Zootaxa, Review and Cladistic Analysis of the Neotropical

Review and cladistic analysis of the Neotropical tarantula genus Ephebopus Simon 1892 (Araneae: Theraphosidae) with notes on the Aviculariinae

RICK C. WEST1, SAMUEL D. MARSHALL2, CAROLINE SAYURI FUKUSHIMA3,4 & ROGÉRIO BERTANI4 1 3436 Blue Sky Place, Victoria, B.C., CA. E-mail: [email protected] 2 Department of Biological Sciences, 317 Bienvenu Hall, Northwestern State University Natchitoches, LA. E-mail: [email protected] 3 Instituto de Biociências, Departamento de Zoologia, Universidade de São Paulo. E-mail: [email protected] 4 Instituto Butantan, Av. Vital Brazil 1500, 05503–900, São Paulo SP. E-mail: [email protected]

Abstract

The tarantula genus Ephebopus Simon 1892 is reviewed and includes the type species, E. murinus (Walckenaer 1837), and E. uatuman Lucas, Silva & Bertani 1992, E. cyanognathus West & Marshall 2000, E. rufescens West & Marshall 2000 and Ephebopus foliatus, sp. nov., from Guyana. Ephebopus violaceus Mello-Leitão 1930 is transferred to Tap- inauchenius Ausserer, where it is a senior synonym of Tapinauchenius purpureus Schmidt 1995 new synonymy. Ephebo- pus fossor Pocock 1903 is considered a nomen dubium. Ephebopus occurs in northeastern South America where it is known only from Brazil, Guyana, Suriname, and French Guiana. Spiders of the genus are generally fossorial; however, Ephebopus murinus has a developmental stage that is arboreal. A cladistic analysis of the Theraphosidae retrieves the Aviculariinae as monophyletic, including Avicularia Lamarck, Iridopelma Pocock 1901, Pachistopelma Pocock 1901, Tapinauchenius, Psalmopoeus Pocock, Ephebopus, Stromatopelma Karsch and Heteroscodra Pocock, having as a synapomorphy the well-developed scopulae on tarsi and metatarsi I–II that is very laterally extended.

Key words: Aviculariinae, systematics, Ephebopus, Guyana, cladistic analysis, theraphosid behavior

Introduction

The Theraphosidae are found on all continents, except Antarctica. More than 900 species are described, con- sisting roughly of a third of all described mygalomorph species (Platnick 2008). Most theraphosids are found in tropical areas but some representatives live in subtropical and temperate regions. They are mainly terrestrial in habit, living in burrows and other natural cavities or under rocks and fallen logs. Arboreal forms are found mainly in the tropical New World but also in Africa and Asia. The taxonomy of this family is very confused and taxonomic revisions are rare (Raven 1990). Cladistic analyses of relationship of species and genera are still rare and there was only one attempt to recover the cladistic relationships of all theraphosid subfamilies (Raven 1985). Therefore, the taxonomic position of several genera in theraphosid subfamilies is controversial, as is the composition of the Aviculariinae. Simon (1892) created the Avicularieae, including only Avicularia and Tapinauchenius, then characterized by the absence of a line of setae dividing the tarsal scopulae (i.e., entire or undivided scopula) of leg IV, legs without spines, and posterior legs longer than anterior ones. Later, Pocock (1901) included Pachistopelma and Iridopelma as well as Ephebopus and Psalmopoeus in the Avicularieae. He also suggested the inclusion of the West African genera Scodra Becker 1879 (= Stromatopelma) and Heteroscodra. Simon (1903) included the

Accepted by R. Raven: 10 Jul. 2008; published: 13 Aug. 2008 35 two African genera in Avicularieae, but transferred Ephebopus to the Phoneyuseae and synonymized Iri- dopelma with Avicularia. Mello-Leitão (1923) considered only American genera belonged in the Avicularii- nae, including his monotypic genus Ancylochiros Mello-Leitão 1920. He also considered Typhochlaena C. L. Koch to be valid and Iridopelma its junior synonym. Roewer (1942) transferred Heteroscodra and Scodra back to the Aviculariinae. He also included the Middle Eastern genus, Avicuscodra Strand (= Chaetopelma Ausserer, Ischnocolinae), in the Aviculariinae but considered Typhochlaena a junior synonym of Avicularia and placed Ancylochiros (= Avicularia) in the Ischnocolinae and Ephebopus in the Eumenophoriinae. Raven (1985) transferred Ephebopus to the Theraphosinae, Stromatopelma and Heteroscodra to the Eumenophorinae, Psalmopoeus to the Selenocosmiinae, revalidated Iridopelma and considered Ancylochiros a junior synonym of Avicularia. Raven (1985) also included only the genera Avicularia, Iridopelma, Pachis- topelma, and Tapinauchenius in the Aviculariinae based on the following shared traits: spinose process between the lobes of the male palpal tarsi, tarsi as broad as or broader than the metatarsi, and the legs being weakly spined or lacking spines. Raven (1985: 119) noted, based on a curator’s drawing of the type, that the type species of Tapinauchenius, T. plumipes (C. L. Koch 1842), appeared to lack a well-developed spinose process on the male palpal tarsi. Lucas et al. (1991) described the male of E. murinus and based on the posses- sion of a sinuous embolus lacking keels (in addition to characters used by Raven 1985), and transferred Ephe- bopus to the Aviculariinae. Based on their observations, they noted that males of both Ephebopus and Tapinauchenius lacked the spinose process on the palpal tarsi, but considered the wide tarsal pads, elongated embolus, and absence of spines on the legs as synapomorphies uniting the Aviculariinae As for Ephebopus itself, no other theraphosid genus has had its taxonomic position so uncertain. First included in the Selenocosmieae by Simon (1892), the genus was later transferred variously to the Avicularieae (Pocock 1901), Phoneyuseae (Simon 1903), back to Aviculariinae (Mello-Leitão 1923), Eumenophorinae (Roewer 1942), Theraphosinae (Raven 1985), and again to Aviculariinae (Lucas et al. 1991). Thus, it was placed at least once in four of the eight theraphosid subfamilies recognized by Raven (1985). The type species of Ephebopus, Mygale murina Walckenaer 1837, was based on a brief but illustration- free description but was fairly accurate. After some decades, Simon (1892) transferred M. murina to Ephebo- pus. Simon (1903) examined the holotype of Santaremia pococki F. O. P.-Cambridge 1896 and concluded it was a junior synonym of Ephebopus murinus. Pocock (1903) described a second species, Ephebopus fossor Pocock 1903 from Ecuador, and Mello-Leitão (1930) added the third, E. violaceus Mello-Leitão 1930 from Cuminá River region, Brazil. Only females of Ephebopus species were known until Lucas et al. (1991) described the male of E. murinus, and males and females of a new species E. uatuman from Presidente Figueredo, Amazonas, Brazil. These authors also transferred E. violaceus to Avicularia. Two more species of Ephebopus were described recently, Ephebopus cyanognathus and Ephebopus rufescens, both from French Guiana (West & Marshall 2000). We describe a new species of Ephebopus from Guyana, and present a cladistic analysis of Ephebopus species, including representatives of seven theraphosid subfamilies and all aviculariine genera, in order to test the cladistic relationships of Ephebopus and the Aviculariinae.

Material and methods

All measurements are given in mm and were made to the nearest 0.01 mm with the aid of a Wild-Heerbrugg dissecting microscope using an ocular micrometer or with dial vernier calipers. Leg and pedipalp measurements were taken from the dorsal aspect on the left side (unless appendages were lost or obviously regener- ated) of all specimens whereas coxae and trochanters are measured from their ventral aspect, on the same side. Leg span is taken on left side in a straight line from the tip of leg I to that of leg IV. The extent of metatarsal scopulae is presented as a fraction (maximum 1.0) measured from the base ventrally. Claws are not included

