Labahitha spiders (Arachnida: Araneae: Filistatidae) from islands in the Indian and Pacific Oceans

The genus Labahitha has hitherto comprised two species from peninsular Malaysia and Christmas Island (Australia). We here demonstrate that the genus is widespread in islands and territories across the Indian and Pacific Oceans, including the following species that have been previously assigned to other filistatid genera: Labahitha marginata (Kishida, 1936) comb. nov. (= Filistata bakeri Berland, 1938 syn. nov.), Labahitha garciai (Simon, 1892) comb. nov. (= Pritha heikkii Saaristo, 1978 syn. nov., = Pritha sechellana Benoit, 1978 syn. nov.), Labahitha nicobarensis (Tikader, 1977) comb. nov., Labahitha littoralis (Roewer, 1938) comb. nov., Labahitha insularis (Thorell, 1891) comb. nov., Labahitha sundaica (Kulczyński, 1908) comb. nov. (all transferred from Pritha, the latter three provisionally, pending re-examination of the type material); Labahitha fuscata (Nakatsudi, 1943) comb. nov. and Labahitha ryukyuensis (Ono, 2013) comb. nov. (both transferred from Tricalamus). Many of these species have been collected in synanthropic settings and from disparate islands thousands of kilometers apart. This suggests either high dispersal capabilities or, more likely, human-mediated introductions. At least L. marginata has been introduced to continental America. Two new species of Labahitha are described: Labahitha platnicki sp. nov. from New Caledonia and the Bismarck Islands and Labahitha incerta sp. nov. from Queensland, Australia. The male of Labahitha gibsonhilli (Savory, 1943) is reported for the first time. Wandella loloata sp. nov. is described from Papua New Guinea, representing the first record of this genus outside Australia. Pritha hasselti (Simon, 1906) from Indonesia is shown to be a Filistatinae, and thus the species is provisionally transferred back to Filistata.


Introduction
(vs anterior margin of the carapace straight, with acute clypeus in males), (2) paraembolic lamina, which has a ragged margin (Figs 5, 12A-D) (vs entire margin), and may be divided into two parts (Figs 5,15,20) or reduced to a small proximal keel almost completely fused to the tegulum (Figs 12A-D, 25) (vs a single free-ending, unfused part); in addition, they usually have an apical macroseta on the ventro-retrolateral face of metatarsal I (vs macroseta usually absent; present in at least W. murrayensis Gray, 1994). Females are distinguished from Wandella and Yardiella by the less contrasting colouration pattern, with weak or absent submarginal bands, leg rings and chevron pattern on the abdomen (Figs 2-3) (vs submarginal bands, chevron and leg rings well-marked). The female genitalia is variable, with paired (Figs 6,12) or unpaired (Fig. 19) receptacles; the median receptacles may be well-developed (Fig. 12) or reduced (Fig. 6).

Relationships
Labahitha forms a clade with Wandella and Yardiella on the basis of shared possession of a paraembolic lamina with micro-teeth (Figs 5, 12A-D) and micro-teeth in the male clypeus ( Fig. 23F) (see also Gray 1994;Magalhaes 2016;Zonstein et al. 2017).

Tentative transfers
The following three species are known only from the females mentioned in the original descriptions, which include poor figures, if any at all. Based on the textual description and their distribution, they are here provisionally transferred to Labahitha, taking into account that their current placement in the mainly Eurasian genus Pritha (proposed by Lehtinen 1967: 260) is poorly justified. Examination of their type material should be carried out to confirm the generic placement and clarify their identities, as it is not unlikely that they are synonyms of the other species treated here. This is especially important in the case of Filistata insularis Thorell, 1891, which has nomenclatorial priority over any of the names treated in this paper.
(2) Labahitha insularis (Thorell, 1891) comb. nov. (F. insularis Thorell, 1891: 17. Female (subadult?) from India, Car Nicobar, repository unknown, not examined). Lehtinen (1967) mentions that the types are deposited in Naturhistorisches Museum Wien, Austria, but the only specimens identified as Filistata insularis in this collection come from Sumatra and were collected in 1938 (C. Hörweg, pers. com.), and thus cannot possibly be the types. The description mentions a dark spider, consistent with some species of this genus; the type locality is the same as that of L. nicobarensis (Tikader, 1977) comb. nov. and it is not unlikely the two species are synonyms ( Fig. 2A).

