Article. urn:lsid:zoobank.org:pub:0e34f9de-b76c d0-5a08a1f7c534. The lineages of Stylocellidae (Arachnida: Opiliones: Cyphophthalmi)

Transcription

1 Zootaxa 3595: 1 34 (2012) Copyright 2012 Magnolia Press Article urn:lsid:zoobank.org:pub:0e34f9de-b76c d0-5a08a1f7c534 The lineages of Stylocellidae (Arachnida: Opiliones: Cyphophthalmi) ISSN (print edition) ZOOTAXA ISSN (online edition) RONALD M. CLOUSE American Museum of Natural History, Division of Invertebrate Zoology, Central Park West at 79th St., New York City, NY 10024, USA. rclouse@amnh.org Abstract The taxonomy of the Southeast Asian mite harvestman family Stylocellidae is updated in light of new molecular and morphometric phylogenies, examinations of type specimens, and a new species from Thailand. A new genus, Giribetia gen. nov., is erected, and Fangensis insulanus Schwendinger & Giribet, 2005, recombined in it as Giribetia insulana new comb. All species in the genus Stylocellus have been recombined in Miopsalis and Leptopsalis, except for the type species, S. sumatranus Westwood, 1874, and a new species, S. lornei, sp. nov., described here. The new recombinations of former Stylocellus species are as follows: Leptopsalis pangrango (Shear, 1993), new comb., Leptopsalis sedgwicki (Shear, 1979), new comb., Leptopsalis laevichelis (Roewer, 1942), new comb., Miopsalis globosa (Schwendinger & Giribet, 2004), new comb., Miopsalis kinabalu (Shear, 1993), new comb., Miopsalis leakeyi (Shear, 1993), new comb., Miopsalis mulu (Shear, 1993), new comb., and Miopsalis pocockii (Hansen & Sørensen, 1904), new comb. Stylocellus spinifrons Roewer, 1942 is now designated as nomen dubium, as the sole type specimen has been found to be a juvenile. Two new subfamilies are proposed, each with two genera: Fangensinae subfam. nov., containing Fangensis and Giribetia, and Leptopsalinae subfam. nov., containing Leptopsalis and Miopsalis. The subfamily sensu strictu Stylocellinae contains the remaining two genera: Stylocellus and Meghalaya Key words: biogeography, Borneo, harvestman, Sibumasu, Southeast Asia, Sundaland Introduction In 1874, when British entomologist John Obadiah Westwood described Stylocellus sumatranus Westwood (Fig. 1), there was no hint of the great number and varied forms of Cyphophthalmi still hidden on Sumatra and throughout Southeast Asia, nor any suggestion of the family s importance to understanding the region s history. Westwood s only mention of other Cyphophthalmi in Southeast Asia is a Javanese specimen he saw in the British Museum, which he contrasts with S. sumatranus by color only. (The Javanese specimen was reddish brown, unlike the purplish black of S. sumatranus, he says.) Appearing on the 200 th of 205 pages of taxonomic descriptions and being illustrated by small sketches on the 37 th plate of 40, the description of S. sumatranus was based on a lone male who was, in fact, deformed: sternites 3 and 4 as well as tergites V and VI were fused on the right side (Westwood 1874). Since Westwood, another 33 species in the family have been described, and due to a recent focus on Cyphophthalmi collecting worldwide (Giribet 2000), scores of new stylocellid species await description. The analysis of DNA sequence and morphometric data from newly collected specimens has led to the understanding that the family includes all Cyphophthalmi found from the Eastern Himalayas to New Guinea and out to Palawan and Mindanao (Clouse et al. 2011; Clouse et al. 2009; Clouse & Giribet 2010). Stylocellidae is also considered the most ancient living family of animals in Southeast Asia (Stelbrink et al. in press; Lohman et al. 2011), having likely arrived on an ancient fragment of the Gondwanan coastline which today underlies the Thai- Malay Peninsula (Clouse & Giribet 2010). Following the most recent phylogenetic analysis of Cyphophthalmi, it was placed with the Laurasian family Sironidae in a new infraorder, Boreophthalmi (Giribet et al. 2012). However, no adequate definition exists for the genera containing most of the family s diversity, and no subfamilial taxonomy exists to record identifications that at least limit the number of possible genera to which any new specimen might belong. Accepted by A. Perez-Gonzalez: 29 Oct. 2012; published: 21 Dec

2 Our concept of Stylocellidae as a whole now appears to be stable, and the familial synapomorphies postulated by Giribet (2002) have not been contradicted by the many new collections subsequently examined. Still, the history of the family is worth recapping, since it leaves its mark as considerable confusion over the family s internal taxonomy and how to recognize the various lineages. Stylocellus sumatranus was described as a member of the family Cyphophthalmidae, which was in an arachnid order Westwood had erected earlier, Adelarthrosomata (Westwood 1874). Less than a decade later, Thorell (1882) described another Sumatran species, Leptopsalis beccarii Thorell, erecting a second monotypic genus in the family. Thorell (1891a) later synonymized L. beccarii with Stylocellus sumatranus but in the same year (Thorell 1891b) described another new species and monotypic genus, Miopsalis pulicaria Thorell. Hansen and Sørensen (1904) subsequently described five additional species in Stylocellus Westwood two more from Sumatra and one each from Java, Borneo, and Sulawesi and agreed with Thorell s synonymy of Leptopsalis beccarii and Stylocellus sumatranus. With these they combined a species described in 1897 (Miopsalis lionota (Pocock)), Thorell s Miopsalis pulicaria (which they noted was weakly described), and a new tropical African species, Ogovea grossa (Hansen & Sørensen), into a new subfamily of the Sironidae named Stylocellini [sic]. Shear (1979) agreed with Hansen and Sørensen (1904) that the description of Miopsalis pulicaria was inadequate while describing a new species from the same locality (Stylocellus sedgwicki Shear), and later Shear (1980) raised their subfamily to family and superfamily, moving the African Ogovea Roewer into a separate family and superfamily (Ogoveidae, Ogoveoidea). Rambla (1991) used Shear s taxonomic scheme when describing the next new species in the family, Stylocellus silhavyi Rambla, and Shear (1993) soon described 11 new species in Stylocellus and discussed continuing uncertainty about the identity of Miopsalis pulicaria. This picture of Stylocellidae was captured in Giribet s (2000) catalog of Cyphophthalmi: the continued synonymization of Stylocellus and Leptopsalis Thorell, designation of Miopsalis pulicaria as nomen dubium, and acceptance of Fangensis Rambla, 1994, as a member of the Laurasian family Sironidae. Soon thereafter, however, Leptopsalis beccarii and Stylocellus sumatranus were recognized as distinct, and Fangensis was transferred to Stylocellidae (Giribet 2002; see also Giribet & Boyer 2002; Schwendinger & Giribet 2005). This settling of Stylocellidae lasted only briefly, for a new genus of Cyphophthalmi from northeastern India, Meghalaya Giribet, Sharma & Bastawade, was soon described, noted as resembling Stylocellidae, but left unassigned to any family (Giribet et al. 2007). Meghalaya corresponded to an obscure mention by D. Bastawade (1992) of a cyphophthalmid species from India and cited as Sironoidea sp. by Giribet (2000), but molecular phyloegenetic analyses clearly place it within Stylocellidae (Clouse & Giribet 2010). The analysis of molecular data from several undescribed species (Clouse & Giribet 2010), morphometric data from those species and type specimens (Clouse et al. 2009; de Bivort et al. 2010), and the examination of longunseen types from the Museo Civico di Storia Naturale "Giacomo Doria", Genoa, Italy (MCSN) (Clouse & Giribet in press), has allowed resurrection of the long-synonymized Leptopsalis and confirmation of the validity of the nomem dubium Miopsalis, as well as the recombination of 17 of the 27 Stylocellus species into these genera and other key insights into the family s internal lineages. The genus Fangensis is the earliest branching lineage in the family but, in its former combination, not always recovered as monophyletic; the large southernmost species, Fangensis insulanus Schwendinger & Giribet, 2005, recombined here in Giribetia, gen. nov., often places as sister to the non-fangensis stylocellids (i.e., the remainder of the family) (Clouse & Giribet 2010) or in some other position away from the rest of Fangensis (Schwendinger & Giribet 2005). A large species recently found on Phayam Island (off the West coast of Thailand), which is described here, is often recovered as sister to Stylocellus sumatranus, and these two, in turn, are recovered as sister to Meghalaya (Clouse et al. 2009). Finally, the genera Miopsalis and Leptopsalis are nearly always recovered as sister groups (Giribet et al. 2012; Clouse & Giribet 2010; Clouse & Giribet 2007), but although the latter genus contains most of the family s diversity, support for its internal relationships is currently too weak to justify further subdivision. Thus the current phylogenetic hypotheses for the genera of Stylocellidae are as follows: ((Fangensis, Giribetia), ((Stylocellus, Meghalaya), (Miopsalis, Leptopsalis))) (Fig. 2, A) or (Fangensis, (Giribetia, ((Stylocellus, Meghalaya), (Miopsalis, Leptopsalis)))) (Fig. 2, B). The recent phylogenetic hypotheses of Stylocellidae have led to a re-examination of the external morphology to see if generic identities consistently recovered in molecular phylogenies are diagnosable without sequencing. This has been a process of assessing types for characters previously overloooked, rethinking certain character homologies, and examining shape characteristics. The presence of anal gland pores was often not recorded in species descriptions, even in the modern era, and when they were noticed, they and the lack of eyes led Fangensis leclerci Rambla, 1994, to originally be placed in a different family (Sironidae). Shear (1980) had listed a lack of anal gland pores and the presence of eyes among stylocellid synapomorphies (the blind Miopsalis pulicaria not 2 Zootaxa Magnolia Press CLOUSE

