The Ebo-like running crab spiders in the Old World ( Araneae , Philodromidae )

A recent phylogenetic analysis within Philodromidae has shown that Ebo Keyserling, in its current limits, is a paraphyletic assemblage of spiders characterized by a strongly elongated second pair of legs and by enlarged anterior median eyes. Here a generic revision of Ebo-like philodromid spiders is provided, with the genera Ebo, Titanebo Gertsch (re-elevated to genus rank), Halodromus gen. n. and Philodromus Walckenaer ad. part. (the histrio species group = Rhysodromus Schick) being redefi ned and diagnosed. Ebo and Titanebo are Nearctic taxa whose occurrence in the Old World remains doubtful. Old World species with a long patellar apophysis on the male palp are included in Halodromus gen. n. (H. patellaris (Wunderlich, 1987), H. patellidens (Levy, 1977), both ex. Ebo). Th ree new species are described from both sexes, Halodromus barbarae sp. n. from the Arabian Peninsula, Egypt and Spain, H. deltshevi sp. n. from Yemen, and H. gershomi sp. n. from Eritrea. Ebo eremus Levy, 1999 is a new subjective synonym of Halodromus patellaris (Wunderlich, 1987). Halodromus is presumably an Afro-Syrian element with wide distribution in the Eremial of northern Africa and the Middle East. Th e Israeli species Philodromus halophilus (Levy, 1977), comb. n. ex. Ebo is transferred to the Philodromus histrio species group.


Introduction
Ebo is a diverse genus of philodromid spiders in North America (Keyserling 1884;Sauer and Platnick 1972).Th eir characteristic trait is the conspicuously elongated second pair of legs, which can be more than twice as long as the other legs.Further diagnostic characters include a fl attened prosoma which is wider than long or as wide as long, and the eye confi guration: AME (anterior median eyes) are distinctly larger than the other eyes, both rows slightly recurved, medians of both rows closer to the laterals than to each other (Schick 1965;Sauer and Platnick 1972;Levy 1977).Based on this set of characters, Tikader (1965), Levy (1977Levy ( , 1999)), Wunderlich (1987) and Lyakhov (1992) included new Philodromidae from India, Israel, the Canary Islands and the mountains of South Siberia in Ebo.However, even a loose examination of the genital traits (cf.Fig. 1 versus Figs 2, 3) shows that the Old World Ebo species have little in common with the Nearctic representatives.Furthermore, the above mentioned combination of characters is also found in other taxa of Philodromidae, e.g. in most species of the Philodromus histrio group (Dondale and Redner 1975;Szita and Logunov 2008).Th us, it seems that elongation of leg II and enlargement of the AME coevolved independently several times within the family, and Ebo apparently became a paraphyletic taxon through inclusion of the Old World species.Th is view is corroborated by a thorough examination of the leg spination pattern.Recent comprehensive studies within Philodromidae have shown that leg spination, in particular that of the tibiae and metatarsi, is presumably the most useful morphological criterion for delineation of supraspecifi c taxa in this family (Muster 2009).Concerning leg spination in Ebo, I found not only the Old World species to be more similar to spiders of the Philodromus histrio group (= Rhysodromus Schick, 1965) than to North American congeners, but also striking diff erences between the Nearctic subgenera Ebo and Titanebo Gertsch, 1933.A fi rst quantitative cladistic analysis of phylogenetic relationships within Philodromidae (Muster 2009) provided further evidence: Ebo emerged as the most basal clade within the family, while Titanebo grouped with Rhysodromus, Old World Ebo and the Th anatini in a clade that was sister to the remaining Philodromus species.Th e aim of this study is a generic revision of the Ebo-like philodromid spiders (those with markedly elongated leg II and enlarged AME) in order to achieve a better concordance of genera with phylogenetic lineages.A group of Old World species with an elongated apophysis at the patella of the male palp is described as a new genus Halodromus gen.n.A taxonomic revision of the included species is provided.

Material and methods
Material from the following institutions and private collections was examined (institutional abbreviations follow Evenhuis 2007): AMNH American Museum of Natural History, New York CB Coll.Robert Bosmans, Gent  Specimens were examined and measured using a Zeiss STEMI 2000 stereoscopic microscope with a micrometer eyepiece.All measurements are in millimetres.For leg measurements, the variation is given for the entire leg followed by average values for the leg segments femur, patella, tibia, metatarsus and tarsus in squared brackets.Th e following morphometric indices turned out to be most useful for delineation of the Ebo-like genera: (i) the Leg II length index, which is the length of femur II divided by length of femur I, (ii) the AME size index, which is the diameter of AME divided by the width of the cephalothorax, (iii) the PME interdistance index, which is the distance between the PME divided by the distance PME-PLE (Schick 1965: 8), and (iv) the clypeus height index, which is the height of the clypeus divided by the width of the cephalothorax.For leg spination, the system of Ono (1988) was adopted.As the most distal pair of ventral spines is often shifted dorsally to a rather lateral position, this distal pair was ascribed to the ventral pairs of spines up to a lateral shift of 90°.