36 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. in measurements of tarsi. Images were made with a Nikon Coolpix 5000 digital camera and a WILD stereomi- croscope. Scanning electron micrographs were obtained with a LEO 440I from the Instituto de Geociências da Universidade de São Paulo. Standard abbreviations and measurements are used for ocular and spination descriptions and follow those of Prentice (1992) and Reichling and West (1996). Male palpal bulb keel terminology follows Bertani (2000); urticating hair type terminology follows Cooke et al. (1972) and Marshall & Uetz (1990). Material of the following institutions were examined: AMNH – American Museum of Natural History, New York; BMNH – The Natural History Museum, London; IBSP – Instituto Butantan, São Paulo; MIZA – Instituto de Zoologia, Universidad Central de Venezuela, Maracay; MNHN – Muséum National d'Histoire Naturelle, Paris; MZSP – Museu de Zoologia, Universidade de São Paulo, São Paulo; NMNH – National Museum of Natural History, Washington, DC; MNRJ – Museu Nacional, Rio de Janeiro; SMF – Senckenberg Museum, Frankfurt. Geographical coordinates: primary sources are between round brackets, secondary sources between square brackets. Cladistics: A data matrix with 21 taxa and 48 characters (Table 3) was used with three programs using equal (Hennig86 1.5, Farris 1988; Nona 2.0 for Windows ,Goloboff 1993a) successive (Hennig86 1.5) and implied character weighting (X-Pee-Wee 1.3 for Windows Goloboff 1997). For Hennig86, the exact algorithm "ie". For X-Pee-Wee 1.3 and Nona 2.0, the commands h100, h/20, amb- and mult*50 were used. For X-Pee- Wee, concavities from 1 to 6 were used. All characters were treated as unordered. Material examined for cladistic analysis: Representatives of seven of the eight theraphosid subfamilies recognized by Raven (1985) were included, as well as all aviculariine genera and other available arboreal taxa with uncertain position (Poecilotheria Simon 1885, Heteroscodra, Stromatopelma). An undescribed species of Melloina Brignoli 1985 (Paratropididae, Glabropelmatinae) was included as the outgroup. It was chosen because paratropidids are the sister-group of Theraphosidae and Melloina is the genus which retains more plesiomorphic characters in common with the Theraphosidae (Raven 1985). Paratropididae: Glabropelmatinae. Melloina sp., VENEZUELA: 1 male, MIZA 520, 1 female, MNRJ 12961, both from Lara, Cueva El Santuario [9°49’ N 70°03’W], 19 April 2000, O. Villarreal. Theraphosidae, Aviculariinae: Avicularia juruensis Mello-Leitão 1923, BRAZIL: male, IBSP 2503, state of Rondônia, Porto Velho [8º45’ S, 63º54’ W]; female, IBSP 10279, state of Mato Grosso, between Vale de São Lourenço and Pontes & Lacerda [14º40’S, 56º51’ W], UHE Guaporé. Ephebopus spp.: Specimens cited in the descriptions were used. Heteroscodra maculata Pocock, AFRICA: male, IBSP 9642, pet trade; GUINEA-BISSAU: female, IBSP 9644, pet trade. Iridopelma hirsutum Pocock 1901, BRAZIL: male, IBSP 8078, state of Paraíba, João Pessoa [7o07’ S, 34º52’ W]; female IBSP 9664, state of Sergipe, São Cristóvão [11º00’ S, 37º12’ W] Pachistopelma rufonigrum Pocock 1901, BRAZIL: male, IBSP 4921, state of Bahia, Salvador [13o00’ S, 38º31’ W], Ondina; female, IBSP 8085, state of Sergipe, Brejo Grande [10º25’ S, 36º27’ W].Psalmopoeus cambridgei Pocock 1895, TRINIDAD–TOBAGO: male, IBSP 9653, pet trade. Psalmo- poeus sp., VENEZUELA: female, IBSP 9655, pet trade. Stromatopelma sp., SIERRA LEONE: male, IBSP 9665, pet trade; AFRICA: female, IBSP 11136, pet trade. Tapinauchenius violaceus (Mello-Leitão 1930), FRENCH GUIANA: male (paratype of T. purpureus Schmidt 1995), SMF 38046; female (holotype of T. purpureus Schmidt 1995), SMF 38042. Eumenophorinae: Citharischius crawshayi Pocock 1900, KENYA: male, IBSP 8530, female IBSP 9643, born in capitivity. Harpactirinae: Pterinochilus sp., ANGOLA: male, IBSP 9647, Buila-Dala [11o10’ S, 20o12’ E]; AFRICA: female, IBSP 8765, pet trade. Ischnocolinae: Holothele rondoni (Lucas & Bücherl 1972), BRAZIL: male (holotype) and female (paratype), IBSP 4090, state of Amazonas, Iauaretê [o36’ N, 69o11’ W]. Ornithoctoninae: Haplopelma longipes von Wirth & Striffler 2005, CAMBODIA: male, MZSP 28761,

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 37 Skuon [1o33’ N, 104o55’E], A. Anderson, 10 July 2003; Haplopelma minax (Thorell 1897), THAILAND: female, IBSP 9645, 1 mi E. Bangkok [13o43’ N, 100o31’ E]. Poecilotherinae: Poecilotheria sp., INDIA: male, IBSP 9660, pet trade; Poecilotheria ornata Pocock 1899, SRI LANKA: female, IBSP 8767. Selenocosmiinae: Phlogiellus sp., VIETNAM: male, MZSP 28762, R. Blauman, April 1989; Phlogiellus sp., VIETNAM: female, AMNH, pet trade. Theraphosinae: Euathlus vulpinus (Karsch 1880), CHILE: 3 males, IBSP 3817–A; 1 female, IBSP 3817– B, Osorno [40o34’ S, 73o09’ W]; Lasiodora sp., BRAZIL: male, IBSP 11143, state of Paraíba, João Pessoa [7o07’ S, 34º52’ W]; Lasiodora sp., female, IBSP 10293, state of Pernambuco, Jaboatão dos Guararapes [8o06’ S, 35º00’ W], Conjunto Murebeca. Other material examined Theraphosidae, Aviculariinae: Psalmopoeus emeraldus Pocock 1903, COLOMBIA: female holotype, BMNH 1894.12.9.1, examined in 2003 by RB; Tapinauchenius sp., PERU: female, MZSP 28763, Loreto (4o09’ S, 74o25’ W), Yarapa River; Ephebopus violaceus, BRAZIL: female holotype, MNRJ, state of Pará, Cuminá River.

Systematics

Aviculariinae Simon 1892

Ephebopus Simon 1892 (Figs 1–20, 30)

Ephebopus Simon 1892: 155; Pocock 1903: 85; Raven 1985: 119; Marshall & Uetz 1990: 120; Lucas, Silva & Bertani 1991:161; West & Marshall 2000: 6; Platnick 2008. Type species: Mygale murina Walckenaer 1837, by original designation.

Diagnosis. Differs from all other theraphosid genera by the apical patch of type V urticating hairs on prolateral pedipalp femora of males and females (Figs 19–20). Description: Cephalothorax longer than wide, cephalic region slightly raised, convex. Cephalic and thoracic striae distinct. Fovea deep, straight. Chelicerae without rastellum. Eye tubercle distinct, wider than long. Clypeus absent. Anterior eye row straight. Labium subquadrate, slightly wider than long, with numerous (100–300) cuspules concentrated on anterior half. Maxilla subrectangular, anterior lobe distinctly produced into conical process, inner angle bearing numerous cuspules (more than 100). Sternum longer than wide. Pos- terior sigilla submarginal. STC of males and female with median row of few small teeth. Tarsi I–IV and metatarsi I–II fully scopulated, metatarsus III scopulated along half its length, metatarsus IV apically scopulated. Scopulae of tarsi and metatarsi I–II extended very laterally giving them a spatulate appearance. Femur IV without retrolateral scopula. Type V of urticating hair in pad on distal prolateral palpal femora (Marshall & Uetz 1990). Legs with few spines on distal ventral tibiae and metatarsi. Stridulatory setae absent. Males. Spur on tibia I bipartite; metatarsus I straight, when flexed closes on outer upper process; male palpal bulb pyri- form; embolus narrow, long, 2–3 times longer than tegulum, keels absent. Females. Two spermathecae weakly sclerotized. Species included: Ephebopus murinus (Figs 1–2, 11–12), E. uatuman (Figs 3–4, 13), E. cyanognathus (Figs 5–6, 16), E. rufescens (Figs 7–8, 17–18) and E. foliatus (Figs 9–10, 14). Distribution & Habitat: Northeastern and Central Brazilian Amazon, French Guiana, Southern Suri- name and Southwestern Guyana (Fig. 30). E. murinus is found in lowland rainforests, in fringing grasslands above riparian flood plains; E. rufescens is found in both lowland and upland rainforest; E. uatuman and E.

38 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. cyanognathus are found in upland rainforest areas; E. foliatus specimens were collected in lowland riparian rainforest. Remarks: Ephebopus fossor is considered a nomen dubium, as its type specimen is lost (pers. comm. J. Beccaloni, BMNH, 2000). Pocock’s (1903) original description is too vague to determine the correct generic placement of this species (e.g., lacking any reference to the brush of urticating hairs on the palpal femur) and its type locality in Ecuador is outside the otherwise known zoogeographical range of Ephebopus (i.e., northeastern South America). Ephebopus violaceus was transferred to Avicularia by Lucas et al. (1991) who failed to locate the holotype. They based the transfer only on the original description: first ocular row procurved, pattern on the dorsal side of abdomen and the division of the posterior tarsal scopulae; the latter suggesting the specimen was a juvenile. Recently, the holotype was rediscovered in the arachnological collection of Museu Nacional, Rio de Janeiro. Although a small specimen, after dissection of the genital region, we found well-formed spermathecae which do not resemble those of Avicularia species, since each receptaculum is short and lacks median curvature (Fig. 21). Furthermore, the specimen lacks type II urticating hairs, has an almost straight anterior ocular row, and has a few spines only on the ventral apex of tibiae and metatarsi. These two characteristics indicate it would belong to Ephebopus or Tapinauchenius. The absence of a type V urticating hair pad on the pedipalp femur suggests it is a Tapinauchenius species. In comparing the holotype of E. violaceus with that of Tapinauchenius purpureus, we found a strong resemblance in spermathecal shape (Figs 21–22). Furthermore, the two species are from the same region (north of Para state, Brazil and French Guiana), and both have the same typical purple color pattern (viz., "violaceus" and "purpureus"). Thus, E. violaceus is transferred to Tapinauchenius, creating the new combina- tion Tapinauchenius violaceus (Mello-Leitão 1930) and T. purpureus is considered a junior synonym of T. violaceus, new synonymy.

Ephebopus murinus (Walckenaer 1837) (Figs 1–2, 11–12, 30)

Mygale murina Walckenaer 1837: 220. Ephebopus murinus: Simon 1892: 155; Mello-Leitão 1923: 316; Lucas, Silva & Bertani 1991: 246, f. 1–4. Santaremia pococki F.O.P.-Cambridge 1896: 746, pl. 33, f. 8–9, pl. 34, f. 20, pl. 35, f. 12; Pocock 1901: 548–549. Holo- type female from Santarém, Pará, Brazil, in BMNH, not examined. First synonymized by Simon 1903: 952. Type: Mygale murina holotype female (MNHP AR4760, 2160) from ‘Le Para’; examined.