3 considered in his analysis). Eyes became a key generic diagnostic feature within the family when Miopsalis received more consideration (Shear 1993), and when Miopsalis and the blind Fangensis were formally recombined into the family (Giribet 2002); but eyes currently appear to be rather variable and prone to loss in small or troglomorphic species throughout Stylocellidae (Clouse et al. 2009). By contrast, the discussion of anal gland pores has expanded recently, having been first recorded for Stylocellus sumatranus 130 years after its formal description (Schwendinger et al. 2004), and it is clear they are a critical element in generic diagnosis. Rambla s organ (Fig. 3) was named and discussed most thoroughly with the description of new Fangensis species (Schwendinger & Giribet 2005), where its homology to a similar structure in Miopsalis silhavyi (Rambla, 1991) was implied; later it was also suggested as homologous to the large depression in the fourth tarsus of Meghalaya males (Giribet et al. 2007). Finally, the second ventral cheliceral process, an extension of the dorsal crest and lateral ridge that lies anterior to the first ventral cheliceral process, is unique in stylocellids and was postulated early as a key to the genera, given its variability in the family (Giribet 2002; Giribet & Boyer 2002). As phylogenetic hypotheses for the family expanded, this was thrown into doubt as species with the second process were recovered as distantly related (Clouse & Giribet 2007). Other cheliceral characters, like the extent of second-article granulations and the relative size of the third article, have been complicated by the scattered phylogenetic placement of species with attenuate chelicerae, which all have the same narrow, smooth appearance (Clouse et al. 2009). Noticing that most stylocellid species are simple ovoid forms that lack any obviously scorable traits, but also that they come in a wide variety of shapes, morphometric data have been used in phylogenetic analyses (Clouse et al. 2009; de Bivort et al. 2010). The result has been the recovery of several relationships also seen in molecular phylogenies, or ones which are consistent with biogeographic hypotheses. For example, morphometric data alone have consistently recovered Fangensis as sister to the remainder of the family, and have even placed Giribetia insulana (Schwendinger & Giribet) new comb., as either sister to Fangensis or, more remarkably, as sister to the non-fangensis stylocellids. Morphometric phylogenies also tend to separate most Leptopsalis and Miopsalis species into separate clades, and recover all but one Meghalaya species in a single clade. These same analyses have also consistently recovered a close relationship between Stylocellus sumatranus and the Phayam Island species, the latter of which has never been recovered with Fangensis or Miopsalis species in molecular phylogenies. Most notably, morphometric data have consistently recovered a close relationship among the Sulawesi species, which is consistent with the history of the island, and even included Leptopsalis novaguinea (Clouse & Giribet, 2007), a relationship first detected by molecular data. Thus, I have begun to provide preliminary identifications of new specimens based on their overall shape, which is often corroborated by molecular data, and I attempt here to describe the key shape features (and other clues) I am using for generic determinations. The most important taxonomic use of shape features has been in the connection of one of the lineages in molecular phylogenies to the sole, female type specimen of Miopsalis pulicaria, which I only recently examined (Clouse & Giribet in press); it is a female, thus eliminating the chance of scoring it for anal gland pores, Rambla s organ, or gonostome shape. However, given its overall shape resemblance to other specimens that have been sequenced, especially an even smaller one from Northwestern Borneo (collected by A. Schulz, Borneo sp. 15 in previous analyses, MCZ DNA101513), the name Miopsalis is now used for a large clade that lives mostly on Borneo (Giribet et al. 2012). Examination of the type specimens available from Genoa (Clouse & Giribet in press) also led me to confirm the similarity between modern specimens from Sumatra used in molecular studies and the type species for Leptopsalis, thus allowing the bulk of stylocellid diversity to be recombined in this large genus. Materials and Methods Genitalia were photographed on a Laser Scanning Confocal Microscope (LCSM) (Zeiss LSM 710 Axiovert Z.1) after first being embedded in glycerine gel. Genitalia were removed and placed in glycerine for several hours, then glycerine gel on a depression slide was warmed until liquid, and the genitalia were transferred into it. This was allowed to cool and reheated as necessary to smooth the surface; no cover slip was used. The ovipositor was also photographed on a Leica compound microscope fitted with a digital imaging system. Legs, tarsi, and ventral details were photographed on a Hitachi S-4700 field emission scanning electron microscope (FE-SEM). Specimens were also photographed under a Leica MZ 12.5 stereomicroscope ( with a mounted JVC KY-F70B digital camera ( the specimens were photographed at different focal planes, and the images were assembled using the application Auto-Montage Pro Version by Syncroscopy ( THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 3

4 Specimens and collections sometimes have multiple identifying code numbers, as they have passed through different institutions and been included in phylogenetic analyses before being described. The Museum of Comparative Zoology (MCZ) DNA database records whole collections under a six-digit number, with the prefix MCZ DNA. All specimens were also recorded separately in a Cyphophthalmi database built in Biota (Colwell ) and housed in the Giribet Laboratory at MCZ; these gave each specimen a unique six-digit number, with the prefix SPM. Undescribed species that have been used in phylogenetic analyses also have had a simple moniker constructed from its collection locality, such as Borneo sp. 11, all of which are cross-referenced to MCZ DNA numbers in those publications. Certain appendage measurements were not possible for all specimens; the full length of the first cheliceral article requires removal of the chelicera, and trochanters are often broken when legs are removed. The depth of the chelicer was measured from behind the dorsal crest to a point between the first and second ventral process, and the width of the fourth tarsus was measured just proximal to the adenostyle. Certain types were not scored for traits that required too much manipulation of the specimen, which can cause damage. Cuticular sculpturing is described using terms from Murphree (1988). Diagnoses of taxonomic groups are written so as to add new information to diagnoses at higher levels. Taxonomy Order Opiliones Sundevall, 1833 Suborder Cyphophthalmi Simon, 1879 Infraorder Boreophthalmi Giribet, et al., 2012 Superfamily Stylocelloidea Hansen & Sørensen, 1904 Stylocellidae Hansen & Sørensen, 1904 Comments: Several morphological features in the family are distributed unevenly among the lineages, resulting in definitions for the genera that rest largely on statistical statements (Table 1). A case may be made for a particular specimen as belonging to one genus or another based on the probability that a member of that genus would present a certain combination of characters. For example, having an anal gland pore could place a species in any genus (when considering undescribed species in Leptopsalis and Meghalaya with anal gland pores), but if Rambla s organ is lacking, the ventral prosomal complex is large and flat, eyes are present, the fourth coxae tapered anteriorally, and the anterior sternal opisthosomal sulci are distinct and parallel, then it is most likely in Miopsalis and cannot be in Fangensis, Giribetia, Stylocellus, or Meghalaya. Known cheliceral characters clearly follow this distributional bias, states for any one aspect never being exclusive to one genus. My initial doubt over the role of the second ventral process is now replaced by a more complex picture: although a distinct second ventral process and a reduced first one is a synapomorphy of the family, a large first process and reduced second is more derived in Stylocellidae, with a number of exceptions. That is, although the ancestral stylocellid had the unique development of a second cheliceral process, the family appears to have generally reverted back to a condition common in the rest of Cyphophthalmi. A reduced first process may be a byproduct of troglomorphism, since mostly the troglobitic Miopsalis M. globosa (Schwendinger & Giribet, 2004) new comb., M. gryllospeca (Shear, 1993), and M. silhavyi (Rambla, 1991) have reduced first processes like troglobitic Fangensis; but so do Giribetia gen. nov. and Stylocellus, which live in leaf litter, as well as the small, non-troglomorphic Miopsalis leakeyi (Shear, 1993) new comb. Extensive granulation on the second article of the chelicera is mostly found in the early lineages, and having a small smooth ridge along the laterodistal area may be exclusive to Fangensinae subfam. nov. and Stylocellinae (not just Stylocellus, as I have seen it in one species of Meghalaya). Several species of Miopsalis also have the same heavy granulations down the second article, a feature that often comes with the other plesiomorphic traits, like anal gland pores, a large ventral prosomal complex, Rambla s organ, a reduced first ventral cheliceral process, and claw-like chelicerae. About the latter, I have simply defined it as having a ratio of the third-to-second cheliceral article of 0.3 or greater, although species in Stylocellus, Meghalaya, and Leptopsalis with this ratio look far less claw-like than those in Fangensis and Giribetia gen. nov. This is because the latter have the second article s widest point past halfway toward the distal end, and a curved distal end of the whole chelicera. However, keeping the definition just as the article ratio is clearer and sufficiently makes the point that claw-like chelicerae are more frequent in the early lineages, which also completely lack attenuate ones. 4 Zootaxa Magnolia Press CLOUSE