Male and female genitalia were dissected and studied as temporary mounts in Hoyer's solution (Kraus 1984) under a Nikon Eclipse E600 microscope with a drawing tube at the MNHN.Montage images were taken at the Zoological institute of the University of Greifswald with a Leica DFC 320 digital camera operating on a Leica Stereomicroscope MZ125.Images taken at diff erent focal planes were assembled with the Helicon Focus Pro software (Helicon Soft Limited).SEM images were taken by the team of Peter Michalik (University of Greifswald) using the following protocol: specimens were dehydrated in graded ethanols, dried in a BAL-TEC CPD 030 critical point dryer using amylacetate as the intermedium, coated with gold-palladium in a Quorum Technologies SC7620 sputtering device and examined in a Leo DSM 940A scanning electron microscope.
Terminology of the genital organs largely follows the recent studies by Muster et al. (2007) and Szita and Logunov (2008).Glandular heads are treated as synonymous with spermathecal organs.As the use of the term "Bursa copulatrix" has been controversial in the past, I here applied a rigid defi nition as proposed by Muster (2009): the "bursa copulatrix" is a canal or three-dimensional region of the vulva that is passed by the embolus during copulation before it enters the receptacula and that is not connected with the duct leading to the glandular heads.If the intromittent canal is merged with the ducts of the glandular heads, this structure is called a "copulatory duct".In this meaning "bursa copulatrix" is synonymous with "insemination duct" as used by Szita and Logunov (2008).Th e white line in Fig. 5 demarcates the transition from bursa copulatrix to copulatory ducts according to this defi nition.Th e various parts of the cephalothorax were named according to Schick (1965: 9).
Remarks.Th e Nearctic Ebo species were thoroughly revised by Schick (1965) and Sauer and Platnick (1972).In both these studies Titanebo described by Gertsch (1933) was regarded a subgenus of Ebo.All Nearctic species indeed share a number of characters, e.g. the lack of PTA, the curved embolus with a fl exible tip and the constitution of the conductor.However, the leg spination pattern is strikingly diff erent, with Titanebo resembling Rhysodromus (the Philodromus histrio group) and Halodromus gen.n.Ebo and Titanebo are also diff erent in the structure of the female genitalia and in some morphometric indices (Table 1).In none of the phylogenetic analyses performed by Muster (2009) did Titanebo emerge as sister to Ebo.In conclusion, Titanebo clearly deserves generic re-establishment.Composition and distribution.Th e centre of diversity is undoubtly North America.Sauer and Platnick (1972) included seven Nearctic species in the subgenus Ebo: E. contrastus Sauer & Platnick, 1972, E. evansae Sauer & Platnick, 1972, E. iviei Sauer & Platnick, 1972, E. lathithorax Keyserling, 1884, E. merkeli Schick, 1965, E. pepinensis Gertsch, 1933, and E. punctatus Sauer & Platnick, 1972.Two further species were added by Platnick (1972) andCokendolpher (1978), E. bucklei Platnick, 1972, andE. redneri Cokendolpher, 1978.Old World species described in Ebo are either transferred to Halodromus gen.n. or Philodromus, or they are regarded together with three Neotropical species as incertae sedis.
Remarks.Titanebo is herewith re-elevated to full genus status; for a justifi cation see remarks on Ebo.Titanebo species from North America were thoroughly revised by Schick (1965) and Sauer and Platnick (1972).
Remarks.Th e eastern-Palaearctic species of the Philodromus histrio group were recently reviewed by Szita and Logunov (2008), but delineation of the taxon remained vague and no evidence for monophyly of the included species was provided.Th e histrio group in its current limits is most likely an artifi cial assemblage of superfi cially similar species.A recent cladistic analysis within Philodromidae (Muster 2009) has shown that histrio and related species are more closely related to the Th anatini (Th anatus and Tibellus) than to other Philodromus species groups.Th us, Rhysodromus clearly deserves re-elevation to genus rank, but the urgently required revision of the histrio group from a phylogenetic perspective is beyond the scope of this study.