Remarks: Walckenaer did not state the county of origin in the original description. Simon (1892) examined the type and placed it into Ephebopus; he also noted the country of origin was unknown, but cited there was the indication "Brazil", "probably by mistake", since E. murinus "resembles the Aviculariids of the Old World" (translated from Simon 1892: 155). When Simon visited Pocock at the BMNH and examined the type specimens of Santaremia pococki , both Pocock and Simon concluded that Santaremia pococki was a junior synonym of E. murinus (Pocock 1901: 547–548; Simon 1903: 952) and concluded that Walckenaer's 1837 'Le Para' specimen came from that region of Brazil (Simon 1903). Actually, Belém, the present capital of the state of Pará was, at that time, called Santa Maria do Grão Belem do Pará, and was commonly abbreviated to 'Para' (Papavero 1973). Santaremia pococki was described from Santarem, also in the state of Pará, where E. murinus is very common (F.O.P.-Cambridge 1896; pers. obs.). Because of the detailed original description of San- taremia pococki, which includes a color plate, together with type locality, we support the synonymy proposed by Simon and Pocock. Diagnosis: E. murinus differs from all other Ephebopus species by the bold, ivory-colored parallel bands dorsally on the patellae and tibiae I–IV of both sexes, more so in females on patellae and tibiae I–II (Fig. 11).

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 39 FIGURES 1–10. Ephebopus, genitalia. (1–2) E. murinus. 1, Male, AMNH, palpal bulb, retrolateral. 2, Female, AMNH, spermathecae. (3–4) E. uatuman. 3, Male, holotype, palpal bulb, retrolateral. 4, Female, paratype, spermathecae. (5–6) E. cyanognathus. 5, Male, paratype, palpal bulb, retrolateral. 6, Female, paratype, spermathecae. (7–8) E. rufescens. 7, Male, paratype, palpal bulb, retrolateral. 8, Female, paratype, spermathecae. (9–10) E. foliatus sp. nov.. 9, Male, holotype, palpal bulb, retrolateral. 10, Female, paratype, spermathecae. Scale line: 1 mm.

40 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. Additionally, males differ from congeners in that the male palpal bulb is large and globular with a strong embolus tapering apically with the tip outwardly geniculate (Fig. 1). Females differ by the spermathecae having two widely separated tall columnar lobes, widest at the base, tapering apically with a small process medially on the outer edge of each lobe (Fig. 2). Description: Holotype female, lengths: total body, 61.3; chelicerae, 12.0; carapace, 24.1; abdomen, 25.2; leg I, 74.6; leg II, 67.0; leg III, 55.7; leg IV, 68.9. Leg formula I, IV, II, III. Color of legs and abdomen dark gray; patellae and tibiae dorsum with bold ivory colored parallel stripes, more so on legs I, II; carapace pale buff. Spermathecae: two tall columnar lobes, widely separated, widest at base constricting medially, continu- ing apically as tall lobes, outer edge of each lobe with small medial process (Fig. 2). Male (IBSP 4937), Bra- zil, state of Pará, Tucuruí [3o45' S, 49o40' W], VI.1986, lengths: total body, 37.8; chelicerae, 3.6; carapace, 15.3; abdomen, 18.9; leg I, 68.88; leg II, 59.14; leg III, 51.53; leg IV, 61.32. Leg formula, I, IV, II, III. Color of legs and abdomen dark reddish brown, legs and abdomen with longer amber setae; carapace with lighter amber pubescence; ivory-colored parallel stripes on all patellae and tibiae but not as bold as that found in females. Palpal bulb large, globular; embolus strong and tapering apically, tip outwardly geniculate (Fig. 1). Additional Material Examined: FRENCH GUIANA: 1 female, Iracoubo District, Iracoubo (05o28’ N, 53o12’ W), III.2001, S. Marshall; 1 female, Montsinéry District, Emerald Jungle Village [4o49’ N, 52o21’ W], IV.1999, S. Marshall; 1 male, Montsinéry District, Emerald Jungle Village [4o49’ N, 52o21’ W], IV.1999 (matured IX.1999), S. Marshall. BRAZIL: 1 female, state of Pará, Ananindeua [1o21’S, 48o22’ W], 18.VII.1974, R. F. de Silva; 2 males, state of Amapá, no other data. Distribution: Northeastern Brazilian Amazon, French Guiana and Southern Suriname (Fig. 30). Natural History: The burrows of E. murinus are distinctive among Ephebopus species. They are a verti- cal tube terminating in a retreat chamber as do some fossorial theraphosids, but the burrow entrance has a large and elaborate trumpet-shaped turret of silk (Fig. 11). We have also observed that early instars of E. murinus live in arboreal refugia, constructing retreats of silk in terrestrial bromeliads (SDM & RCW, pers. obs.) (Fig. 12). Unlike spiderlings of E. murinus, spiderlings of other Ephebopus species have not been found living in silk tubes in vegetation.

Ephebopus uatuman Lucas, Silva & Bertani 1992 (Figs 3–4, 13, 30)

Ephebopus uatuman Lucas, Silva & Bertani 1992: 161, f.1–7 Types: BRAZIL: Male holotype IB 4939, female paratype IB 4940, state of Amazonas, Presidente Figueiredo [2o02’ S, 60o01’ W], Uatuman River, Balbina Hydroelectric Power Station, 19.II.1988, M. Costa; examined.

Diagnosis: Similar to E. cyanognathus but this species differs from all congeners by the coloration in females, lacks the metallic blue chelicerae (Fig. 13), and shape of the genitalia of both sexes. The male ale differs from those of all other congeners by the palpal bulb being globular with a long embolus that abruptly tapers and is slightly bent apically (Fig. 3). Spermathecae of females of E. uatuman (Fig. 4) are similar to those of E. rufescens but differ from that of other congeners by having two stout columnar lobes that are widely separated, widest at base, and by being narrower apically. Description: Holotype male, lengths: total body, 34.0; chelicerae, 4.9; carapace, 12.0; abdomen, 17.1; leg I, 56.5; leg II, 50.3; leg III, 44.2; leg IV, 52.9. Leg formula I, IV, II, III. Color of carapace and legs dark amber brown; abdomen same color but with numerous longer reddish hairs; transverse narrow buff yellow banding between all femora and patellae. Palpal bulb globular, embolus long, abruptly tapers and slightly bent apically, and stouter than in E. cyanognathus (Fig. 3). Paratype female, lengths: total body, 37.9; chelicerae, 5.1; carapace, 14.8; abdomen, 18.0; leg I, 54.4; leg II, 47.84; leg III, 38.88; leg, 52.48. Leg formula I, IV, II, III. Color

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 41 of carapace pale brown; legs darker than abdomen and carapace; abdomen with greenish metallic pubescence. Spermathecae: two columnar lobes, widely separated, widest at base and narrowing more apically (Fig. 4) than in E. rufescens.

FIGURES 11–16. Ephebopus, habitus and burrow. (11–12) E. murinus. 11, Female in burrow entrance. 12, Antepenulti- mate female in retreat in bromeliad leaf. (13, 15) E. uatuman. 13, Female. 15, Detail of burrow entrance. (14) E. foliatus sp. nov., female paratype, in life. (16). E. cyanognathus, female. Photos: 14, M. Kuntner, other by R. C. West.

42 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. FIGURES 17–18. Ephebopus rufescens. 17, Female exposed in retreat under overturned rock. 18, Female in arboreal retreat two meters above ground in palm tree. Photos: R. C. West.

FIGURES 19–20. Ephebopus murinus. 19, Threat display showing urticating hair pad on apical prolateral pedipalp femur (arrow). 20, Detail of type V urticating hairs. Photo: R. Bertani.

FIGURES 21–22. 21, Tapinauchenius violaceus, holotype, spermathecae. 22, Tapinauchenius purpureus, holotype, spermathecae.

Additional Material Examined: 3 males (IB 7853–7855) and 3 females (IB 7850–7852) from same data as type material. Distribution: The type locality and Brazil, Amazonas, 30 km W. of Manaus, Jacaré Creek (Fig. 30).

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 43 Natural History: Spiders were found in upland forest areas in Brazil: Amazonas, Reserva Balbina and 30 km W. of Manaus, Jacaré Creek. Antepenultimate, penultimate and female E. uatuman constructed a simple flare-mouthed tubular burrow in rainforest in soft year long damp soil covered in leaf litter (Fig. 15), similar to that of E. cyanognathus. The entrances were all a raised flared collar of leaves bound together with silk and opened above the litter. The burrow terminated in an enlarged chamber and was about 30–40 cm deep. Mature males were found in February.

FIGURES 23–26. Some characters used in cladistic analysis. (23–24) Eye tubercle and clypeus. 23, Heteroscodra sp., female, low eye tubercle and narrow clypeus. 24, Haplopelma sp., female, high eye tubercle and wide clypeus. (25–26) Cymbium and male palpal bulb. 25, Avicularia sp., subtriangular cymbium, thin embolus with curvature to retrolateral side. 26, Lasiodora sp., male, rounded cymbium, thick embolus.

FIGURES 27–28. Tapinauchenius sp., female, maxilla, prolateral. 27, Area with modified fimbria setae (arrow) resem- bling weakly developed lyra. 28, Detail of setae.

Ephebopus cyanognathus West & Marshall 2000 (Figs 5–6, 16, 30)

Ephebopus cyanognathus West & Marshall 2000: 6, f. 1–2; West & Marshall 2002: 7. Types: FRENCH GUIANA: Male holotype and female paratype, Roura District [4o43’ N, 52o19’ W], Tresor Mountains, Tresor Reserve, 17.VI.1999 (male matured 10.IX.1999), R. West; 2 males paratypes, Roura District [4o43’ N, 52o19’ W], Roura Mountains, 22.IV.1999 (matured in captivity 28–30.VIII.1999), R. West; 1 female, Kourou District, C.I.R.A.D. Field Station [5o09’ N, 52o38’ W], 17.IV.1999, J. Huff; 1 female, Roura District [4o43’ N, 52o19’ W], Tresor Mountains, Tresor Reserve, 17.VI.1999, R. West. All types in AMNH.