Distribution
The genus is mainly distributed in Oceania and adjacent areas (Fig. 1). Species occur in a wide range spanning the Seychelles, Malaysia, Indonesia, Australia, New Guinea and several islands in the Pacific Ocean; records from the American continent likely represent human-mediated introduction. Material examined by us from India, Sri Lanka, China, Laos, Cambodia, and Thailand belongs to other Prithinae genera, thus we suspect that Labahitha is not diverse in continental Asia.

Notes
Although neither Ono (2011) nor we have examined the type specimens, we have seen plenty of specimens from the type locality and neighbouring islands that tally with the original descriptions of this species. Ono (2013) recorded this species in the Ogasawara Islands in Japan. Gray (1995: fig. 15) was the first to figure the male of this species. See Ono (2011) for a discussion regarding the correct authorship of this species and the priority of Nakatsudi's (1943) name over Kishida's (1947).

Diagnosis
The male is similar to that of Labahitha gibsonhilli in the sharp, triangular apex of the paraembolic lamina and the gently curved distal portion of the sperm duct. It differs by the shorter, more stout palpal bulb and the relatively larger paraembolic lamina (Fig. 5B, J-L) (vs palpal bulb longer and slender, with smaller paraembolic lamina in L. gibsonhilli). Females can be distinguished from all congeners by the very small median receptacles, which are reduced to a bump with pores, placed posteriorly to the larger, oval lateral receptacles (Fig. 6). The deep brown and uniform colouration is also characteristic (Fig. 4), although also present in L. ryukyuensis (Ono 2013 Komatsu (1936).  (Nakatsudi, 1943) (Nakatsudi, 1943)

Description
Male (from Palau, Koror Island, Entomology Laboratory, JBJB) Colouration (in ethanol). Carapace grey-brown, with darker irregular patterning in the median postocular area and the weakly defined mid-lateral and lateral marginal bands. Chelicerae, labium, endites and sternum brown. Legs with dark grey-brown pigmentation on femora, paler on tibiae and reduced to absent on remaining segments. Abdomen colour brownish grey with and several indistinct chevrons, but patterning poorly preserved.
Habitus. Anterior margin of carapace subrounded, sclerotized but apical part with an unsclerotized curved edge. Sternum subrounded, posteriorly bluntly pointed, a pair of posterior sigilla present.  (Nakatsudi, 1943) comb. nov., endogyne, dorsal, lactic acid cleared. A. Brunei, Tutong (JK 110416.1907 PalP (Fig. 5). Cymbium horseshoe shaped; bulb enlarged with a median constriction, sperm duct Nshaped with a single coil; tegulum with a large prolateral excavation, its surface adorned with strongly spined short ridges (each typically bispinate); large crest-like paraembolic process with a fimbriated dorsal margin bent more or less prolaterally, and narrowing distally to end in an extension above the embolus; comb-like microtooth arrays present on paraembolic process; embolus slender, slightly curved.
Female (from Palau, Koror Island, Entomology Laboratory, JBJB) Colouration. Carapace similar to male, but anterior margin of carapace pigmented. Abdomen greyish brown with a longitudinal pale patch mid-dorsally and several indistinct pale chevrons laterally. One pair of posterior sternal sigilla present.  (Fig. 6). Median lobes very small with a few clustered pores; lateral lobes large, ovoid in side view, with scattered pores.

Variation
Male genitalia is similar across islands (Fig. 5). The shape of the lateral receptacles varies from rounded to oval, and the pores in the lateral receptacles may be concentrated in the ectal side or dispersed throughout the receptacle (Fig. 6).

Natural history
Specimens have been collected in habitats as diverse as wooden wall crevices, in coconut litter, in crevices on buildings, by beating trees in Pandanus forest, in banana leaves, in web on tree trunk, in webs in crevices on tree bark, in coastal cliffs and in cave entrance rock cracks.

Notes
The original description of Simon (1892) mentions at least two females, but the vial with the syntypes contains one female, a male with both palps attached, and a loose male palp. We did not examine the holotype of F. pulchella, but Lehtinen (1967: fig. 23) provided an excellent illustration of the palp of the holotype and the type localities of both species are nearby (Simon 1892(Simon , 1893. This indicates that the synonymy is correct. We did not examine the type material of Pritha heikkii, but the illustrations in the original description and the proximity between type localities indicate that the synonymy proposed by Saaristo (2010) is correct.