Composition and distribution.Szita and Logunov (2008) include 16 species from the eastern Palearctic in the Philodromus histrio group.In the Nearctic the taxon is represented by three polytypic species (Schick 1965;Dondale and Redner 1975).Th e group also occurs in the Mediterranean and in (northern) Africa, but species from this region still await a proper revision.Philodromus halophilus (Levy, 1977) comb.n. is transferred from Ebo to this group.Th e male lacks a patellar apophysis and the morphometric indices fall well within the range of the histrio group.
Composition and distribution.Five species from Northern Africa (including the Canary and Cape Verde islands) and the Middle East (one presumably reaching the Iberian Peninsula) are included in the new genus.Two of them were hitherto placed in the genus Ebo, three species are newly described.While all species are present in the region around the Red Sea, three are rather widespread in the area outlined above (Fig. 37).Diagnosis.Males are characterized by the shape of the embolus ("foxtailed") and by the RTA reduced to a small bulge (Fig. 17).Females show a unique shape of the receptacula (Fig. 20).
Colour.Pale species.Dorsal shield of prosoma (Fig. 14) light brown with yellowish median band that extends to posterior margin, posterior edges with whitish pubescence, allatum with radiating, furcated stripes of dark spots, whitish patches along longitudinal allatal stripes, metadiscus a central whitish W-sign, mesodiscus with inconspicuous pattern.Clypeus whitish, chelicerae uniformly beige.Sternum uniformly whitish with long pubescence.Legs yellowish-brown, mottled (particularly at prolateral-ventral side of femora), faint annulations distal at femora and basal and distal at tibia.Opisthosoma whitish grey with lanceolate cardiac mark, some dark patches and four to six chevrons in posterior half.Venter whitish grey.
Pedipalp .Patella with shortest apophysis among all known congeners, barely half as long as tibia.Tibia with RTA reduced to a low bulge.Cymbial tip less than one third of CyL.Cymbium length (CyL) 0.44-0.58,width (CyW) 0.2-0.3,20).Median septum with almost parallel margins, more than three times as long as wide, stretched grooves at both sides of median septum.Posterior guide pockets wide at epigastric furrow, anterior guide pockets small, moderately sclerotized.Receptacula kidney-shaped, in close contact.Glandular heads in lateral-distal position.
Remarks, distribution and habitat.Specimens from the Middle East are considerably smaller than those from Cartagena.However, there are no diff erences in the structure of the male and female genitalia, consequently there is no reason to doubt their conspecifi ty.Th e occurrence on the Iberian Peninsula needs to be confi rmed, since the original labels of the type series from the Simon collection contained no other information than "13388 Cartagena!".Th is locality is somehow suspect because it is the single record of the genus from Europe.Given the wide range of H. patellaris and H. patellidens, this distribution is not entirely implausible, but at the same time it can not be excluded that Simon referred to the ancient city of Carthage in Tunisia.Th e species may also occur on the eastern Canary Islands.Schmidt and Krause (1996) illustrated the epigynum of a female from El Jable, Fuerteventura (mounted epigynum lost, Schmidt and Krause 1996: 268) that they provisionally allocated to Ebo patellaris, but they referred to diff erences from the original description by Wunderlich (1987).Th is fi gure corresponds well to H. barbarae, while it certainly does not show H. patellaris.No details on habitat were available from any of the original labels accompanying the specimens.Diagnosis.Males are characterized by the prominent bulge at the base of the patellar apophysis , females are unique in the shape of the epigynal grooves (Fig. 23).
Colour.Dorsal shield of prosoma brown with orange-brown median band that extends to posterior margin, posterior edges with whitish pubescence, allatal stripes discontinuous, metadiscus a central whitish V-sign, mesodiscus with conspicuous pattern (similar to Fig. 16).Clypeus brown with whitish patch, chelicerae light brown with black spots.Sternum whitish with dots and brown patches at the margin.Legs yellowish-brown, mottled and with double annulations at femora, single annulations at patel- lae, threefold annulations at tibiae and weak annulations at metatarsi.Opisthosoma densely covered with whitish pubescence, interspersed with dark setulae.Dorsum grey with lanceolate cardiac mark, two posterior dots are usually fused with the cardiac mark, fl anks darkish.Venter whitish with grey patches and sometimes grey median stripe.