5 TABLE 1. Current species of Stylocellidae and states for certain morphological features. Locality codes are as follows: BO = Borneo, CP = Central Thai-Malay Peninsula, EH = Eastern Himalayas, JA = Java, NG = New Guinea, NP = Northern Thai-Malay Peninsula, PI = Philippines, SP = Southern Thai-Malay Peninsula, SU = Sumatra, SW = Sulawesi. Cheliceral proportions are shown as the ratio of third article (mobile digit) to the second article, followed by C ( clawlike ) if the ratio is equal to or greater than 3.0, and A ( attenuate ) if less than or equal to 2.0. Species Fangensinae subfam. nov. Fangensis cavernarus Schwendinger & Giribet, 2005 Locality Anal gland pore Rambla's organ First cheliceral process Second cheliceral process Cheliceral proportions NP yes small, distinct, raised reduced distinct 0.34 (C) Fangensis leclerci Rambla, 1994 NP yes small, distinct, raised reduced distinct 0.37 (C) Fangensis spelaeus NP yes small, distinct, raised reduced distinct 0.35 (C) Schwendinger & Giribet, 2005 Giribetia insulana (Schwendinger & Giribet, 2005) CP yes large, distinct, sunken reduced distinct 0.39 (C) Stylocellinae Stylocellus lornei, sp. nov. CP yes large, diffuse, flat reduced distinct 0.30 (C) Stylocellus sumatranus Westwood, 1874 SU yes?? 0.34 (C) Meghalaya annandalei Giribet, Sharma & Bastawade, 2007 EH large, diffuse, sunken reduced distinct 0.32 (C) Leptopsalinae subfam. nov. Miopsalis collinsi (Shear, 1993) BO yes large reduced 0.14 (A) Miopsalis globosa (Schwendinger & Giribet, 2004) SP yes reduced distinct 0.35 (C) Miopsalis gryllospeca (Shear, 1993) BO small, distinct, flat absent large 0.33 (C) Miopsalis kinabalu (Shear, 1993) BO large reduced 0.25 Miopsalis leakeyi (Shear, 1993) BO reduced large 0.31 (C) Miopsalis lionota (Pocock, 1897) BO yes small, distinct, flat??? Miopsalis mulu (Shear, 1993) BO yes distinct distinct 0.20 (A) Miopsalis pocockii (Hansen & Sørensen, 1904) BO?? large distinct 0.28 Miopsalis pulicaria Thorell, 1890 SP?? distinct reduced 0.19 (A) Miopsalis sabah (Shear, 1993) BO yes large reduced 0.21 (A) Miopsalis silhavyi (Rambla, 1991) BO small, distinct, flat reduced distinct 0.32 (C) Miopsalis tarumpitao (Shear, 1993) PI large reduced 0.32 (C) Leptopsalis beccarii Thorell, 1882 SU distinct distinct 0.20 (A) Leptopsalis dumoga (Shear, 1993) SW large distinct 0.35 (C) Leptopsalis hillyardi (Shear, 1993) SW distinct distinct 0.28 Leptopsalis javana (Thorell, 1882) JA large reduced 0.26 Leptopsalis laevichelis (Roewer, 1942) SP????? Leptopsalis lydekkeri (Clouse & Giribet, 2007) NG large reduced continued on the next page THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 5

6 TABLE 1. (Continued) Species Leptopsalis modesta (Hansen & Sørensen, 1904) Leptopsalis novaguinea (Clouse & Giribet, 2007) Locality Anal gland pore Rambla's organ First cheliceral process Second cheliceral process Cheliceral proportions SW large distinct 0.30 (C) NG distinct distinct 0.33 (C) Leptopsalis pangrango (Shear, 1993) JA large reduced 0.33 (C) Leptopsalis ramblae (Giribet, 2002) SP large distinct 0.29 Leptopsalis sedgwicki (Shear, 1979) SP large reduced 0.13 (A) Leptopsalis sulcata (Hansen & Sørensen, 1904) JA large distinct 0.31 (C) Leptopsalis tambusisi (Shear, 1993) SW large distinct 0.28 Leptopsalis thorellii (Hansen & SU?? 0.18 (A) Sørensen, 1904) Leptopsalis weberii (Hansen & Sørensen, 1904) Nomen dubium Stylocellus spinifrons Roewer, 1942 SU?? 0.20 (A) BO???? 0.20 (A) Rambla s organ (Fig. 3) is also found in more than one lineage, although it is more restricted than anal gland pores, and its different states appear to be exclusive to each lineage. Appearing as a small, raised patch in Fangensis (Schwendinger & Giribet 2005), it is enormous in Giribetia gen. nov. and Meghalaya (Fig. 3, E F), although in the former it is more clearly demarcated, the larger tuberculate-granules and microgranules stopping more abruptly at its proximal border and large tuberculate-granules extending for only a short distance past its distal border. In Stylocellus Rambla s organ is barely visible (keeping in mind that we can only examine it in one species, as the type of S. sumatranus is missing its fourth legs), but under a scanning electron microscope what can be seen as a shiny patch under a light microscope is recognized as an area where microgranules are missing and larger granules are reduced. In Miopsalis, where the first observation of this organ was made by Rambla in her description of M. silhavyi, the area is well-demarcated and small, not raised or depressed, and the microgranules, larger tuberculate-granules, and general cuticle combine to form a regular series of triangular projections that resemble scales (Fig. 3, A D). Although not noted in its original description (Shear 1993), this form of Rambla s organ can also be seen in the type of M. gryllospeca (Fig. 3, A), and I have photographed it in three other large, cave-dwelling Miopsalis. These species differ in size and even the presence of anal gland pores, but their Rambla s organ retains the same general appearance. Rambla s organ is completely missing so far in Leptopsalis, even among those with anal gland pores. Subfamiles included: Stylocellinae Hansen & Sørensen, 1904, Fangensinae subfam. nov., Leptopsalinae subfam. nov. Distribution: Southeast Asia, from northeastern India to western New Guinea, including the Thai-Malay Peninsula, the Indo-Malay Archipelago, and the Philippine islands of Palawan and Mindanao. Stylocellinae Hansen & Sorensen 1904 Comments: The first large molecular analysis to include both genera of Stylocellinae (Stylocellus represented at that time by the Phayam Island species, described below and Meghalaya) recovered the subfamily as monophyletic under all cost schemes, with 99% parsimony bootstrap support and 100% likelihood bootstrap support (Clouse & Giribet 2010), and the same result was obtained in the most recent combined molecular and 6 Zootaxa Magnolia Press CLOUSE