Diagnosis
Males are similar to those of L. oonopiformis, L. ryukyuensis and L. nicobarensis by the teardrop-shaped bulb with a keel-shaped paraembolic lamina. They differ from L. oonopiformis by the shorter embolus (vs longer), from L. ryukyuensis by the more globose base of the bulb (vs base of the bulb more tubular) and from L. nicobarensis by the longer, more curved embolus (vs embolus short and straight) (Fig. 10).
Females are more similar to those of L. oonopiformis and L. gibsonhilli by the large membranous base of the receptacles and well-developed median receptacle; they differ from L. oonopiformis by the median receptacles subequal in size to the laterals (vs median receptacles notably enlarged) and from L. gibsonhilli by the globose median receptacles with evenly scattered pores (vs median receptacles with pores restricted to medial face) (Fig. 11).

Description
Male (from Kapit, Sarawak, Malaysia, AMNH IFM-0936) Colouration. Carapace light brown, with brown median pattern and lateral borders and faint submarginal bands. Chelicerae orange brown. Labium, endites and sternum cream. Legs cream, except for orange brown femur I and light brown tarsus I, without rings. Abdomen light brown; dorsum with two patches of white setae, one large and anterior and one smaller and posterior; venter cream.
State of the specimen: lost many abdomen setae, many legs, and abdomen separated from the cephalothorax; left palp dissected.
Female (from Dairy Farm Nature Park, Singapore, JK 1308230003) Colouration. As in male, except where noted. Carapace cream, sparsely stippled with brown. Legs yellow, except for white coxae. Abdomen cream, dorsum with sparse white setae on the abdomen.
Habitus. Anterior margin of the carapace unmodified. Eye apodemes present. Sternum suboval, with a well-marked pair of posterior sigilla.

Natural history
The species has been collected in synanthropic settings in Singapore and in the Seychelles ("very common in minute crevices of the walls of the Reef Hotel"; Saaristo 1978: 100).

Notes
This hitherto undescribed species, known from a single male, shares some characters with Wandella, such as a contrasting colouration (Fig. 14A) and the lack of a true macroseta in metatarsus I (Fig. 14D).
On the other hand it presents characters typical of Labahitha, such as a subrounded clypeus (Fig. 14E) (vs straight or sharp in Wandella), and the paraembolic lamina divided in two parts and apparently with a ragged margin (Fig. 15B, D). We decided to allocate this species in Labahitha provisionally until more data is collected, but the possibility that it belongs in the mainly Australian genus Wandella cannot be excluded.

Diagnosis
The male is most similar to L. marginata in the paraembolic lamina divided in two parts, with a rounded apex near the embolus; it differs by the less obvious fimbriations in the paraembolic lamina and the bulb without a median constriction (Fig. 15) (vs fimbriations more prominent, bulb with a clear median constriction).

Etymology
The name is an adjective meaning 'uncertain' and refers to the tentative placement of this species in this genus.

Description
Male holotype (from Nipping Gully, Queensland, Australia, QM S.78659) Colouration. Carapace cream with dark brown median pattern and clypeal markings, light brown median area, and slightly dark submarginal bands. Chelicerae, labium and endites cream, with a small light brown patch. Sternum cream with diffuse light brown patches along the border. Legs cream, with incomplete light brown rings in the base and apex of the femora, tibiae and metatarsi (except for leg I, with barely visible markings in femur and tibia only). Abdomen light brown, clothed with whitish setae, with a cream, oval, elongate patch in the first half and four transverse cream patches posteriorly; venter cream.
PalP (Fig. 15). Cymbium horseshoe shaped, bulb suddenly narrowing towards embolus, sperm duct N-shaped, with a single coil, prolateral excavation very large and high, apparently with comb-like arrays of micro-teeth, paraembolic lamina free, prolaterally curved, with rounded apex, ending close to embolus apex, embolus short.
State of the specimen: has lost some setae from the legs and abdomen, left palp dissected, right legs I and II separated from the body at the patella-tibia joint.

Female
Unknown.

Distribution
Known only from the type locality in Australia, Queensland (Fig. 1B).