Diagnosis: Similar to E. uatuman but this species differs from all congeners by the coloration of females, the metallic blue chelicerae (Fig. 16), and the different shape of the genitalia of both sexes. Males differ from

44 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. those its congeners by the palpal bulb being globular with a long slender gradually tapering embolus slightly bent outward apically (Fig. 5), similar to that in E. uatuman but straighter, thin and tapering. Spermathecae differ from those of its congeners by having two short truncate columnar lobes, separated, being widest at base and constricting medially then widening apically (Fig. 6). Description: Holotype male, lengths: total body, 41.17; chelicerae, 7.0; carapace, 16.67; abdomen, 17.5; leg I, 72.16; leg II, 65.34; leg III, 54.5; leg IV, 68.5. Leg formula, I, IV, II, III. Color of legs dark brown; legs with narrow yellow transverse band between all femora and patellae; chelicerae with lavender pubescence; abdomen pale gold intermixed with longer gray setae; carapace and trochantera with rose and brown pubescence setae, more so on carapace. Palpal bulb globular, with long slender tapering embolus, similar to that of E. uatuman but straighter and thin, bent upward apically (Fig. 5). Paratype female, lengths: total body: 45.84; chelicerae, 6.67; carapace, 19.67; abdomen, 19.5; leg I, 65.16; leg II, 57.14; leg III, 46.20; leg IV, 58.90. Leg formula, I, IV, II, III. Color amber brown; chelicerae entirely with metallic blue pubescence; carapace lighter brown with greenish pubescence; legs and abdomen darker amber brown with narrow transverse yellow band between all femora and patellae, similar narrower transverse band of amber between all tarsi and metatarsi. Spermathecae: two shorter truncate columnar lobes, separated, widest at base, constricting medially then widening apically (Fig. 6). Distribution. Known only from French Guiana (Fig. 30) where, in addition to the above localities, it has also been found in Maripasoula [3o43’ N, 54o04’ W] and Montsinéry District [4o54’ N, 52o30’ W], Saül [3o37’ N, 53o12’ W] and Chevaux Mountains. Natural History. The spiders appear to be widely distributed in upland rainforest areas. They were rarely common, with the exception of a large (30+) population observed along a 10 meter long x 2 meter high road bank in Saül in September 1981 (SDM, pers. obs.). Burrows were constructed in a fully shaded laterite clay embankment, facing northeast, at about 210 meters elevation in upland rainforest. The type specimens from the Tresor Reserve were found in burrows constructed in fully shaded upland rainforest on steep, north-facing slopes in year long damp clay soil covered in leaf litter at elevations of 200–300 meters. Antepenultimate, penultimate and female E. cyanognathus construct a simple flare-mouthed tubular burrow, similar to that of E. uatuman (Fig. 15). Adult female burrows were found to terminate in an enlarged chamber about 30 cm deep (RCW and SDM, pers. obs.). We observed an adult female E. cyanognathus in her burrow with early instars in April of 2001. This specimen and her young were not collected. We have also observed an individual in an upland rainforest retreat constructed of a dead leaf in a plant in the Chevaux Mountains in March of 1999. The retreat had no obvious entrance and was about 20 cm above the ground.

Ephebopus rufescens West & Marshall 2000 (Figs 7–8, 17–18, 30)

Ephebopus rufescens West & Marshall 2000: 8, f. 3–4. Types: FRENCH GUIANA: Male holotype and female paratype, Roura District, Kaw Mountains, vicinity of Camp Cäi- man [4o37’ N, 51o55’ W], 17.IV.1999 (male matured 14.IX.1999), R. West & J. Huff; paratypes: 1 male, Guiana, Montsinéry District [4o54’ N, 52o30’ W], Chevaux Mountains, 17.IV.1999 (matured 18.VIII.1999), R. West & J. Huff; 1 male, Montsinéry District, Emerald Jungle Village [4o49’ N, 52o21’ W], 15.IV.1999 (matured 21.IX.1999), R. West & J. Huff; 1 female, Roura District [4o43’ N, 52o19’ W], Roura Mountains, 19.IV.1999, R. West & J. Huff; 1 female, Roura District, Kaw Mountains, vicinity of Camp Cäiman, 18.VI.1999, R. West. All types in AMNH.

Diagnosis: Females can be distinguished from those of their congeners by distinct dark reddish brown coloration and lighter narrow parallel stripes dorsally on the patellae and femora and by the lack of yellow banding on the patellal–femoral joints of all legs (Figs 17–18) and shape of the genitalia of both sexes. The male palpal bulb differs from that of its congeners by being large and globular with a shorter tapering embolus that is

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 45 slightly bent apically (Fig. 7). Spermathecae are similar to those of E. uatuman but differ by being two stout separated columnar lobes (Fig. 8) that are not as wide as at base and narrowing apically as in E. uatuman. Description: Holotype male, lengths: total body, 35.75; chelicerae, 6.42; carapace, 13.0; abdomen, 16.33; leg I, 57.99; leg II, 53.32; leg III, 45.32; leg IV, 57.17. Leg formula, I, IV, II, III. Color of legs dark brown; chelicerae, carapace, trochantera and base of all femora with longer short golden setae. Abdomen uni- formly dark brown with longer reddish setae dorsally and laterally. All dorsal leg joints with narrow transverse band of short golden setae. Palpal bulb large, globular, short slender embolus tapering apically (Fig. 7). Paratype female, lengths: total body: 46.17; chelicerae, 7.33, carapace, 16.17; abdomen, 22.67; leg I, 59.16; leg II, 52.67; leg III, 44.83; leg IV, 55.0. Leg formula I, IV, II, III. Color of legs, abdomen and carapace dark reddish brown with longer lighter setae, more so on chelicerae and carapace; leg femora darker brown, dorsum of all patellae, tibiae and metatarsi with pale narrow parallel or single stripes. Spermathecae: two stout columnar lobes, separated, widest at base, similar to that in E. uatuman but not as wide at base and narrowing apically (Fig. 8). Distribution: The type localities and French Guiana: Maripasoula [3o43’ N, 54o04’ W], Saül [3o37’ N, 53o12’ W]. Brazil: Amazonas, 30km W. of Manaus, Jacaré Creek (Fig. 30). Natural History: The spiders were widespread and, apparently, the most ecologically versatile in their retreat placement of all Ephebopus species (Figs 17–18). Spiders were found in retreats in hollow logs and stems, both on and above the ground, in holes and hollows of standing trees, in arboreal termitaria and in mosses on the sides of shaded rock faces in rainforest. One retreat was found 3 meters above ground in a decaying palm tree. Wherever found, the spider built its retreat in association with wood or mosses. This was in or on dead or fallen trees, roots or moss covered trees, logs or rocks. A second trait of E. rufescens retreats that makes them distinct from those of other congeners is that when burrowing in the ground, the spider (par- ticularly adult females) built an extended tubular silk retreat. This extended tubular silk retreat was generally one-quarter to one-third of the length of the burrow; it may be oriented vertically or horizontally. The retreat was usually camouflaged with soil and vegetative debris, incorporated into the silk retreat, and was attached to a woody or moss substratum. The retreat entrance had a slightly flared collar constructed of silk.

Ephebopus foliatus sp. nov. (Figs 9–11, 14, 30. Tables 1, 2)

Types: GUYANA: Male holotype and 5 paratype females, Upper Takutu–Upper Essequibo Region, 4.42 km S. Gunn’s Strip, bank of Essequibo River, 240 metres, (1°38’45.7” N, 58°38’14.6” W), 6–15.VII.1999, J.A. Coddington, G. Hormiga, J. Miller, I. Agnarsson & M. Kuntner, deposited in NMNH.

Etymology: From the Latin folium, a leaf, and refers to the subtle four-point oak leaf pattern on the dorsum of the abdomen. Diagnosis: This species differs from all other Ephebopus by having a leaf-like pattern on the abdominal dorsum, when viewed from the side (Fig. 14), and by the shape of the genitalia of both sexes. The male palpal bulb is similar to E. cyanognathus but differs from its male congeners by having a long, thinner, tapering embolus, slightly bent apically (Fig. 9). The spermathecae differs from those of all other female congeners by having two tall columnar lobes, not widely separated, widest at base, gradually narrowing then rounding apically, sides with raised folds (Fig. 10). Description: Holotype male, lengths: total body, 29.56; carapace, 11.36 long x 10.72 wide; abdomen, 12.0 long x 6.72 wide; chelicerae, 6.2 long x 2.28 wide; cheliceral macroteeth: (left side, right side) 10, 9 with basal granules. Sternum, 5.60 long x 4.48 wide; glabrous sigilla opposite coxae II and III, posterior sigilla the largest, 0.14 sternum width. Labium, ovoid 1.44 long x 1.92 wide; labiosternal suture a narrow groove with