Pedipalp .Patella with bifi d apophysis: a rounded bulge at ventral base and a long (three quarters the length of the tibia), blunt dorsal process.Tibia with short, rectangular processing RTA with rounded tip.Cymbial tip approximately one third of CyL.Cymbium length (CyL) 0.66, width (CyW) 0.34, ratio CyL/cephalothorax width 0.4.Subtegulum visible in ventral view.Anterior border of tegulum indistinct, PTA originating in central position, hooked, tip pointing to RTA.Sperm duct symmetric, opening in 7:30 o'clock position.Conductor largely hidden behind embolus, in ventral view protruding at retrolateral side of embolus.Embolus at transition from basal to distal part with distinctive prolateral bulge that projects above prolateral margin of bulb, embolus tip a thin, slightly curved spur.
Epigyne-vulva .Epigyne longer than wide.Median septum narrowed posteriorly, epigynal sutures sigmoid.Epigynal grooves almost as wide as median plate at epigastric furrow.Posterior guide pockets relatively small, anterior guide pockets externally only moderately sclerotized, in dorsal view visible as voluminous pockets.Receptacula bagpipe-shaped, separated from epigastric furrow by approximately their diameter.Glandular heads in inner-distal position, almost touching one another.
Distribution and habitat.Only known from the type locality (Fig. 37), no information on habitat available.Etymology.Th e species is named in honour of Dr Gershom Levy, who described the fi rst species of the genus from Israel and unfortunately passed away during preparation of this manuscript.Noun in genitive case.
Diagnosis.Males are distinguishable from similar species by the shape of patellar apophysis and PTA .Th e epigynum is wider than long with exceptionally wide posterior guide pockets (Fig. 27).
Colour.Dorsal shield of prosoma (Fig. 16) brown with wide orange-brown median band that extends to posterior margin, median band wider than the dark sides which show blackish venation, posterior edges with whitish pubescence, allatal stripes discontinuous, metadiscus a central whitish V-sign, mesodiscus with inconspicuous pattern.Clypeus brown with bi-humped beige area, chelicerae brown with black spots, distally more light.Sternum whitish with dots and brown patches at the margin.Legs yellowish-brown, mottled and with double annulations at femora, single annulations at patellae, threefold annulations at tibiae and weak annulations at metatarsi.Opisthosoma densely covered with whitish pubescence, interspersed with dark setulae.Dorsum grey with conspicuous black cardiac mark, followed by a separated black dot and grey chevrons posteriorly, fl anks darkish.Venter whitish with a conspicuous, grey median stripe.
Pedipalp .Patellar apophysis less than two thirds the length of the tibia, with inconspicuous groove at ventral border, tip rounded, pointing to dorsal margin of tibia.Tibia with rectangular processing RTA with rounded tip which merges with a second, more dorsally situated bulge.Cymbial tip relatively long and narrow, approximately one third of CyL.Cymbium length (CyL) 0.6, width (CyW) 0.3, ratio CyL/cephalothorax width 0.36.Subtegulum visible in ventral view.PTA in central position, hooked, tip pointing in ventral-proximal direction, not reaching lateral margin of cymbium.Sperm duct symmetric, opening in 7 o'clock position.Conductor largely hidden behind embolus, in ventral view protruding at both sides of embolus.Embolus at transition from basal to distal part with distinctive prolateral bulge that projects above prolateral margin of bulb, embolus tip a thin, almost straight spur.
Epigyne-vulva .Epigyne distinctly wider than long.Median septum triangular, narrowed posteriorly, epigynal sutures straight.Epigynal grooves large.Posterior guide pockets exceptionally wide and well developed, anterior guide pockets externally moderately sclerotized, in dorsal view visible as voluminous pockets.Receptacula bagpipe-shaped, somewhat separated from epigastric furrow, not touching each other.Glandular heads in inner-distal position, pointing in lateral-distal direction.
Epigyne-vulva .Epigyne wider than long.Median septum almost quadrangular, epigynal sutures short.Epigynal grooves covered by lateral plates.Posterior guide pockets small, anterior guide pockets prominent, heavily sclerotized, with almost parallel slits, extending anteriorly far beyond receptacula.Receptacula at epigastric furrow, separated from each other by their diameter.Glandular heads in innerdistal position, with short ducts pointing in lateral-distal direction.
Remarks.Due to the serious illness of Dr Gershom Levy I failed to receive the type series of Ebo eremus Levy from the Hebrew University of Jerusalem for examination.However, the species was suffi ciently characterized by Levy (1999Levy ( , 2007) ) to propose synonymy with Halodromus patellaris (Wunderlich).