7 morphological analysis of Cyphophthalmi (Giribet et al. 2012). Clearly, molecular data contain a robust set of characters for this subfamily, but finding a morphological synapomorphy is more difficult. Meghalaya s morphology, the most distinctive in the family, is quite autapomorphic: the large, rectangular gonostome, scooplike male tarsus IV, thickened male tibia III, and deep ventral opisthosomal depression are not found in other genera at all (Giribet et al. 2007). Both Meghalaya and Stylocellus lornei, sp. nov., do have common features in their modified areas on the male fourth tarsi (what I interpret here as homologous to the Rambla s organ in Fangensinae), and perhaps the deeply sinusoidal anterior sternal opisthosomal sulci of Stylocellus are a result of the same ecological or developmental process that has resulted in the ventral depression of Meghalaya. FIGURE 1. Stylocellus sumatranus Westwood, 1874, male type specimen: A, dorsal; B, dorsal illustration from Hansen & Sørensen (1904); C, ventral; D, lateral. Scale bar = 2.00 mm for all views. Description: Eyes present; anal gland pores variable; chelicerae with distinct second ventral process, first process reduced; chelicerae claw-like; granulations on second cheliceral article variable, low ridge variable; Rambla s organ diffuse, large, lacking microgranulations but often with larger tuberculate-granulations encroaching irregularly, with micropores and microtubercles bearing pores; ozophores pronounced and pointing forward; posterior prosoma not especially large or bulging, the widest part of body usually across the opisthosoma (Fig. 4, B C); lateral edges of fouth coxae tapered anteriorally (Fig. 5, B C); ventral prosomal complex variable; sternum variable; body profile deep, arching of dorsal scutum variable (Fig. 6, B C); sternal opisthosomal sulci between sternites 3 and 4, 4 and 5, and 5 and 6 variable (Fig. 5, B C); posterior gonostome edge weakly concave. Included genera: Stylocellus Westwood, 1874 (Type genus) and Meghalaya Giribet, Sharma & Bastawade, 2007 Distribution: Sumatra (more specific locality information not recorded), southern, central, and northern Thailand, and northeastern India. Stylocellus Westwood, 1874 Stylocellus Westwood, 1874: 200; Pocock 1897: ; Roewer 1926: 263; Giribet 2000: 68. Removed from synonymy: Leptopsalis Thorell, 1882: [synonymized by Thorell (1890: 106), but restored by Clouse and Giribet (in press)] Type species: Stylocellus sumatranus Westwood, 1874, by monotypy. Other species included: Stylocellus lornei sp. nov. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 7

8 Comments: The two species in this genus are nearly identical in size and proportion (Fig. 7; males approximately 6 mm long and 3 mm wide across the opisthosoma), and the wider measurement across the ozophores of S. sumatranus could easily be caused by its pinning, which has clearly split and widened the ventral prosomal complex. Still, given the multiple evolution of various sizes and associated features in the family, the similar proportions could be causing these species to appear more closely related than they truly are. Thus, the diagnosis relies on other features I think give the two species a strong overall resemblance and which have clearly influenced their placement in morphometric phylogenies. Diagnostic features from the male fourth tarsus are based solely on S. lornei, sp. nov., as the type of S. sumatranus is missing this appendage. Description: Stylocellinae, with the following: anal gland pores and medial anal plate sculpturing present; second cheliceral article more than 2/3 granulated, granulations stopping before joint with third article, bordered by long lateral smooth patch and low ridge that extends to third article joint; Rambla s organ at same level with surrounding tarsus, lacking microgranulations, the larger, pointed tuberculate-granulations sometimes smaller or missing at the center of the organ but mostly continuing across as on the surrounding cuticle; ozophores long and parallel-sided, almost widening at the tip; ventral prosomal complex large, with fourth coxae meeting for a length longer than the gonostome, distinct sternum present, and second coxae broadly meeting; sternal opisthosomal sulci between sternites 3 and 4, 4 and 5, and 5 and 6 distinct and sinusoidal. Distribution: Sumatra and Phayam Island, off the west coast of Peninsular Thailand FIGURE 2. The two most commonly recovered phylogenetic hypotheses for the genera of Stylocellidae. The only difference is the position of Giribetia, which is sometimes recovered as sister to Fangensis (A), forming a monophyletic Fangensinae, and sometimes recovered as sister to (Stylocellinae + Leptopsalinae) (B). Stylocellus sumatranus Westwood, 1874 (Fig. 1) Stylocellus sumatranus Westwood, 1874: 200, Plate XXXVII, figs. 7, 7a, 7b. See complete synonymy in Giribet 2000: 68. Comments: Along with the widening of the prosoma, the pinning of S. sumatranus may have also caused the ozophores to point perpendicular to the long body axis (Hansen and Sørensen draw them pointing anteriorally, Fig. 1B), as well as the dorsal scutum to appear less arched than S. lornei, sp. nov. Likewise, desiccation of the specimen may have caused the sternal sulci to become more pronounced. Still, the sulci have the same sinusoidal appearance, the ozophores have the same shape, and these features, combined with the anal gland pore, cheliceral sculpturing, and overall body proportions, make the species extremely similar. In fact, this raises the question of whether S. sumatranus and S. lornei, sp. nov., are the same species (unlikely given the biogeography of Cyphophthalmi, but mix-ups in labels and locality identifications among early specimens could have happened). Three features of S. sumatranus clearly distinguish it from S. lornei: the lack of any evidence of a groove in the middle of the smooth strip on the anal plate, the lack of pronounced dorsal opisthosomal sulci, and the fact that its spiracles are completely free of the fourth coxae, not overlapped. 8 Zootaxa Magnolia Press CLOUSE

9 FIGURE 3. Rambla s organ in Miopsalis and Meghalaya: A, Miopsalis gryllospeca (Shear 1993) type; B, undescribed species of Miopsalis from Borneo ( Borneo sp. 18, MCZ DNA103262); C D, undescribed species of Miopsalis from Borneo ( Borneo sp. 13, MCZ DNA101517); E F, undescribed species of Meghalaya from the Thai-Malay Peninsula ( Peninsula sp. 19d, MCZ DNA101094, ). THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 9

10 FIGURE 4. Dorsal prosomal views of Stylocellidae genera: A, Fangensis spelaeus Schwendinger & Giribet, 2005; B, Giribetia insulana (Schwendinger & Giribet, 2005); C, Stylocellus lornei, sp. nov.; D, Meghalaya annandalei Giribet, Sharma & Bastawade, 2007; E, Miopsalis silhavyi (Rambla, 1991); F, Leptopsalis cf. beccarii ( Sumatra sp. 4d in previous analyses, MCZ DNA101478, collected by P. Schwendinger). All images were scaled to the same ozophore width and aligned along the ozophore ends. Stylocellus lornei, sp. nov. (Figs. 6 17; Table 2) Peninsula sp. 16 in: Clouse & Giribet 2010, Clouse et al Meghalaya sp in: Giribet et al Material examined: Holotype, Male (MHNG sample TH-07/02, MCZ DNA103251, MCZ Cyphophthalmi database SPM006161), southern Thailand, Ramong Province, Ko (= Island) Phayam ( N, E), m alt., northern tip of island, in remnant of semi-evergreen rainforest, leg. P. Schwendinger. Paratypes, 7 males (MHNG sample TH-07/02, MCZ DNA103251, SPM [SEM], , [dissected for genitalia], ) and 13 females (MHNG sample TH-07/02, MCZ DNA103251, SPM , 6179 [dissected for genitalia]) from same collection as holotype. All types are deposited in MHNG except the one used for SEM (SPM006159), which is deposited in the MCZ Cyphophthalmi SEM collection. Description: Holotype male. Body Size (Fig. 7, 8; Table 2): Body length (5.90 mm) about twice as long as the widest point (3.10 mm), between the second and third opisthosomal segments, and 2.33 times as wide at the width across the ozophores (2.65 mm). 10 Zootaxa Magnolia Press CLOUSE