46 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. distinct lateral mounds; about 149 cuspules in tight cluster. Maxillae about (left, right) 180, 159 cuspules; cuspules in tight cluster on ventral inner heel of maxillae, heel rounded, anterior lobe conical and slightly elongated. Fovea: a deep circular pit. Cephalothorax: caput not rising abruptly from thoracic region but in gradual arch. Caput length 7.52, width 5.92. Ocular area 2.28 wide x 0.96 long, tubercle elevated. Anterior eye row straight; AME round, diameter 0.56, 0.23 apart; ALE elliptical, 0.33 x 0.44, 1.47 apart. Posterior eye row slightly recurved; PME ovoid, 0.21 x 0.37, 1.19 apart; PLE ovoid, 0.28 x 0.48, 1.49 apart. Clypeus 0.48 wide. Leg span 112.0mm, taken on right side (leg IV missing. Leg formula I, IV, II, III (Table 1). Femur III not noticeably swollen when viewed from above. All tarsal scopulae entire, without setal division. Metatarsal scopulae: I, 0.77; II, 0.85; III, 0.67; IV, 0.28. Type V urticating hairs in dense pad on apical prolateral face of palpal femur. No plumose setae on palp or leg segments. Spination: leg I, metatarsus 1v(1ma), tibia 4v (3 megaspines, one along entire inner lower process length, two on apical upper process, one on inner and one on outer, 1 ra); leg II, metatarsus 1v(1ma), tibia 2v(1pa, 1ra); leg III, metatarsus 3v(1pa, 1ma, 1ra), tibia, 2v(1pa, 1ra); ;leg IV, metatarsus 2v(1pa, 1ra); palp, aspinose. Spur on tibia I bipartite; lower process 1.20 long, with single basal megaspine along entire inner face; upper process 1.80 long, with preapical megaspine on oppos- ing inner and outer curved face. Metatarsus I closes on outer upper process. Palpal bulb similar to E. cyan- oganthus but globular, with long thinner tapering embolus, slightly bent apically (Fig. 9). Color (in alcohol): carapace golden brown woolly pubescence; abdomen golden brown with longer reddish setae; chelicerae golden brown; legs brown with narrow transverse yellow band between all femora and patellae. Paratype female, lengths: total body: 38.08; carapace, 14.88 long x 12.8 wide; abdomen, 16.32 long x 12.0 wide; chelicerae, 6.88 long x 3.36 wide; cheliceral teeth: (left side, right side) 10,13. Sternum: length 6.24, width 5.4; glabrous sigilla opposite coxae II and III, posterior sigilla the largest, 0.15 of sternum width. Labium: 2.46 long, 2.04 wide; labiosternal suture with distinct lateral mounds; ca. 198 cuspules in tight cluster; maxillae with ca. 184–207 cuspules. Fovea straight, deep. Cephalothorax: caput not rising abruptly from thoracic region but in gradual arch. Caput length 9.12, width 8.0. Ocular area 2.64 wide x 1.20 long, tubercle relatively low. Anterior eye row straight; AME round, diameter 0.67, apart 0.30; ALE elliptical, 0.26 x 0.58, apart 1.84. Posterior eye row slightly recurved; PME ovoid, 0.21 x 0.37, apart 1.51; PLE elliptical, 0.23 x 0.44, apart 1.98. Clypeus 0.47 wide. Leg span 105.0 mm. Leg formula I, IV, II, III (Table 2). All tarsal scopulae entire, without setal division. Metatarsal scopulae: I – II, entire; III, 0.73; IV, 0.44. Type V urticating hairs in dense pad on apical prolateral face of palpal femur. No plumose setae on palp or leg segments. Spination: Leg I, tibia 2v (1pa, 1ra); II, metatarsus 1v (ma), tibia 1v(pa); III, metatarsus 3v(1pa, 1ma, 1ra), tibia 2v(1pa, 1ra); IV, metatarsus 2v(1pa, 1ra), tibia 2v(1pa, 1ra); palp 4v(2pa, 2ra). Spermathecae, two tall columnar lobes, not widely separated, widest at base, abruptly narrowing then rounded apically, with raised folds (Fig. 10). Color: (in alcohol): chelicerae, carapace and abdomen rufus brown with scattered longer reddish brown setae on the abdomen; legs darker rufus brown with narrow transverse yellow bands between all leg joints, widest between femora and patellae.

TABLE 1. Leg and palp segment lengths in mm of Ephebopus foliatus, holotype male.

coxae trochanter femur patella tibia metatarsus tarsus total palp 4.0 2.4 7.2 2.72 6.88 - 2.56 25.76 legI 5.28 3.2 13.76 7.04 12.16 11.2 5.12 57.76 legII 4.32 3.04 12.0 6.08 10.08 10.4 4.96 50.88 legIII 3.68 2.4 9.92 4.8 8.0 10.56 4.32 43.68 legIV 4.0 2.88 12.64 5.28 11.2 14.4 4.32 54.72

Variation: Males: only the holotype. Females: 5 (including the paratype), length 29.28–36.16; carapace length 11.68–14.88, width 10.72–12.8; carapace width/length 0.86–0.92; chelicerae length 5.6–6.88, width

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 47 2.88–3.36; abdomen length 11.68–17.12, width 8.48–13.12; leg span 95.0–105.0 mm. Cheliceral teeth 9,10– 11,13. Metatarsal scopulae: I–II, entire; III, 0.70–0.77 (0.74); IV, 0.34–0.44 (0.39). Type V urticating hairs in dense pad on apical prolateral face of palpal femur. No plumose setae on palp or leg segments. Spermathecae, as in paratype. Distribution: Known only from type locality in Guyana (Fig. 30). Natural History: All specimens were collected at night from 1–4 m above ground on the sides of trees in riparian rainforest (Fig. 14). None were found on the ground or observed in silken retreats. We lack the field observations on the natural history of E. foliatus, however, collecting records indicate this species is arboreal (M. Kuntner, per. comm., USNM, 2002).

TABLE 2. Leg and palp segment lengths in mm for paratype female of Ephebopus foliatus.

coxae trochanter femur patella tibia metatarsus tarsus total palp 4.96 3.68 8.8 4.96 5.44 - 5.6 33.44 Leg I 6.4 3.2 12.48 7.36 9.6 8.96 4.8 52.8 Leg II 5.6 3.36 11.2 6.72 7.84 8.32 4.48 47.52 Leg III 3.84 2.88 9.12 5.12 9.12 8.0 4.16 42.24 Leg IV 5.28 3.2 11.52 5.6 9.44 11.62 4.16 50.88

Cladistics

Characters: In characters marked with an asterisk (*; viz., 6, 10, 12, 13, 18, 20, 31, 39, 42), fit and steps are not given as the character is not significant to the analysis but was kept here to indicate an autapomorphy for a terminal taxon. (0) Eye tubercle in males and females: (0) low (Fig. 23), (1) high (Fig. 24); fit = 66.6. Steps/ Extra steps = 4/3. The eye tubercle can be absent in some mygalomorphs (e.g., Atypidae, Rastelloidina, Raven 1985; Goloboff 1993b, 1995) or well-defined in most mygalomorph taxa. Intermediate states have been proposed for some cyrtaucheniids, hexathelids, ctenizids (Raven 1985:125, 70, 141; Bond & Opell 2002), and diplurids (Ischnothele Ausserer) (Bond & Opell 2002). In theraphosids, two states have been proposed: absent, for some Hemirrhagus Simon 1903, (Spelopelma Gertsch; Raven 1985) and distinct, for remaining species. An intermediate state was herein identified: Heteroscodra (Fig. 23) shows a somewhat flattened eye tubercle, with anterior lateral eyes being in almost same plane as anterior median eyes (state 0). Spiders with an arched eye tubercle (Fig. 24), with the anterior median eye positioned clearly in superior plane in relation to the anterior lateral eye were considered as having state 1. (1) Anterior row of eyes in males and females, measured from anterior margins: (0) procurved, (1) straight; fit = 85.7. Steps/ Extra steps = 2/1. (2) Clypeus in males and females: (0) absent, (1) narrow (Fig. 23) (2) wide (Fig. 24); fit = 60.0. Steps/ Extra steps = 6/4. The clypeus was considered absent when there is no space between the eye tubercle and the anterior carapace edge which can slightly project forward beyond the anterior carapace edge. A wide clypeus is found in some theraphosids, e.g. Harpactirinae, Ornithoctoninae and some theraphosines. In this case, the distance between the anterior edge of the eye tubercle and the anterior edge of carapace can be almost the same as the length of the eye tubercle (Fig. 24). The narrow state of the clypeus occurs in intermediate cases (Fig. 23). (3) Fovea in males and females, curvature: (0) straight, (1) recurved, (2) procurved; fit = 85.7. Steps/ Extra steps = 3/1. (4) Fovea in males and females, closure: (0) slit-like (closed), (1) pit-like (open); fit = 100. Steps/ Extra steps = 1/0. In Mygalomorphae, a pit-like (open) fovea seems to be plesiomorphic, since Mesothelae shows

