotypic genus Caporiaccosa Roewer, 1960 was described based on a subadult female and is distributed within the range of T. urbana. Thus, we synonymize both the genus and species (comments below).
Below, we list former or current subspecies of Trochosa urbana in alphabetic order. In the heading we provide the species name in combination with the genus name as it appears in the World Spider Catalog (2020). (Strand, 1906 Remarks: This species was described based on a female from Gofa (Ethiopia) collected by the German ornithologist Oskar Neumann. He also collected the holotype male of Tarentula urbana molensis not far away (c. 40 miles) from the type locality of T. u. gofensis. Figures of the epigyne by Strand (1908b) and Lessert (1926) are somewhat different from one another, and Lessert (1926) doubted his identification.
Remarks: This species was described based on numerous specimens from Mt Meru (Tanzania) collected at elevations of 3000-3500 m. The female is 9 mm long (carapace 4.5 mm), and the male is 7 mm long (carapace 4 mm long). The description is accompanied by appropriate figures showing that this (sub)species is similar to T. urbana, the type of which was studied and illustrated by Lessert (1915). There is contradictory information about the type deposition. The World Spider Catalog (2020) stated that syntypes 2♂ and 3♀ are in the Senckenberg Museum, but Roewer (1960) mentioned 2♂ and 5♀ from Genf (= Geneva). Considering that Lycosa urbana meruensis is a type species of Trochosippa Roewer, 1960, we designate this genus a junior subjective synonym of Trochosa. It is likely that the seven species of Trochosippa Roewer, 1960 syn. n. known from Africa belong to Trochosa, while the species Trochosippa malayana (Doleschall, 1859) (Indonesia, Ambon) and T. obscura (Mello-Leitão, 1943) (Argentina) belong elsewhere.
Remarks: This species was described based on the holotype male from southern Ethiopia (Mole Tal). It has never been illustrated. The type locality of Tarentula urbana molensis is near to (40 miles) those of Tarentula urbana gofensis, a species described based on a female. According to Strand (1908b) the total length of the holotype is 7 mm, and the carapace 4.1 mm long. Strand (1908b) compared the details of his subspecies with T. urbana. Trochosa differs considerably from Pirata by the male palp and size (Pirata is smaller). Therefore, the transfer of this species to Pirata by Roewer (1955a) was unjustified. Considering that the holotype of this species is lost (Nentwig et al. 2019), the size of the species corresponds to those in T. urbana, the similarity of epigyne to T. urbana, and the fact that the type locality lies within the range of T. urbana, we synonymize these two names.
Remarks: This species was transferred to Hogna via the synonymy of Lycorma and Hogna. Roewer studied the type of this species in Berlin, probably before WWII. According to Nentwig et al. (2019), the types of this species are lost. Based on the small size of the carapace (4.2 mm), it does not belong in Hogna. Considering that the types are lost, the type locality lies within the range of Trochosa urbana, and the carapace length of the type is within the size range of T. urbana, we synonymize these two names.
Remarks: This species was described from Morocco, and the holotype is in Basel (Roewer 1960). It is unclear why Roewer (1955aRoewer ( , 1960 transferred this species to Pirata. It is much too large (16.5 mm) to belong to Pirata, and the epigyne has a distinct fovea, anterior hoods, and septum, which are lacking in Piratini (Zoicinae). Roewer (1955aRoewer ( , 1960 did not examine the holotype of this species but rather reproduced Schenkel's figure with some modifications (see Supplementary Fig. 4A-B). Based on the large size, the type locality (Morocco), and the shape of the epigyne, this species likely belongs in Hogna or Geolycosa. Because of its similarity to Hogna ferox (Lucas, 1838), we transfer this species to Hogna. It might be a synonym of Hogna ferox. (Strand, 1907 Remarks: This species was described based on the holotype female from the Nossibe (= Nosy Be, an island of Madagascar). The size of the holotype, according to Roewer (1960), is 12 mm, a 4 mm carapace and 8 mm abdomen. Such body proportions (abdomen twice as long as the carapace) are unknown in Lycosidae. Thus, Roewer's measurements seem to be incorrect, and the abdomen can be only a maximum of 6 mm in length. According to Nentwig et al. (2019), the holotype of this species was destroyed in Lübeck in 1942; however, according to Roewer (1960), he examined and illustrated the type stored in the Senckenberg Museum (No. 2253). Considering that the carapace length is the same as that of T. urbana, and the type locality is near the known range of T. urbana, we synonymize these two names.