11 FIGURE 5. Ventral opisthosomal views of Stylocellidae genera: A, Fangensis spelaeus Schwendinger & Giribet, 2005; B, Giribetia insulana (Schwendinger & Giribet, 2005); C, Stylocellus lornei, sp. nov.; D, Meghalaya annandalei Giribet, Sharma & Bastawade, 2007; E, Miopsalis silhavyi (Rambla, 1991); F, Leptopsalis cf. beccarii ( Sumatra sp. 4d in previous analyses, MCZ DNA101478, collected by P. Schwendinger). All images were scaled to the same total body length and aligned along the posterior edge. Eyes, ozophores, and sculpturing (Fig. 9): Eyes present directly anterior to ozophores, large, without cornea; cuticle over eye slightly raised, yellowish area visible underneath, surface texture like surrounding area. Ozophores lateral, raised above carapace edge ( type 2 for Cyphophthalmi); pointing anteriorly, not noticeably tapering, with an infolded opening. Body with microgranules over entire surface and tuberculate-granules and granules mostly on ventral prosomal complex, medial ventral opisthosoma, anal region, appendages, ozophores, medial dorsal soma, and tergites VIII and IX; granules missing on parts of chelicerae, male anal plate groove, and certain sulci; microgranulations missing on Rambla s organ. Microtubules with pores present, especially on tarsi, sparser than tuberculate-granules. Granule shape more frequently round on soma, more oblong and even pointed distally on legs. Larger tuberculate-granules distributed unevenly on dorsal surface of first cheliceral article, trochanters, and femurs III and IV. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 11

12 TABLE 2. Body and appendage measurements for Stylocellus lornei, sp. nov. Article abbreviations are as follows: Tr = trochanter, Fe = femur, Pa = patella, Ti = tibia, Mt = metatarsus, and Ta = tarsus. All measurements are expressed in mm. holotype male [ratio] (SPM006161) length Width (widest) Width (ozophore) body [1.90] 2.65 [2.23] Tr Fe Pa Ti Mt Ta leg I 0.45 [1.00] 1.93 [4.81] 0.93 [2.31] 1.33 [3.79] 0.48 [1.58] 1.18 [3.13] leg II 0.40 [1.00] 1.83 [4.29] 0.80 [1.78] 1.08 [3.31] 0.53 [1.75] 1.50 [5.00] leg III 0.40 [1.00] 1.17 [2.65] 0.80 [1.67] 1.13 [2.96] 0.53 [1.54] 1.35 [2.70] leg IV 0.58 [1.35] 2.22 [5.54] 0.98 [2.60] 1.38 [3.54] 0.55 [1.83] 1.73 [4.60] I (crest) I (whole) II III chelicer [7.17] 0.65 [5.70] Tr II III IV Ta palp [4.67] 0.77 [3.53] 0.78 [4.28] 0.75 [4.75] average male [ratio] (n = 5) length Width (widest) Width (ozophore) body [1.95] 2.65 [2.26] Tr Fe Pa Ti Mt Ta leg I 0.46 [1.10] 1.94 [4.77] 0.90 [1.95] 1.36 [3.83] 0.50 [1.66] 1.59 [4.18] leg II 0.42 [0.98] 1.82 [4.50] 0.88 [1.69] 1.12 [3.13] 0.51 [1.74] 1.50 [5.06] leg III 0.39 [1.03] 1.49 [3.62] 0.76 [1.70] 1.06 [2.78] 0.49 [1.61] 1.34 [3.88] leg IV 0.55 [1.24] 2.11 [5.00] 0.91 [1.90] 1.36 [3.32] 0.49 [1.58] 1.69 [4.12] I (crest) I (whole) II III chelicer 1.13 [2.37] 1.71 [3.22] 2.16 [6.95] 0.69 [5.95] Tr II III IV Ta palp 0.58 [2.58] 1.25 [4.58] 0.78 [3.02] 0.81 [4.23] 0.76 [3.87] average female [ratio] (n = 2) length Width (widest) Width (ozophore) body [1.98] 2.70 [2.37] Tr Fe Pa Ti Mt Ta leg I 0.50 [1.04] 2.08 [4.96] 0.98 [2.34] 1.54 [4.06] 0.60 [1.69] 1.78 [4.69] leg II 0.44 [1.00] 1.92 [4.57] 0.84 [1.91] 1.24 [3.28] 0.56 [1.70] 1.58 [4.99] leg III 0.46 [1.00] 1.68 [3.66] 0.80 [1.74] 1.14 [2.89] 0.52 [1.44] 1.48 [3.74] leg IV 0.56 [1.27] 2.28 [5.18] 0.96 [2.19] 1.38 [3.27] 0.52 [1.44] 1.86 [4.91] I (crest) I (whole) II III chelicer 1.17 [2.14] 2.27 [5.46] 0.7 [2.23] palp Tr II III IV Ta 0.47 [1.99] 1.02 [4.30] 0.78 [4.08] 1.05 [7.50] 0.75 [5.21] 12 Zootaxa Magnolia Press CLOUSE

13 FIGURE 6. Lateral views of Stylocellidae genera: A, Fangensis spelaeus Schwendinger & Giribet, 2005; B, Giribetia insulana (Schwendinger & Giribet, 2005); C, Stylocellus lornei, sp. nov.; D, Meghalaya annandalei Giribet, Sharma & Bastawade, 2007; E, Miopsalis silhavyi (Rambla, 1991); F, Leptopsalis cf. beccarii ( Sumatra sp. 4d in previous analyses, MCZ DNA101478, collected by P. Schwendinger). All images were scaled to the same total body length and aligned along the ozophore ends. Dorsal scutum (Figs. 7, 8): Transverse prosomal-opisthosomal sulcus and opisthosomal sulci distinct, deep, lacking sculpturing; mid-dorsal, longitudinal opisthosomal sulcus absent; dorsal midline with granules on either side on the prosoma, lacking microgranules narrowly down the opisthosoma. Dorsal prosoma meeting cheliceral ridge with pronounced lip, then rising steeply; in lateral view prosoma angling gradually to highest point at first opisthosomal tergite, rounding evenly to the posterior. Widest point at sulcus between tergites II and IIII. Ventral opisthosoma (Figs. 7, 8): Transverse sulci complete after sternite 3, deep and lacking sculpturing; first three sulci after tergite 3 sinusoidal, with medial region bending forward to same point as lateral ends. Sutures between tergites 6 and 7 and 7 and 8 gently curved back laterally. Spiracles large and C-shaped, their anterior edges slightly under swollen posterior edge of fourth coxae. Ventral prosomal complex (Fig. 10): Granules distinct and completely covering coxae I, lacking on distal ends of coxae II, distal fourth of coxae III, and distal half of coxae IV. Coxae IV meet along midline for distance slightly longer than gonostome opening; coxae III not meeting, distinct sternum in between proximal ends; coxae II meeting for about the same length as coxae IV, with distinct endites. Gonostome wider than long, somewhat rectangular; first opisthosomal sternite rounded, forming concave posterior edge to gonostome; lateral walls of gonostome formed by distinct, elevated posteroproximal processes of fourth coxae. Anal region (Fig. 10): Sternites 8 and 9 and tergite IX free; tergite IX wider than sternite 9. Tergite VIII with large pore along suture with tergite IX; tergite IX with smooth area medially alongside tergite VIII pore; anal plate with distinct medial longitudinal area lacking granulations, containing fine medial longitudinal groove. Anal plate groove with slightly irregular or wrinkled appearance, ending anteriorally in short array of smaller grooves, like a river delta. Anal plate and surrounding sternites and tergites otherwise with pronounced, large granulations. Chelicerae (Figs. 7, 10B, 11): Third article slightly less than one-third the length of second article, second article widest just proximal of mid-length (Fig. 8). First article with distinct dorsal crest, ridge continuing laterally to pronounced, moderately sized, second ventral process. First ventral cheliceral process highly reduced to broad angle (Fig. 11). Ventro-lateral edge of first article with distinct ridge running from second ventral process to distal end (Fig. 12, A B). Second article with tightly packed, mostly dorsal, large granulations that stop abruptly before joint, remainder of the second article and third article smooth. Smooth area around granulations on laterodistal end of second article with slightly raised region, creating low ridge (Fig. 11, B; Fig. 12, B C). Palps, legs, and tarsi (Figs. 7, 10, 12 14; Table 2): Palps completely granulated, without process on trochanter (Fig. 11A). Legs I IV completely granulated on all articles, except large granules missing in patches on dorsal and ventral trochanters and femurs I and IV (Fig. 13). Tarsus I with fairly distinct solae that extends more than half tarsus length, distinguished by distinct angle in ventral edge of tarsus I and dense area of short hairs distally (Fig. 14). Median dorsal surface of tarsi I and II with row of evenly spaced, curved hairs, more distal on tarsus I. Tarsus IV wider distal of adenostyle; adenostyle pointing vertically and located 55% of tarsus length from joint. Retrolateral surface with faintly visible Rambla s organ about halfway between adenostyle and claw, covering oval region with width about half this distance (Fig. 14, D; Fig. 15). Surface of Rambla s organ not noticeably raised or depressed, lacking microgranulations, with wrinkled cuticle and microtubules with pores and micropores; edge of Rambla s organ indistinct, marked by the patchy loss of microgranulations (Fig. 15, A D). THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 13