48 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. this state as well as representatives of Atypidae, Antrodiaetidae and Dipluridae (Raven 1985). For theraphosids, Gallon (2003) suggested the presence of pit-like fovea to be a synapomorphy of the clade Xenoden- drophila Gallon 2003 (now Encyocratella Strand) (Stromatopelma + Heteroscodra). This proposal is followed here. (5) Fovea in males and females, depth: (0) shallow; (1) deep; fit = 66.6. Steps/ Extra steps = 4/3. This character is introduced here and refers to the depth of the fovea. Plaster was applied to the specimen's fovea and the resulting mould was analyzed. Those with deep fovea (state 1) presented a mould with trapezoid shape. Specimens with shallow fovea (state 0) showed a mould with a short line. (6) Labial cuspules in males and females, number*: (0) 30–300, (1) 0–20, (2) 350–450 Within the Thera- phosidae examined here, only Euathlus has a low number of labial cuspules. The most common state is between 30-300 labial cuspules and the only taxon with more than this number is Phlogiellus. The state 30– 300 was coded as 0 because it is present in the outgroups and thus should be the plesiomorphic state. (7) Sigilla, posterior pair in males and females, position: (0) marginal, less than 1.5 times its own diameter from margin, (1) located closer to center of sternum separated from margin by twice its own diameter; fit = 75. Steps/ Extra steps = 3/2. In Eumenophorinae, the posterior sternal sigilla are medially placed (Pocock 1897; Gallon 2003). We consider Phlogiellus and Haplopelma also have this state. (8) Tarsus IV in males, cracked: (0) cracked, (1) integral; fit = 85.7. Steps/ Extra steps = 2/1. A cracked tarsus is found in some mygalomorph taxa, e.g., some diplurids, nemesiids, barychelids and theraphosids (Raven 1985). In specimens examined, this condition was found in males and females of Melloina and males of Holothele and Phlogiellus. (9) Tarsal scopulae in males and females: (0) no true scopula; (1) scopula of sparse hairs; (2) dense scopula that does not extend much laterally; (3) scopula very extensive laterally, giving the tarsi and metatarsi I and II a spatulate appearance; fit = 85.7. Steps/ Extra steps = 4/1. Melloina has no true scopula (state 0) (Raven 1999). In Holothele and other Ischnocolinae, the scopula is very sparse (state 1), and frequently divided by rows of setae. Most theraphosids have well-developed scopulae (state 2), that, however, is not so developed as those having state 3, where their lateral development give the tarsi and the metatarsi I and II a spatulate appearance. This last state was used by Simon (1892) to characterize the Avicularieae and that was followed by most authors thereafter. (10) Scopulae on metatarsus IV in males and females*: (0) divided by hairs or spines; (1) not divided. Gallon (2003) considered the undivided scopula on metatarsus IV to be apomorphic for Eumenophorinae. The only taxon with the derived state in the analysis is Citharischius Pocock 1901. (11) Leg spines in males and females: (0) present on apex and other faces of tibiae and metatarsi, (1) present only on ventral apices of tibiae and metatarsi, (2) absent; fit = 75. Steps/ Extra steps = 4/2. The theraphosid sister-groups, e. g., Melloina and barychelids, have spines distributed largely on ventral, prolateral, retrolateral and sometimes dorsal faces of the legs. This suggests that the condition is plesiomorphic for Theraphosidae. Among the studied taxa, a lower number of leg spines is seen in Tapinauchenius, Psalmo- poeus, Ephebopus, Citharischius and Phlogiellus where there are only two small spines on the ventral apical edge of the tibiae and metatarsi. In Avicularia, Iridopelma, Pachistopelma, Stromatopelma, Heteroscodra and Poecilotheria, spines are totally absent. (12) Palpal femora in males and females, scopula on retrolateral face*: (0) absent, (1) present. Raven (1985) included Stromatopelma in the Eumenophorinae because they share the presence of a brush of hair on the retrolateral face of the palpal femora. Gallon (2003) stated that only Stromatopelma females present this scopulae and proposed its presence to be homoplastic with respect to Eumenophorinae. We failed to find this structure in males and females of Stromatopelma examined, and so it was here coded as absent . The only genus in the analysis clearly having the scopulae was Citharischius. (13) Femora IV in males and females, scopulae on retrolateral face*: (0) absent, (1) present.The scopulae on the retrolateral face of femora IV is an apomorphy of Lasiodora C. L. Koch 1850 and other theraphosine

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 49 genera not included in the analysis (Pérez-Miles et al. 1996). (14) Chelicerae in males and females, scopulae on retrolateral face: (0) absent, (1) present; fit = 85.7. Steps/ Extra steps = 2/1. (15) Maxillae, spiniform setae on lower prolateral face: (0) absent, (1) present; fit = 100. Steps/ Extra steps = 1/0. Spiniform setae on the lower prolateral face of maxillae are found in Ornithoctoninae and Thrig- mopoeinae (Raven 1985). We here consider that spines in the same position in Poecilotheria are putative primary homologs. (16) Stridulatory bristles in males and females form maxillae lyra: (0) absent, (1) present; fit = 75. Steps/ Extra steps = 3/2. The derived condition (1) of this character was used to characterize the Selenocosmiinae (e.g., Pocock 1895). It is also found in other mygalomorph taxa apart from Theraphosidae, e.g., the diplurids Diplura C. L. Koch 1850 and Trechona C. L. Koch 1850 (Raven 1985). (17) Stridulatory bristles on coxae I of males and females: (0) absent, (1) present; fit = 85.7. Steps/ Extra steps = 2/1. Stridulatory bristles on coxae were used to characterize the Eumenophorinae (e.g., Raven 1985) and some Theraphosinae genera such as Lasiodora, Grammostola Simon 1892 and Theraphosa Thorell 1870 (Pérez-Miles et al. 1996). (18) Longitudinal white stripes on patellae and tibiae of males and females*: (0) absent, (1) present (Fig. 11). The conspicuous white stripes in Ephebopus murinus were considered an autapomophy. (19) Leg rings on distal femora, tibiae and metatarsi of males and females: (0) absent, (1) white, (2), yellow (Figs 13–14, 16); fit = 85.7. Steps/ Extra steps = 3/1. Most theraphosid species have a band of short setae around the distal femora, tibiae and metatarsi of all legs, more visible on legs I and II. Some taxa, e. g. Holothele, have no detectable rings, whereas in most theraphosids, they are white, if present. Some species of Ephebopus and Avicularia have these rings but are yellow in color. (20) Chelicerae of females, hair: (0) not iridescent, (1) iridescent (Fig. 16). State 1 is an autapomophy of E. cyanognathus. (21) Black marking dorsally on tibiae, metatarsi and tarsi of males and females: (0) absent, (1) present; fit = 100. Steps/ Extra steps = 1/0. This character was introduced by Gallon (2003) and his proposal is followed here. (22) Female abdominal pattern: (0) all one color or irregularly mottled, (1) with pattern; fit = 60. Steps/ Extra steps = 5/4. Gallon (2003) considered that ornithoctonines, harpactirines and "stromatopelmines" (Het- eroscodra, Stromatopelma and Xenodendrophila, now Encyocratella) to have abdomen dorsums marked with a herring-bone pattern. We follow that here and include also Poecilotheria and E. foliatus as well as Avicu- laria, Pachistopelma, Iridopelma, Tapinauchenius and Psalmopoeus Pocock 1895 in which genera, imma- tures have a dorsal abdominal pattern. (23) Spermathecae, number: (0) two, completely separated, (1), two, fused at base, (2) one, totally fused; fit = 100. Steps/ Extra steps = 2/0. (24) Spermathecal lobes: (0) absent, (1) present; fit = 66.6. Steps/ Extra steps = 4/3. (25) Cymbium: (0) without spiniform process between lobes, (1) with spiniform process between lobes; fit = 100. Steps/ Extra steps = 1/0. This character was proposed by Raven 1985 as a putative synapomorphy of Aviculariinae. (26) Prolateral cymbium lobe shape: (0) rounded (Fig. 26), (1) subtriangular (Fig. 25); fit = 85.7. Steps/ Extra steps = 2/1. (27) Subtegulum: (0) small, (1) large, extending down the bulb for half of the tegulum (Raven 1985); fit = 85.7. Steps/ Extra steps = 2/1. A large subtegulum is considered a Theraphosinae synapomorphy (Raven 1985; Pérez-Miles et al. 1996). This character state 1 is also present in Haplopelma Simon 1892. (28) Embolus curvature: (0) straight (Figs 1, 3, 5, 7, 9), (1) curved to the retrolateral side (Fig. 25); fit = 66.6. Steps/ Extra steps = 4/3. (29) Embolus length: (0) 1.5-2.5 times longer than tegulum, (1) shorter than tegulum, (2) more than 3

50 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. times longer than tegulum; fit = 60. Steps/ Extra steps = 6/4. (30) Embolus distal width: (0) thin, less than 1/5 of tegulum height (Fig. 25), (1) thick, more than 1/3 of tegulum height (Fig. 26); fit = 85.7. Steps/ Extra steps =2/1 (31) Embolus shape*: (0) not flattened, (1) slightly flattened, (2) very flattened. The states follow Bertani (2000; 2001). Haplopelma was found to have a very flattened embolus, similar to some theraphosine species, such as Theraphosa, Megaphobema Pocock 1901 and Sericopelma Ausserer 1875. (32) Prolateral inferior keel of bulb: (0) absent, (1) present; fit = 100. Steps/ Extra steps =1/0. Several mygalomorph taxa have strongly developed keels on the male palpal bulb, e.g., species of Pycnothelinae (Nemesiidae), Barychelinae (Barychelidae) and Theraphosinae (Raven 1985). In Theraphosidae, the presence of keels was considered a synapomorphy of Theraphosinae (Raven 1985; Pérez-Miles et al. 1996). Bertani (2000) created a terminology for theraphosine male palpal bulb keels, which is followed here. However, other theraphosid taxa have strongly developed keels, such as species of Haplopelma and Poecilotheria. These keels are similar in position to some theraphosine species and thus they are being putatively considered primary homologs in the cladistic analysis. (33) Prolateral superior keel on embolus (Bertani 2000): (0) absent, (1) present; fit = 100. Steps/ Extra steps =1/0. (See character 32). (34) Apical keel on embolus (Bertani 2000): (0) absent, (1) present; fit = 85.7. Steps/ Extra steps = 2/1. (See character 32). (35) Retrolateral keel on embolus (Bertani 2000): (0) absent, (1) present; fit = 85.7. Steps/ Extra steps = 2/ 1. (See character 32). (36) Male tibial spurs: (0) present, two-branched, (1) present, with apical megaspine, (2) present, with spiniform setae, (3) absent; fit = 66.6. Steps/ Extra steps = 6/3. Male tibial spurs are a very homoplasious character in Theraphosidae (Bertani 2001) and most other mygalomorph families (Raven 1985). However, the presence of two-branched spurs seems to be plesiomorphic in Theraphosidae since it is present in all baryche- lid subfamilies and in Melloina (Paratropididae), the proposed sister-groups of Theraphosidae (Raven 1985). (37) Hairs on metatarsi and tibiae I–IV of males: (0) normal, (1) long hairs laterally projected, forming a brush; fit = 85.7. Steps/ Extra steps = 2/1. Males of Tapinauchenius, Psalmopoeus, Stromatopelma and Het- eroscodra all share the presence of hairs projecting laterally. (38) Urticating hairs of males and females: (0) absent, (1) on abdomen (Cooke et al. 1972), (2) on pedipalps (Marshall & Uetz 1990) (Fig. 19). Fit = 85.7. Steps/ Extra steps = 3/1. (39) Type I urticating hairs in males and females*: (0) absent, (1) present. (40) Type II urticating hairs in males and females: (0) absent, (1) present; fit = 100. Steps/ Extra steps = 1/ 0. (41) Type III urticating hairs in males and females: (0) absent, (1) present; fit = 100. Steps/ Extra steps = 1/0. Considered a synapomorphy of Theraphosinae (Pérez-Miles et al. 1996). (42) Type IV urticating hairs in males and females*: (0) absent, (1) present. (43) Type V urticating hairs in males and females: (0) absent, (1) present (Figs 19–20); fit = 100. Steps/ Extra steps = 1/0. Type V hair (Fig. 20) is present in Ephebopus murinus (Marshall 1990) and all known species of this genus. (44) Burrow entrance (fossorial) of female: (0) simple with little or no silk, (1) collar of silk bound with surrounding debris (Fig. 15), (2) trumpet-shaped (Fig. 11); fit = 85.7. Steps / Extra steps = 3/1. (45) Eggsac type: (0) mobile, (1) fixed hammock, (2) fixed flat; fit = 100. Steps / Extra steps = 2/0. Here we follow the classification of Gallon & Gabriel (2006) (46) Habitat: (0) evergreen forest, (1) deciduous forest, (2) desert-scrub; fit = 85.7. Steps/ Extra steps = 3/ 1. (47) Habits of female: (0) hide in surface layers of soil, (1) arboreal, (2) opportunistic (Figs 17–18), (3) fossorial (Fig. 11). Fit = 75. Steps / Extra steps = 5/2. State 0 was reported in paratropidids (Raven 1999).