Notes
Zonstein & Marusik (2019: 88) considered the type locality of Filistata marginata as "Japan, Honshu, Akita Prefecture, Senboku". This was apparently a mistake: H. Ono (in litt.) informed us that the original text in Komatsu (1936) reads "Distribution: Taiwan. In August of 1935, Mr. Kyukichi Kishida found this spider in a hotel at Tainan where he stayed.
[…] Afterwards, Mr. Izumi Kayashima announced that this spider is very common in houses in Taihoku." Taihoku was an administrative region of the island while it was under Japanese rule, and corresponds to modern-day Taipei. Apparently, the old Japanese kanji 'tai' was misread by Zonstein & Marusik as 'sen', causing the confusion. Thus, the correct type locality of this species is Taiwan, Tainan. This species is cited in books on Taiwanese spiders (Kayashima 1943;Lee 1966), supporting this view. H. Ono (in litt.) kindly informed us that "after the death of Dr. Kishida in 1968, no specimen has been found, which he presumably held", and thus we may consider that the type specimens of F. marginata have been lost. We have examined specimens from Taiwan, Pingtung  (50 km to the south of Tainan) whose habitus morphology fits well with Komatsu's (1936) illustration, thus ascertaining the identity of this species. Since this is the only filistatid known to occur in Taiwan and the original illustrations match our specimens, this name can be objectively applied to this species, precluding the need for a neotype (see International Code of Zoological Nomenclature, article 75.1). Wang (1987: fig. 1i) was the first to illustrate the female genitalia of this species, showing the single pair of receptacles with a trilobulate structure; however, the locality of the illustrated specimens is unclear.
The types of Filistata bakeri are in a bad state of preservation, and apparently have dried at some point. The palp of the holotype is deformed, especially the sperm duct; the colours are faded, but Berland's illustration match the colour pattern of fresh specimens examined by us. Berland (1938) mentions a male paratype,   but the two vials we saw contain only one male, six females and one immature. Records of this species by Berland (1942) from Austral Islands could not be checked and are not included in the map; his report of this species for Christmas Island probably refers to L. gibsonhilli. Lehtinen (1967) included two species in his Pritha bakeri group: Filistata bakeri from Vanuatu and a species from Taiwan he thought to be new (his revision did not account for F. marginata). We here show that specimens from both islands are conspecific; specimens from Taiwan (Fig. 19F) and Vanuatu (Fig. 19C) show similar female genitalia, and the differences observed in males (Fig. 18D, G) are due to the bad preservation of the holotype of Filistata bakeri.

Diagnosis
The male is most similar to L. incerta sp. nov. in the paraembolic lamina divided in two parts, with a rounded apex near the embolus; it differs by strong fimbriations in the paraembolic lamina and the bulb with a median constriction (Fig. 18) (vs smaller fimbriations, bulb without clear median constriction). Females are unique among all filistatids by the genitalia with a single pair of stalked, trilobate receptacles (Fig. 19).    Habitus. Anterior margin of the carapace subrounded, with unsclerotized tip. Eye apodemes present. Sternum subrounded, sigilla not visible.
State of the specimen: good, lacking some setae; left palp and right leg I dissected.
Female (from Parque Nacional La Cangreja, San Jose, Costa Rica, INBIO 79582) Colouration. As in male, except for lack of brown markings on carapace and chevron in the abdomen.

Variation
Female seminal receptacles may be fully sclerotized or hyaline (Fig. 19)

Notes
We have not examined the type specimens, but the teardrop-shaped bulb with a keel-like paraembolic lamina figured in the original description (Tikader 1977: fig. 1C) clearly indicates this species belongs to Labahitha. We provided a tentative diagnosis and consider this species as valid, but re-examination of the type material would be desirable to confirm its validity. It should be noted that there is an older available name based on specimens also from the Nicobar islands, Filistata insularis Thorell, 1891, which we were unable to examine; it is not unlikely that they could be synonyms.

Diagnosis
Males are similar to those of L. oonopiformis, L. ryukyuensis and L. garciai by the teardrop-shaped bulb with a keel-shaped paraembolic lamina. They differ from all these species by the shorter, more straight and robust embolus (Tikader 1977: fig. 1c). The female genitalia has never been illustrated and thus we are unable to provide a diagnosis for the female at this time. Tikader (1977).

Notes
This is the type species of the genus. The holotype from Parit Buntar is lost, and a neotype was designated by Zonstein et al. (2017) based on specimens from ca 240 km to the southeast of Parit Buntar. We have now examined specimens from Penang, ca 50 km to the northwest of Parit Buntar, providing additional evidence that Zonstein et al. (2017) interpreted Bristowe's description correctly. We provide the first SEM images of the male of this species (Fig. 25A-B).