14 body males (5) females (2) D leg I A length widest width ozophore width length/ widest 2.0 E leg II 1.5 chelicer F leg III B I (crest) I (whole) II III palp 2.0 G leg IV C Tr II III IV Ta Tr Fe Pa Ti Mt Ta FIGURE 8. Body and appendage article measurements for five male (dark gray) and two female (white) specimens of Stylocellus lornei, sp. nov., expressed as the average ± the standard deviation. Body measurements and the length-to-width ratio (as measured across the widest part) are shown for the type specimen of S. sumatranus Westwood, 1874, as a narrow black bar to the left of S. lornei, sp. nov., male measurements (A). Cheliceral measurements for article I are done both as the whole article and just from the dorsal crest, as the entire article cannot be seen on most complete specimens (B). Palp and leg article abbreviations are as follows: Tr = trochanter, Fe = femur, Pa = patella, Ti = tibia, Mt = metatarsus, and Ta = tarsus (C G). 14 Zootaxa Magnolia Press CLOUSE

15 FIGURE 7. Stylocellus lornei, sp. nov., male paratype (SPM006159): A, dorsal; B, ventral, C, lateral. Scale bar = 1.00 mm for all views. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 15

16 FIGURE 9. Detailed views of Stylocellus lornei, sp. nov.: A, ventral prosoma; B, anterior lateral prosoma; C, spiracle and surrounding region; D, dorsal anterior. Scale bars = 0.50 mm for all panels. Spermatopositor (Figs ): Microtrichial formula: 4, 10, (Fig. 16); dorsal microtrichia attached along angle, medial three on each side attached distinctly proximal to the lateral dorsal microtrichia; ventral microtrichia in close proximity and fairly level, medial two slightly more distal; apical microtrichia with two groups of four separated by a medial invagination, lateral two distinctly smaller and positioned along sides at widest point; all microtrichia with denticles on shafts, ventral surface with denticles. Gonopore complex (Fig. 17) with broad lobus medialis, with small lobuli laterales; lacinia dorsalis extending distally to as far as lobus medialis and flanked by digiti that extend further. Ovipositor (Fig. 18): With 49 circular articles, ending in two apical lobes, each with several setae over whole surface, ending in an apical seta, and just subapical a sensitive process containing a dense set of slightly curved setae of moderate length (shorter than apical seta but longer than setae on the lobe); with distinct pigmented body on posterior half of the insides of each lobe. Variation (Fig. 8): Females slightly larger than males but with same body proportions; length of males measured mm (n=5); measured females mm (n=2). Smaller males have smaller appendages, females have larger appendages, thus all specimens, whether male or female, have the same general appearance and body proportions. Sculpturing, ventral opisthosomal sulci, and ozophore shape similar on males and females. Females lack anal gland pore but do retain bilobed tergite IX and lack of granules on the dorsomedial anal plate (as in Fangensis and Giribetia). Some specimens have more pronounced eyes than the male holotype, with a more distinct cornea. Distribution: Known only from the type locality, which is the northern end of Phayam Island, off the western coast of Peninsular Thailand. Natural history: Stylocellus lornei sp. nov. lives in humid leaf litter, but its behavior and ecology are otherwise unknown. Etymology: This species is named in honor of Lorne Michaels, a loyal and generous supporter of the American Museum of Natural History and a member of its Board of Trustees since Zootaxa Magnolia Press CLOUSE

17 FIGURE 10. Stylocellus lornei, sp. nov., male paratype (SPM006159) anal region (A) and gonostome (B). Scale bar = 0.50 mm for both panels. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 17

18 FIGURE 11. Stylocellus lornei, sp. nov., male paratype (SPM006159) palp (A) and chelicer (B). Scale bar = 1.00 mm for both. 18 Zootaxa Magnolia Press CLOUSE

19 FIGURE 12. Stylocellus lornei, sp. nov., male paratype (SPM006159) cheliceral details: A, ventral ridge on article I; B, lateral ridge on article II (upper arrow) and ventral ridge on article I (lower arrow); C, lateral ridge on article II. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 19

20 Meghalaya Giribet, Sharma & Bastawade, 2007 Meghalaya Giribet, Sharma & Bastawade, 2007 [originally family uncertain, placed in Stylocellidae by Clouse and Giribet 2010: 9]. Type species: Meghalaya annandalei Giribet, Sharma & Bastawade, 2007 by original designation. Other species included: None Description: Stylocellinae, with the following: anal gland pores almost always absent, small ones known in one undescribed species; granulations on second cheliceral article highly reduced, low ridge known only in the same species with the anal gland pore; Rambla s organ large and depressed (missing in the species with the anal gland pore), when present forming a scoop-shaped fourth tarsus in males (Fig. 3, E F); ozophores tapered, large, and pointing forward; ventral prosomal complex usually compressed; sternum absent; body profile deep, dorsal scutum arched; sternal opisthosomal sulci between sternites 3 and 4 indistinct due to deep ventral depression; posterior gonostome edge weakly concave, anterior edge broad, forming large, rectangular or squarish gonostome; adenostyle thick; male tibia III thickened. Distribution: Northeastern India and northern to southern Thailand. FIGURE 13. Stylocellus lornei, sp. nov., male paratype (SPM006159) legs: A, leg I; B, leg II; C, leg III; D, leg IV. Scale bar = 1.00 mm for all legs. Fangensinae subfam. nov. Comments: Fangensinae subfam. nov. consists of the four species originally described under the genus Fangensis, and it has been recovered as monophyletic only when molecular markers are analyzed separately or are few (Clouse & Giribet 2007; Schwendinger & Giribet 2005), combined molecular analyses are done with low transversion and indel costs (Clouse et al. 2009; Clouse & Giribet 2010), or not all of the species have been included (Giribet et al. 2012). Even when Fangensinae is monophyletic, Giribetia insulana, new comb., has always been recovered as sister to the other species, and under analyses that simultaneously combine several markers, use transformation costs that cause the least variance of tree lengths among the partitions, and include all 20 Zootaxa Magnolia Press CLOUSE

21 four species, G. insulana falls outside the subfamily, often as sister to the remainder of the family (Clouse et al. 2009; Clouse & Giribet 2007; Clouse & Giribet 2010; Schwendinger & Giribet 2005). Whether paraphyly of Fangensinae relative to the rest of the family is a result that strengthens with further species discoveries and more characters, or becomes understood as an analytic artifact, the fact remains that morphologically its members form the most easily recognized and cohesive group of the three subfamilies recognized here. FIGURE 14. Stylocellus lornei, sp. nov., male paratype (SPM006159) tarsi: A, tarsus I; B, tarsus II; C, tarsus III; D, tarsus IV. Scale bar = 1.00 mm for all tarsi. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 21

22 Description: Eyes absent or poorly developed; anal gland pores and anal plate modifications present; distinct second ventral cheliceral process, first process reduced to a broad angle; chelicerae claw-like; second cheliceral article almost completely granulated, granulations contrasting sharply with laterodistal smooth patch that forms low ridge against granulations; Rambla s organ clearly defined, lacking granulations, not at the same level as the surrounding tarsus; ozophores tapered, pointing perpendicular to long body axis (Fig. 4, A B); posterior prosoma clearly bulging (Fig. 4, A B), constituting the widest part of the body; coxae IV large, positioning spiracles well behind gonostome, with lateral edges parallel to long body axis (Fig. 5, A B); ventral prosomal complex large, with fourth coxae meeting for a length longer than the gonostome, distinct sternum present, and second coxae broadly meeting; sternal opisthosomal sulci between sternites 3 and 4, 4 and 5, and 5 and 6 parallel; posterior gonostome edge straight (Fig. 5, A B). Included genera: Fangensis Rambla, 1994 (Type genus) and Giribetia gen. nov. Distribution: Caves in northern and western central Thailand and islands of the western coast of southern Thailand. FIGURE 15. Stylocellus lornei, sp. nov., male paratype (SPM006159) detailed views of inner tarsus IV: A, region of interest is shiny patch on tarsus between claw and adenostyle; B, SEM image of region; C, detailed image of cuticle at border of region showing sculpturing change; D, highly magnified image of border region. Fangensis Rambla, 1994 Fangensis Rambla, 1994: ; Giribet 2000: 56; Schwendinger & Giribet 2005 [originally in Sironidae, moved to Stylocellidae by Giribet (2002: 3)]. Type species: Fangensis leclerci Rambla, 1994 by original designation. Other species included: Fangensis cavernarus Schwendinger & Giribet, 2005; Fangensis spelaeus Schwendinger & Giribet, 2005 Diagnosis: Fangensinae, with the following: eyes absent; Rambla s organ small and raised; gonostome norrow with strongly rounded anterior edge; dorsal scutum arched, body elliptical in lateral view. Distribution: Klaeb Yai Cave in northern Thailand (Fangensis leclerci) and Kaeng Lawa Cave (Fangensis spelaeus) and Tham Nam Cave (Fangensis cavernarus) in western central Thailand. 22 Zootaxa Magnolia Press CLOUSE