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 51 52 · Zootaxa 1849 © 2008 Magnolia Press WEST ET AL. Most theraphosids make a burrow (state 3) or hide in retreats built in trees using silk (most aviculariines and Poecilotheria) (state 1). An opportunistic burrowing behavior (state 2) is found in Tapinauchenius, Psalmo- poeus and Ephebopus rufescens (RCW and SDM pers. obs.). They build a silken retreat which is not always made high in trees or other elevated structures. Furthermore, they commonly use soil and/or rocky substrata.

FIGURE 29. Single tree of Theraphosidae obtained with X-Pee-Wee, fit 3284.0, 117 steps.

Discussion

A single tree (fit = 3284.0, length = 117) was obtained with concavity 6 on X-Pee-Wee (Figs 29, Table 4). Searches with concavities 1 (fit = 2300.0, length = 120) and 2 (fit = 2693.3, length = 120) resulted in trees with the same topology of concavity 6 except for a basal position of Phlogiellus (appearing as the sister group of node 38), the resolution of node 25 [Euathlus {Lasiodora (Haplopelma + Poecilotheria)}], and the topology of Aviculariinae (node 36). With concavities 1 and 2, Ephebopus had a basal position in the node being the sister group of [Pachistopelma (Avicularia + Iridopelma)] + [(Stromatopelma + Heteroscodra)(Psalmo- poeus + Tapinauchenius)]. With concavity 3 (fit = 2929.9, length = 119) the topology is very similar to that found with concavities 1 and 2. The exception is the resolution of node 25 that is the same as that found with concavity 6. Concavities 4 (fit = 3085.7, length = 118) and 5 (fit = 3199.0, length = 118) resulted in a tree similar to that found with concavity 6 except for the internal resolution of Aviculariinae (node 36), which is the same found with concavities 1–3.

Taxa or node Character Change Taxa or Node Character Change Taxa or node Character Change Holothele 20→1 Poecilotheria 01→0380→1 30→122→1410→1 24 0→192→325 320→1 Pterinochilus 14 0→1110→2330→1 22 0→1160→126 0 0→1 36 3→1290→120→2 Citharischius 30→2460→127 0 0→1 50→1473→111→0 70→1 Avicularia 19 1→250→1 10 0→1 Psalmopoeus 16 0→128 250→1 11 0→1 Iridopelma 24 0→1280→1 12 0→1 Euathlus 22→1363→2 17 0→160→1380→1 24 0→1240→1400→1 Haplopelma 70→1420→129 111→2 14 0→1460→130 370→1 31 0→2 E. murinus 18 0→132 191→2 35 0→1 E. cyanognathus 20 0→134 0 0→1 36 3→221 4 0→1380→2 Phlogiellus 30→2210→1430→1 60→2370→135 5 0→1 70→1450→2363→0 16 0→122 450→136 9 2→3 Lasiodora 13 0→1460→237 1 0→1 17 0→123 150→1110→1 23 0→1220→138 190→1 29 0→1230→2260→1 31 0→124 5 0→1360→3 35 0→1261→0 39 0→1363→0

Whatever the concavities used, the node with Ephebopus species resulted in the same topology: E. foliatus + E. uatuman + E. cyanognathus in a monophyletic node and E. murinus and E. rufescens collapsed. Searches using Hennig86 with successive weights resulted in 4 trees (l = 428, ci = 80, ri = 83), the strict consensus tree being that found with X-Pee-Wee with concavity 6. Searches using equal weights resulted in 6 trees with Hennig86 (l = 117, ci = 56, ri = 66) and 2 trees with Nona (l = 117, ci = 56, ri = 66). The strict consensus tree found by both programs was the same and similar to that found with X-Pee-Wee. The difference was that with equal weights, all Ephebopus species are collapsed into a polytomy. The results were robust for all search strategies used, the main differences being the position of Phlogiel- lus, of Ephebopus respective to the other Aviculariinae genera (to be or not basal in the node) and the relationships among the Ephebopus species.

The discussion below is based on the preferred tree (Fig. 29) obtained with X-Pee-Wee and concavity 6. As pointed out by Ramirez (2003), concavities 1 and 2 normally give bizarre solutions, and equal weights gives low performance; better performance comes from those with the mildest weighting functions. Further- more, the chosen tree is the shorter one with higher fit. Ephebopus was found to be monophyletic based on the presence of type V urticating hair on the pedipalps (Figs 19–20) and an elevated eye tubercle (Fig. 24) (the latter having convergences with both nodes 26 and 27). In the present analysis, Ephebopus murinus and E. rufescens are part of a trichotomy with the clade having E. uatuman, E. cyanognathus and E. foliatus. These three species share the presence of yellow rings on the legs (Figs 13–14, 16). Ephebopus murinus has as autapomorphy, the presence of conspicuous white longitudinal stripes on the legs. Ephebopus cyanognathus has females with iridescent chelicerae (Fig. 16). Ephebopus foliatus and E. uatuman lacks autapomorphies. The present analysis confirms that Ephebopus belongs to the Aviculariinae (Fig. 29). However, the monophyly of the subfamily itself is weakly supported by a single character—the presence of well-developed scopula on tarsi and metatarsi very extended laterally, mainly those of legs I and II. This character state appears independently in Poecilotheria, an arboreal taxon. Since several aviculariine species are also arboreal, we

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 55 suggest from this shared trait that the evolutionary convergence would be due to similar habits. However, the function of scopula hairs on the ventral tarsi and metatarsi is not yet well understood. The composition of Aviculariinae obtained in this analysis is the same as that of Pocock (1901), including the genera Psalmopoeus, Tapinauchenius, Ephebopus, Stromatopelma, Heteroscodra, Avicularia, Iridopelma and Pachistopelma. According to the analysis, the sister group of Ephebopus is the node Tapinauchenius + Psalmopoeus. The node Ephebopus (Tapinauchenius + Psalmopoeus) is supported by two homoplasious characters: presence of deep fovea and of a two-branched spur on the ventral tibia I of males. The node Tap- inauchenius + Psalmopoeus is supported by the character “metatarsi and tibiae I to IV in males with long hairs projecting laterally, forming a brush”. This character is a parallelism with the node having the aviculariines Stromatopelma + Heteroscodra. The single autapomorphy for the Psalmopoeus is the presence of stridulatory bristles forming a maxillae lyra, whereas Tapinauchenius lacks any autapomorphy. The presence of a maxillae lyra is homoplasious, since in this analysis it appeared independently three times: in Phlogiellus, Poecilothe- ria and Psalmopoeus. Furthermore, some Psalmopoeus species have the lyra weakly developed, as is the case with the types of Psalmopoeus emeraldus Pocock 1903. A weakly developed lyra was also found in an undescribed species of Tapinauchenius (Figs 27–28). In fact, in Psalmopoeus and Tapinauchenius, the genitalia shapes are very similar and the distinction between Tapinauchenius and Psalmopoeus is evidently based solely on the presence or absence of a lyra. The existence of species having this weakly developed structure makes the differentiation between the two genera a gray area. There is even the possibility of Psalmopoeus and Tapinauchenius being a single genus with lyrate and alyrate species. The sister group of node 35 is node 29, weakly supported by a single character change: legs with segments ventrally having some small apical spines to legs being completely aspinose. Node 29 has two well-supported nodes, one with the African genera, Stromatopelma and Heteroscodra, and the other one with the New World genera Pachistopelma (Avicularia + Iridopelma). Stromatopelma and Heteroscodra share four apomorphies: fovea pit-like, black leg markings dorsally on tibiae, metatarsi and tarsi, metatarsi and tibiae I to IV in males with long hairs projecting laterally, forming a brush (parallelism with Psalmopoeus + Tapinauchenius) and a fixed flat eggsac type. Stromatopelma and Heteroscodra have been assigned to Aviculariinae, Eumenophori- nae, and Stromatopelminae. Gallon (2003) presented evidence of Stromatopelminae being monophyletic and comprising also Encyocratella. However, he did not compare these taxa with Aviculariinae, Theraphosinae or Ischnocolinae. The present analysis indicates that these two genera should be included in the Aviculariinae. Concerning Encyocratella olivacea Strand, the sole species of the genus, the data provided by Gallon (2003, 2005) shows a discrepancy with other aviculariines: presence of several leg spines, a male palpal bulb with three keels, an elevated eye tubercle and leg scopulae not very well-developed. The male also lacks the typical leg brush of hairs projected laterally, present in these two African genera and in the South American Psalmo- poeus and Tapinauchenius. Although African genera, Stromatopelma and Heteroscodra are found in tropical Western Africa, a region which was in broad contact with northeastern Brazil when they were part of Gond- wanaland. The node Pachistopelma (Avicularia + Iridopelma) shares five apomorphies: cymbium with a spiniform process between lobes, embolus curved to the retrolateral side (homoplasious), male tibial spur with spiniform setae (homoplasious in Haplopelma), urticating hairs on the abdomen dorsum (homoplasious in the Therapho- sinae), and presence of Type II of urticating hair. Avicularia and Iridopelma share the presence of an elevated eye tubercle (homoplasious), the anterior eye row procurved (homoplasious), and a deep fovea (homoplasious). The cladistic analysis carried out here, even though it included representatives of most theraphosid subfamilies, can only be considered preliminary. For a satisfactory analysis, it should include more representatives of each subfamily to test their monophyly, as well as include all ischnocoline genera.