Diagnosis
Males are similar to those of L. garciai, L. ryukyuensis and L. nicobarensis by the teardrop-shaped bulb with a keel-shaped paraembolic lamina. They differ from L. garciai by the longer embolus (vs shorter), from L. ryukyuensis by the more globose base of the bulb (vs base of the bulb more tubular) and from L. nicobarensis by the longer, more curved embolus (vs embolus short and straight) (Fig. 25A). Females are more similar to those of L. garciai and L. gibsonhilli by the large membranous base of the receptacles and well-developed median receptacle; they differ from L. garciai by the median receptacles larger in size than the laterals (vs median and lateral receptacles subequal in size) and from L. gibsonhilli by the globose median receptacles with evenly scattered pores (vs median receptacles with pores restricted to medial face) (Fig. 25C). (Bristowe, 1938)

Variation
The shape and size of the median receptacles of females varies ( Fig. 25C; Zonstein et al. 2017: fig. 5).

Diagnosis
The male is unknown. Females can be diagnosed from congeners by the colouration pattern (carapace deep brown with lighter margins and abdomen dark with light grey transversal markings; Fig. 26A) (vs colouration usually uniform, especially in the abdomen) and by the genitalia: the median receptacles are sclerotized, with many pores packed together, and have a median constriction (Fig. 27) (vs median receptacles without a median constriction).

Etymology
The species is named after the eminent arachnologist Norman Platnick, who collected part of the type series and has recently passed away in an untimely manner, in memory of his innumerable contributions to arachnology and systematics.

Other material examined
None.

Unknown.
Female (holotype from Koné, Foué, Nord, New Caledonia, AMNH IFM-0890) Colouration. Carapace reddish brown on median area, lateral borders yellow, with dark brown V-shaped median pattern, submarginal bands and clypeal markings. Chelicerae, labium and endites orange brown. Sternum orange brown stippled with dark brown. Legs yellowish cream, becoming orange brown distally, with longitudinal brown bands on femora, tibiae and metatarsi. Abdomen dark brown, dorsum with five grayish cream chevron-like markings posteriorly, venter greyish cream with brown pigment around spinnerets.
state of tHe sPeCiMen. Genitalia dissected, right legs I-II and left legs III-IV missing from patella.

Natural history
Label data indicate specimens have been collected under bark of Alstonia R.Br. trees in the Bismarck Islands, and in coastal cliffs and under bark in coastal trees in New Caledonia.

Distribution
New Caledonia and Bismarck Islands, Papua New Guinea (Fig. 1B).

Notes
The structure of the male palp (drop-shaped bulb with a keel-shaped paraembolic lamina; Ono 2013: figs 1-3) and somatic morphology indicate this species belongs in Labahitha and is a close relative of L. oonopiformis, L. garciai and L. nicobarensis, thus we propose the new combination.

Diagnosis
Males are similar to those of L. oonopiformis, L. garciai and L. nicobarensis by the teardrop-shaped bulb with a keel-shaped paraembolic lamina. They differ by the more slender palpal tibia, the less globose base of the bulb (  (Nakatsudi, 1943) comb. nov. from Brunei, Tutong (JK 110416.1907). D-E. Labahitha marginata (Kishida, 1936)  smaller than the laterals (Ono 2013: fig. 9) (vs median receptacles subequal in size or larger than the laterals).

Natural history
Ono (2013) reports that specimens were collected in delicate webs with a central tubular retreat in inclined ground in the edge of a forest. One egg-sac had approximately 40 eggs.

Distribution
Australia, here newly recorded in Papua New Guinea.

Diagnosis
This species is most similar to Wandella grayi Magalhaes, 2016 due to the tapering, long palpal bulb with erect paraembolic lamina in the male, and the female with columnar median receptacles attached to the rounded lateral receptacles via a stalk. Males can be diagnosed by the shorter, less curved embolus and the larger paraembolic lamina with an acute tip ( Fig. 30A-B) (vs embolus longer and more curved, and smaller, curved paraembolic lamina in W. grayi). Females can be diagnosed by the digitiform median receptacles, widest at the base, and by the stalk of the lateral receptacles connecting apically ( Fig. 30E-F) (vs median receptacles with even width throughout, and stalks of the lateral receptacles connecting basally).

Etymology
The name is a noun in apposition and is taken from the type locality.

Paratype
PAPUA NEW GUINEA • 1 ♀; same collecting data as for holotype; AM KS32717.