23 FIGURE 16. Spermatopositor of Stylocellus lornei, sp. nov., paratype SPM006164: A, dorsal; B, lateral; C, ventral. Scale bar = 0.10 mm for left view only. Giribetia gen. nov. Fangensis (partim): Schwendinger & Giribet (2005): Type species: Fangensis insulanus Schwendinger & Giribet, Other species included: None Diagnosis: Fangensinae, with the following: eyes present but poorly developed, without a distinct lens; Rambla s organ large and depressed; gonostome wide and regtangular; dorsal scutum flat, dorsal and ventral outlines parallel medially in lateral view (Fig. 6, B). Etymology: This genus is named in honor of Professor Gonzalo Giribet of Harvard University. His extensive work on the molecular systematics and biogeography of Cyphophthalmi has raised a once obscure group of Opiliones to prominence. Distribution: Ko Siray and Ko Phuket, two adjacent islands off the western coast of southern Thailand. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 23

24 FIGURE 17. Spermatopositor of Stylocellus lornei, sp. nov., paratype SPM with dorsal microtrichia removed, showing gonopore complex: A, dorsal; B, lateral and slightly anterior view. LM = lobus medialis; LD = lacinia dorsalis; LL = lobulus lateralis; D = digitus; Scale bar = 0.10 mm for left view only. Leptopsalinae subfam. nov. Comments: More than Stylocellinae, Leptopsalinae subfam. nov. provides a contrast between clear molecular support and subtle morphological synapomorphies, although here the morphological difficulty extends to the genera as well. The subfamily is best defined by its loss or moderation of character states easily noticed in other groups they lack the enlarged coxae IV and swollen posterior prosomas of Fangensinae subfam. nov., the combination of sinusoidal sternal opisthosomal sulci, anal gland pores, large size, and knob-like ozophores of Stylocellus, and the deep ventral depression and flattened male tarsi IV of Meghalaya. Between the genera Leptopsalis and Miopsalis, most species can be easily placed by simply knowing the collection locality and whether males have anal gland pores: most species from Borneo, and miniature leptopsalines from Peninsular Malaysia and Sumatra with anal gland pores are in Miopsalis, and all species from Java, Sulawesi, and New Guinea, all of which lack anal gland pores, have been consistently recovered in Leptopsalis (Clouse et al. 2009; Clouse & Giribet 2007; Clouse & Giribet 2010). There are species from Sarawak (northwestern Borneo), the Lingga-Riau islands and lower Peninsular Malaysia that resemble Miopsalis (in some cases even having anal gland pores) but are consistently recovered as members of Leptopsalis (usually as early lineages) in molecular phylogenies, but these are few. Nonetheless, given how many species are known from only one or two specimens, and the large areas yet unexplored, it seems probable that the above pattern will not hold, so we need to have more than a biogeographic diagnosis. The species from northern Borneo (Fig. 19) challenge any easy definitions of Miopsalis and Leptopsalis, as some of them have apparently converged strongly on the general appearance of Leptopsalis. Nine species have been described from Sabah and the nearby region around Gunung Mulu (Hansen & Sørensen 1904; Rambla 1991; Shear 1993). The large species with known males Miopsalis gryllospeca and M. silhavyi have been recovered as Miopsalis easily in morphometric analyses, despite not having anal gland pores. Likewise, two of the smaller, more rounded species M. collinsi (Shear, 1993) and M. sabah (Shear, 1993) have anal gland pores and are also 24 Zootaxa Magnolia Press CLOUSE

25 easily recovered as Miopsalis in morphometric phylogenies. However, smaller species that lack anal gland pores, M. kinabalu (Shear, 1993) new comb. and M. leakeyi (Shear, 1993) new comb., as well as the smallest species collected in the region with an anal gland pore, M. mulu (Shear, 1993) new comb., are consistently recovered as members of Leptopsalis in morphometric phylogeneis. It can easily be seen why M. mulu new comb. groups with Leptopsalis despite its anal gland pore, for it has curved sternal opisthosomal sulci, an extremely concave posterior gonostome, compressed ventral prosomal complex, huge eyes, and rather uninformative, attenuate chelicerae. The reason why I consider M. mulu new comb., M. kinbalu new comb. and M. leakeyi new comb. convergent members of Miopsalis and not in Leptopsalis is that similar species collected by S. Kurbatov in 2002 from Mt. (Gunung) Kinabalu ( Borneo sp. 11 in recent phylogenetic analyses, MCZ DNA103249) and by A. Schulz from Gunung Mulu ( Borneo sp. 4, MCZ DNA101525, by) consistently place as Miopsalis in molecular phylogenies. Certain derived Leptopsalis species have also converged on typical Miopsalis features, but these are more instructive and not so confounding in morphometric studies. A recently discovered species from New Guinea, L. lydekkeri (Clouse & Giribet, 2007), is never recovered outside of Leptopsalis in molecular studies, but it has a distinctly flattened dorsal scutum and a wide, flattened gonostome with a relatively straight posterior edge, as in Miopsalis. In Northern Sulawesi, L. hilyardi (Shear, 1993) is a large species with small, tapered ozophores, long prosoma, and an elongate, elliptical overall soma shape, quite reminiscent of large Miopsalis species like M. silhavyi (Fig. 20). From the dorsal view, these two species have nearly identical proportions, but the ventral prosomal complex is distinctly different and highlights one of the most general differences between the two leptopsaline genera. For L. hillyardi, the meeting of the first coxae, the posterior corners of the gonostome, and the posterior corners of the fourth coxae are all more anterior than in M. silhavyi. Combined with the concave posterior gonostome, lack of a distinct sternum, simple chelicerae, and lack of male anal gland pores, L. hillyardi is confidently placed in Leptopsalis. Indeed, morphometric phylogenetic analyses have grouped it with the other (much smaller and more rounded) species on Sulawesi, which is highly concordant with the history of the island (Clouse et al. 2009). Description: Eyes usually distinct and sometimes large, absent or reduced only in highly troglomorphic or miniaturized species; anal gland pores usually absent but variable; chelicerae often with reduced second ventral cheliceral process, first process often distinct or large; chelicerae ranging from claw-like to attenuate; granulations on second cheliceral article often highly reduced, low ridge absent; Rambla s organ usually missing, if present then small, distinct, at the same level as surrounding cuticle, and with modified, scaly sculpturing (Fig. 2, A D); ozophores usually tapered and pointing forward; posterior prosoma not bulging; lateral edges of fouth coxae tapered anteriorally; ventral prosomal complex variable; sternum variable; body profile variable, arching of dorsal scutum variable; sternal opisthosomal sulci between sternites 3 and 4, 4 and 5, and 5 and 6 ranging from straight to curved, sometimes indistinct; posterior gonostome highly variable, ranging from straight to deeply concave. Included genera: Leptopsalis Thorell, 1882 (Type genus) and Miopsalis Thorell, 1890 Distribution: Central Thailand to western New Guinea, including the islands of Sumatra, Java, Sulawesi, Borneo, Palawan, and Mindanao. Leptopsalis Thorell, 1882 Leptopsalis Thorell, 1882: 23 24; Karsch 1884: 145; Thorell 1890: 106 (as junior synonym of Stylocellus Westwood, 1874); Giribet 2000: 68 (as junior synonym of Stylocellus Westwood, 1874); Clouse and Giribet (in press) (restored). Type species: Leptopsalis beccarii Thorell, 1882, by monotypy. Other species included: Leptopsalis dumoga (Shear, 1993), Leptopsalis hillyardi (Shear, 1993), Leptopsalis javana (Thorell, 1882), Leptopsalis laevichelis (Roewer, 1942) new comb., Leptopsalis lydekkeri (Clouse & Giribet, 2007), Leptopsalis modesta (Hansen & Sørensen, 1904), Leptopsalis novaguinea (Clouse & Giribet, 2007), Leptopsalis pangrango (Shear, 1993) new comb., Leptopsalis ramblae (Giribet, 2002), Leptopsalis sedgwicki (Shear, 1979) new comb., Leptopsalis sulcata (Hansen & Sørensen, 1904), Leptopsalis tambusisi (Shear, 1993), Leptopsalis thorellii (Hansen & Sørensen, 1904), Leptopsalis weberii (Hansen & Sørensen, 1904) Description: Leptopsalinae, with the following: eyes distinct and often large; anal gland pores almost always absent; granulations on second cheliceral article usually highly reduced; Rambla s organ absent; prosoma almost always compact, not elongate; ventral prosomal complex usually highly compressed; sternum usually indistinct; THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 25