We are grateful to A. Miles (Reuters Limited) for logistical assistance in French Guiana and G. Tavakilian of IRD (formerly ORSTOM) for his assistance in locating collecting sites. We are especially grateful for the lodging and kind hospitality provided by Joep and Marijka Moonen of Emerald Jungle Village. Permission to collect in the Trésor Reserve was granted by the Director of Tresor Reserve, Joep Moonen. The assistance of Jeremy Huff was invaluable in finding specimens in the field for this study. Thanks go to Drs. Matjaz Kuntner at the Slovenian Academy of Sciences and Arts, Slovenia, Jonathan Coddington at the National Museum of Natural History, Washington, and Christine Rollard at the Muséum National d'Histoire Naturelle, Paris, for loaning type material. H. D. Cameron kindly checked the Latin names. Support for RB: Fapesp 03/12587–4 and for CSF: Coordenação de Aperfeiçoamento do Pessoal de Nivel Superior – CAPES and Fapesp 06/ 58326–5. Drs Peter Jäger (SMF), Adriano Kury (MNRJ), Norman Platnick (AMNH), Osvaldo Villarreal, Ricardo Pinto da Rocha (MZSP), Mrs Janet Beccaloni, Mr Paul Hillyard (BMNH), are also thanked for material loans. Claudio Riccomini and Isaac Jamil Sayeg (USP-IGc) are thanked for the SEM photographs. Katia M. Faria kindly made the figures 23–26. Dr Martín Ramirez and the editor Dr Robert Raven made important suggestions to improve the manuscript.

References

Bertani, R. (2000) Male palpal bulbs and homologous features in Theraphosinae (Araneae, Theraphosidae). Journal of Arachnology, 28, 29–42. Bertani, R. (2001) Revision, cladistic analysis, and zoogeography of Vitalius, Nhandu, and Proshapalopus; with notes on other theraphosine genera (Araneae, Theraphosidae). Arquivos de Zoologia do Estado de São Paulo, 36, 265–356. Bond, J.E. & Opell B.D. (2002) Phylogeny and taxonomy of the genera of south-western North American Euctenizinae trapdoor spiders and their relatives (Araneae: Mygalomorphae: Cyrtaucheniidae). Zoological Journal of Linnean Society, 136, 487–534. Cambridge, F.O.P.- (1896) On the Theraphosidae of the lower Amazons: being an account of the new genera and species of this group of spiders discovered during the expedition of the steamship ‘Faraday’ up the River Amazons. Pro- ceedings of the Zoological Society of London, 716–766. Cooke, J.A.L., Roth, V.A. & Miller, F. H. (1972) The urticating hairs of theraphosid spiders. American Museum Novi- tates 2498, 1–43. Farris, J.S. (1988) Hennig86 reference. Documentation for version 1.5. Port Jefferson, New York. Gallon, R.C. (2003) A new African arboreal genus and species of theraphosidae spider (Araneae, Theraphosidae, Stro- matopelminae) which lacks spermathecae. Bulletin of the British Arachnological Society, 12, 405–411. Gallon, R.C. (2005) Encyocratella olivacea Strand, 1907, a senior synonym of Xenodendrophila gabrieli Gallon, 2003 (Araneae: Theraphosidae: Stromatopelminae) with a description of the male. Zootaxa, 1003, 45–56. Gallon, R.C. & Gabriel, R. (2006) Theraphosidae Egg-sac Types. Newsletter of the British arachnological Society, 106, 5–10. Goloboff, P.A. (1993a) Nona 2.0. Program and documentation available from www.zmuc.dk/public/Phylogeny/Nona- PeeWee/ (accessed 30 October 2006). Goloboff, P.A. (1993b) A reanalysis of mygalomorph spider families (Araneae). American Museum Novitates, 3056, 1– 32. Goloboff, P.A. (1995) A revision of the South American spiders of the family Nemesiidae (Araneae, Mygalomorphae). Part I: species from Peru, Chile, Argentina, and Uruguay. Bulletin of the American Museum of Natural History, 224, 1–189. Goloboff, P.A. (1997) X-Pee-Wee 1.3. Program and documentation available from www.zmuc.dk/public/Phylogeny/ Nona-PeeWee/ (accessed 30 October 2006). Lucas, S., da Silva Junior P.I. & Bertani R. (1991) The Brazilian species of the spider genus Ephebopus Simon, 1892 (Araneae, Theraphosidae, Aviculariinae), with descriptions of E. uatuman n. sp. Memórias do Instituto Butantan, 53, 161–166. Lucas, S., da Silva Junior P.I. & Bertani R. (1992) The genus Ephebopus Simon, 1892. Description of the male of Ephe- bopus murinus (Walckenaer, 1837). Spixiana, 14, 245–248. Marshall, S.D. & Uetz G.W. (1990) The pedipalpal brush of Ephebopus sp. (Araneae, Theraphosidae): evidence of a new

NEOTROPICAL TARANTULA GENUS EPHEBOPUS Zootaxa 1849 © 2008 Magnolia Press · 57 site for urticating hairs. Bulletin of the British Arachnological Society, 8, 122–124. Mello-Leitão, C.F. de (1920) An interesting new genus of Aviculariidae. Annals and Magazine of Natural History, 9, 141–143. Mello-Leitão, C. F. de (1923) Theraphosoideas do Brasil. Revista do Museu Paulista, 13, 1–438. Mello-Leitão, C. F. de (1930) Aranhas do Cuminá. Archivos do Museu Nacional do Rio de Janeiro, 32, 51–75. Papavero, N. (1973) Essays on the history of Neotropical Dipterology. São Paulo, Museu de Zoologia, Universidade de São Paulo, 2, 217–446. Pérez-Miles, F., Lucas, S.M., da Silva Junior P.I., & Bertani R. (1996). Systematic revision and cladistic analysis of Ther- aphosinae (Araneae: Theraphosidae). Mygalomorph, 1, 33–68. Platnick, N.I. (2008) The world spider catalog, version 8.5. American Museum of Natural History. Available from http:// research.amnh.org/entomology/spiders/catalog/index.html (accessed 28 March 2008) Pocock, R.I. (1895) On a new and natural grouping of some of the Oriental genera of Mygalomorphae, with descriptions of new genera and species. Annals and Magazine of Natural History, 6, 165–184. Pocock, R.I. (1897) On the spiders of the suborder Mygalomorphae from the Ethiopian Region, contained in the collection of the British Museum. Proceedings of the Zoological Society of London, 1897, 724–774. Pocock, R.I. (1901) Some new and old genera of South American Aviculariidae. Annals and Magazine of Natural His- tory, 7, 540–555. Pocock, R.I. (1903) On some genera and species of South American Aviculariidae. Annals and Magazine of Natural His- tory, ser 7, 81–115. Prentice, T.R. (1992) A new species of North American tarantula, Aphonopelma paloma (Araneae, Mygalomorphae, Theraphosidae). Journal of Arachnology, 20,189–199. Ramírez, M.J. (2003) The spider subfamily Amaurobioidinae (Araneae, Anyphaenidae): a phylogenetic revision at the generic level. Bulletin of the American Museum of Natural History, 277, 1–262. Raven, R.J. (1985) The spider infraorder Mygalomorphae (Araneae): cladistics and systematics. Bulletin of the American Museum of Natural History, 182, 1–180. Raven, R.J. (1990) Comments on the proposed precedence of Aphonopelma Pocock 1901 (Arachnida, Araneae) over Rhecostica Simon 1892. Bulletin of Zoological Nomenclature, 47, 126. Raven, R.J. (1999) Review of the mygalomorph genus Melloina Brignoli (Paratropididae: Araneae). Memoirs of the Queensland Museum, 43, 819–825. Reichling, S.B. & West R.C. (1996) A new genus and species of theraphosid spider from Belize (Araneae, Thera- phosidae). Journal of Arachnology, 24, 254–261. Roewer, C.F. (1942) Katalog der Araneae von 1758 bis 1940. Bremen, Kommissions–Verlag von "Natura", v. 1, 1040 pp. Schmidt, G.E.W. (1995) Eine weitere Tapinauchenius–Art aus Französisch-Guyana, Tapinauchenius purpureus sp. n. (Araneida: Theraphosidae: Aviculariinae). Arachnologisches Magazin 3, 11–17. Simon, E. (1892) Histoire naturelle des araignées. Paris, Librarie Encyclopédique de Roret, 1, 1–256. Simon, E. (1903) Histoire naturelle des araignées. Paris, Librarie Encyclopédique de Roret, 2, 669–1080. Walckenaer, C.A. (1837) Histoire naturelle des insectes. Apteres 1, Paris, Librarie Encyclopédique de Roret, 1–682. West, R.C. & Marshall S.D. (2000) Description of two new species of Ephebopus Simon, 1892 (Araneae, Theraphosidae, Aviculariinae). Arthropoda, 8, 6–14. West, R.C. & Marshall S.D. (2002) Description of two new species of Ephebopus Simon, 1892 (Araneae, Theraphosidae, Aviculariinae) [erratum]. Arthropoda, 10, 7.