Description
Male holotype (AM KS32718) Colouration. Carapace mostly pale brownish-cream, lateral margins translucent; postocular area with a middorsal grey patterned triangular patch that narrows toward the foveal area around which it divides; lateral carapace with three grey pigment patches forming incomplete lateral bands; a pair of larger anterolateral grey patches flank the mostly cream-coloured clypeus. Chelicerae, labium, endites and sternum cream; abdomen pale creamy brown. Legs pale creamy brown, with dark grey partial ringlike patches restricted to proximal and distal femora and tibiae. Abdomen with well-defined brownish patterning on a cream ground. Frontal abdomen dark brown connecting dorsally to a broad middorsal brown patch with a paler central area; from this arise 5-6 pairs of brownish-grey dorsal chevrons reducing in size posteriorly; these are flanked below by a set of 4-5 smaller lateral chevrons on each side. Venter pale brown.
Notes Lehtinen (1967) claims to have examined a type specimen of F. hasselti deposited in the Zoölogische Museum der Universiteit van Amsterdam, Netherlands. This is unusual since type specimens of Simon are usually deposited in the MNHN. The specimen we examined is correctly labelled and fits with the description given by Simon (1906) and we have no reason to believe it is not the holotype. We examined photos of this specimen as its distribution suggested it might belong to Labahitha, but it clearly belongs in the Filistatinae, not in the Prithinae, due to the tarsal macrosetae, calamistrum in a crest and with a median gap, and the deep thoracic fovea (Fig. 31). We did not perform a dissection to examine and illustrate its genitalia. The subfamily Filistatinae is otherwise not represented in the Oriental region, and thus the generic affinity of Filistata hasselti is unclear. The only other filistatine specimen known to us from the Oriental region is a juvenile from Malaysia, Selangor, Kepong, Aug. 1947, R. Traub leg. deposited in the AMNH, whose generic affinity is also unclear. We allocate F. hasselti in Filistata to restore the original combination, and because it shares a calamistral gap with Filistata and Zaitunia Lehtinen, 1967.

Discussion
Labahitha (replacement name for Mystes) was originally described in another family, Pholcidae C.L. Koch, 1850, an error that was not corrected until very recently (Huber et al. 2014). For that reason, it was completely ignored for many years by anyone working on filistatids. We here radically change the composition of Labahitha, known in 2017 from a single species. This is done mainly by revising already described species that had been incorrectly allocated in other genera, mainly Pritha and Tricalamus. The genus is now morphologically heterogeneous, but some characters are shared by all species: a rounded male clypeus (vs sharp in the related genera Wandella and Yardiella), a ragged margin of the paraembolic lamina, legs with faint rings or uniform colouration, and uniform colour of the abdomen. Males also typically show a single macroseta at the apex of metatarsus I, but this macroseta is absent in L. incerta sp. nov. -one of the reasons why it is allocated provisionally in the genus. We here present additional morphological data on Labahitha using SEM images. Labahitha marginata shows the same micro-projections in the clypeus (Fig. 23F) as do Labahitha garciai (Zonstein et al. 2017) and Wandella (Magalhaes 2016), corroborating this as a synapomorphy uniting these genera (it is also likely present in Yardiella; ILFM, unpublished). The spinnerets (Fig. 24) are also very similar in structure to those of Wandella, as in the lack of aciniform gland spigots (a character shared with other genera, such as Pritha). Finally, we present SEM images of the male palp of L. oonopiformis (Fig. 25A-B), confirming the existence of micro-teeth and a ragged margin of the paraembolic lamina.
The most curious aspect of the genus is its geographic distribution (Fig. 1). The closest relatives, Wandella and Yardiella, are largely restricted to Australia, with a single record from Papua New Guinea reported here. Labahitha is distributed from Malaysia to Australia, but most commonly occurs in islands in the Indian and Pacific Oceans. Some species seem to be narrow endemics (such as L. gibsonhilli in Christmas Island, and L. ryukyuensis in Okinawa), but at least L. marginata, L. fuscata and L. garciai have extremely wide distributions spanning several islands and thousands of kilometres. This was noted by Berland (1942: 24) regarding L. marginata (under Filistata bakeri): "Whether this dispersal is natural, or due to artificial introduction by man is difficult to decide; it seems impossible to see direct relations between so widely separated points in the Pacific." Because some species are endemic to volcanic islands, it seems that at least some natural overwater dispersal is achieved by these spiders. On the other hand, in some areas their presence is almost surely due to artificial introduction, such as in Brazil. Studying the poorly known species (such as L. sundaica, L. littoralis, L. nicobarensis and L. insularis) will be important to establish the limits of distribution of species in this genus. A phylogeny of the genus might reveal the timing and pattern of their arrival to the islands, but this will be a Herculean undertaking requiring extensive sampling in numerous disparate locations.