26 body profile usually deep, dorsal scutum usually arched; sternal opisthosomal sulci between sternites 3 and 4, 4 and 5, and 5 and 6 usually curved, anterior medial sulci often indistinct; posterior gonostome often deeply concave. Distribution: Central Thailand and south down the Thai-Malay Peninsula, Sumatra, Java, Sulawesi, western New Guinea, and, to a small extent Borneo. FIGURE 18. Ovipositor of Stylocellus lornei, sp. nov., paratype SPM Scale bar = 0.10 mm for left panel only. 26 Zootaxa Magnolia Press CLOUSE

27 FIGURE 19. The nine described species from northern Borneo and two undescribed species included in recent molecular phylogenies. Described species are depicted as dorsal views of the type specimens and adjusted to the same scale (scale bar = 1.00 mm for all specimens). Locality number (1 4, as indicated on the central map) for each specimen is in the upper left corner of each photo, followed by an indication of whether males have anal gland pores ( a ) or are known from female specimens only ( f ). Miopsalis sabah (Shear, 1993) appears to have been collected at the same time as M. leakeyi (Shear, 1993), which the collection label indicates as G. Silum (mistakenly transcribed as Gunung Sabah in the description); this locality is most likely Gunung Silam. Leptopsalis pangrango (Shear, 1993) new comb. Stylocellus pangrango Shear, 1993: , figs Comments: This species was highly unstable in previous morphometric phylogenies, but this was probably due to the difficulty in taking good measurements from the disarticulated and broken type, which also has the anal region THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 27

28 distended. However it is clearly a member of Leptopsalis, having the condensed ventral prosomal complex, large first ventral cheliceral article, and lack of an anal gland pore or Rambla s organ. Material examined: Holotype, Male (AMNH), Jawa Barat (= West Java), Gunung Pangrango, elev m, August 1921 (no collector named). Leptopsalis sedgwicki (Shear, 1979) new comb. Stylocellus sedgwicki Shear, 1979: , figs Comments: The lack of an anal gland pore or Rambla s organ, compressed ventral prosomal complex (including the lack of a sternum), and large first ventral cheliceral process place this species in Leptopsalinae subfam. nov., and most likely in Leptopsalis. The ozophores are long and thick, with a straight anterior edge; with the posterior slightly lowered they can appear perpendicular to the long body axis (as in Shear s (1993) original drawing), but they are not entirely shaped like those of Stylocellus nor as perpendicular as those of Fangensinae subfam. nov. I have also analyzed a collection of juveniles by G. Cuccodoro and I. Löbl from the same area as the type locality (Penang Island [= Pulau Pinang]) that are consistent with this species in size (too large, it seems, to be M. pulicaria) and shape, and which are consistently recovered as Leptopsalis in molecular phylogenies ( Peninsula sp. 24, MCZ DNA101492). Material examined: Holotype, Male (AMNH), Malaysia, Penang Island, 7 June 1976, col. W. Sedgwick. Leptopsalis laevichelis (Roewer, 1942) new comb. Stylocellus laevichelis Roewer, 1942: 278. See complete synonymy in Giribet 2000: 68. Comments: Roewer s description of this species from southwestern Peninsular Malaysia did not allow us to place it in a genus, since it contained little information on the important characters and lacked illustrations. In fact, the description contradicts an important synapomorphy of the family (Giribet 2002), granulation on the second cheliceral segment ( Glied frontal ebenfalls glänzend glatt, nirgends bekörnelt = Second limb, from the front, likewise glossy, with no granulations ). Roewer s description of laevichelis is mostly concerned with granulations, so we can assume Roewer examined this closely, but many species in Leptopsalis have secondarticle granulations of the chelicerae reduced to only a few proximal bumps that are difficult to see under a normal light microscope. The adenostyle is described as having a little bristle on top ( einem Spitzenbörstchen ), which would make more sense as a stylocellid feature if described in the plural (e.g., tuft of bristles ), although we might again attribute this to low magnification during Roewer s examination. I have thus obtained the type specimen and can see that it is a large member of Leptopsalis and has smooth, extremely attenuate second cheliceral articles, with only one or two granules on the proximal end. The adenostyle is small but tufted, the specimen has small eyes, and the posterior gonostome is deeply concave and thus heart-shaped. Some measurements of the type, in mm, are as follows: body length 5.33, width across the ozophores 2.73, width across the widest part of the opisthosoma 2.87, second cheliceral article length 1.9, and third cheliceral article length Material examined: Holotype, Male (Senckenberg Museum, Frankfurt (SGN), Roewer collection no. 7221/ 22), Malaysia, no collection date given, determined by Roewer Miopsalis Thorell, 1890 Miopsalis Thorell, 1890: 381; Hansen and Sørensen 1904: 99; Roewer, 1923: 47 48, fig. 50; Roewer, 1926: 263; Shear, 1979: ; Giribet 2000: 72 (as nomen dubium); Clouse & Giribet (in press) (validated). Type species: Miopsalis pulicaria Thorell, 1890, by monotypy. 28 Zootaxa Magnolia Press CLOUSE

29 FIGURE 20. Comparison of large Miopsalis and Leptopsalis species: A & C, M. silhavyi (Rambla, 1991); B & D, L. hillyardi (Shear, 1993). Specimens were scaled to the same length between the posterior edge and anterior coxae II; vertical lines indicate the posterior extent of coxae IV. Scale bars = 1.00 mm for both species. THE LINEAGES OF STYLOCELLIDAE Zootaxa Magnolia Press 29

30 Other species included: Miopsalis collinsi (Shear, 1993), Miopsalis globosa (Schwendinger & Giribet, 2004) new comb., Miopsalis gryllospeca (Shear, 1993), Miopsalis kinabalu (Shear, 1993) new comb., Miopsalis leakeyi (Shear, 1993) new comb., Miopsalis lionota (Pocock, 1897), Miopsalis mulu (Shear, 1993) new comb., Miopsalis pocockii (Hansen & Sørensen, 1904) new comb., Miopsalis sabah (Shear, 1993), Miopsalis silhavyi (Rambla, 1991), Miopsalis tarumpitao (Shear, 1993). Description: Leptopsalinae, with the following: anal gland pores usually present; chelicerae often with extensive granulations on the second article; Rambla s organ sometimes present in large species; prosoma often relatively long; ventral prosomal complex usually large, with fourth coxae meeting for a length longer than the gonostome, distinct sternum present, and second coxae broadly meeting; body profile variable, but often flat in species of various sizes; sternal opisthosomal sulci between sternites 3 and 4, 4 and 5, and 5 and 6 often straight, parallel; posterior gonostome usually straight or weakly concave. Distribution: Mostly Borneo and to a lesser extent Peninsula Malaysia, central Sumatra, and the Philippine islands of Palawan and Mindanao. FIGURE 21. Distribution of Stylocellidae lineages. The Stylocellus range is incomplete due to a lack of information about where on the island S. sumatranus Westwood, 1874, was collected. Inset: detailed view of the Kra Isthmus, where all the subfamilies overlap and every genus can be found within a 1500 km stretch of the Thai-Malay Peninsula. Miopsalis globosa (Schwendinger & Giribet, 2004) new comb. Stylocellus globosus Schwendinger & Giribet, 2004: 1423, figs Comments: This highly troglomorphic species has moved between Fangensinae subfam. nov. and Miopsalis in various morphometric phylogenetic analyses (Clouse et al. 2009). Although it has placed most commonly with the 30 Zootaxa Magnolia Press CLOUSE