Distribution and habitat.Th e species is known from the eastern Canary Islands Fuerteventura (Wunderlich 1987(Wunderlich , 1992) ) and Lanzarote, from the Negev desert in southern Israel (Levy 1999(Levy , 2007)), and from Tunisia (Fig. 37).All specimens from Israel were taken using pitfall traps.Th e specimens from Tunisia were beaten from prostrate halophytic shrubs in saline habitats at both coastal and inland sites.Th ey were perfectly camoufl aged in the dense vegetation.(Levy, 1977) Figs 13, 33-36 Ebo patellidens Levy, 1977: 210-212, fi gs 36-39 (description  Diagnosis.Males are characterized by the exceptionally large patellar apophysis and by the large PTA that projects beyond the retrolateral margin of the cymbium .Females are recognizable by the coiled anterior guide pockets and the shape of the receptacula .
Colour.Dorsal shield of prosoma (Fig. 13) brown with bright orange-brown median band that extends to posterior margin, allatum with dark radiating stripes and whitish patches along longitudinal allatal stripes, metadiscus a whitish V-sign, mesodiscus with characteristic pattern.Clypeus with a beige area, chelicerae orange brown with dark spots.Sternum whitish with numerous small spots.Legs orange-brown, intensely mottled and with double annulations at femora and threefold annulations at tibiae.Opisthosoma densely covered with whitish pubescence, interspersed with dark setulae.Dorsum grey with conspicuous black cardiac mark that broadens in posterior third, followed by two dark spots and some partially fused chevrons, fl anks darkish, in posterior half with four to fi ve chevrons and two lateral dark patches.Venter greyish marbled with three faint longitudinal stripes.
Pedipalp .Patella exceptionally long and voluminous, longer than tibia.Tibia with almost rectangular processing RTA with rounded tip.Cymbial tip relatively short, less than one fourth of CyL.Cymbium length (CyL) 0.6-0.68,width (CyW) 0.35-0.36,ratio CyL/cephalothorax width 0.4-0.47.Subtegulum visible in ventral view.Anterior border of tegulum indistinct, PTA large, laminar, projecting be- Remarks.Due to the serious illness of Dr Gershom Levy I failed to receive the type series of Ebo patellidens Levy from the Hebrew University of Jerusalem for examination.However, the detailed fi gures provided by Levy (1977) allow the unequivocal identifi cation of the examined material with this species.
Distribution and habitat.Th e available evidence suggests that H. patellidens is rather widespread in the Middle East, but also along the north-African coast (Fig. 37).
Th e occurrence of H. patellidens on the Cape Verde islands (Schmidt 1990;Schmidt and Krause 1995) was doubted by Wunderlich (1992: 504-505), yet it could be confi rmed through examination of material from the SMF.On the other hand, the species has not been recorded from the Canary Islands.Th e record from Fuerteventura (Schmidt 1990) results from the erroneous synonymization with H. patellaris (already rejected by Wunderlich 1992: 504).Habitat information is scarce, but Schmidt and Krause (1995) collected the species from halophytes.

Incertae sedis
Th e following species, originally described in Ebo, show striking diff erences in genitalic and/or somatic characters compared to Ebo latithorax and its congeners from North America.However, these species are diffi cult to place in the system, and at the current state of knowledge I refrain from suggesting new generic assignments.

Remarks.
No information on leg spination and morphometric indices is given in Tikader (1965Tikader ( , 1971Tikader ( , 1980)).Without re-examination of the type series the relationship of this species cannot be inferred.

Discussion
A close examination of morphometrics, leg spination as well as male and female genitalia has clearly demonstrated the polyphyletic character of Ebo as hitherto delineated.Th e Nearctic subgenera Ebo and Titanebo are distinct enough to deserve genus rank.Th eir occurrence in the Old World remains uncertain, as the placement of Ebo bharatae from India and Ebo distinctivus from the Altai Mountains requires further studies.Th e remaining species from the Old World are either transferred to the Philodromus histrio group or to the new genus Halodromus, which is characterized by at least one putative autapomorphy, the long apophysis at the patella of the male palp.Th e presence of a patellar apoyphsis is very unusual within Philodromidae, only males of the Philodromus subgenus Locupletes Schick, 1965 also bear a short retrolateral projection at the palpal patella (Schick 1965).Th e proposed transfers are a contribution to the eff ort of making more genera of Philodromidae monophyletic.However, phylogenetic relationships among the Ebo-like philodromid spider genera remain largely unresolved, as diff erent character sets provide confl icting evidence.Concerning male genitalic traits, Ebo and Titanebo appear closely related.Leg spination is similar in Titanebo, Halodromus and Rhysodromus, but strikingly diff erent in Ebo.Female genitalia are unique in Titanebo, but almost indistinguishable between Halodromus and Rhysodromus.Morphological and biogeographic considerations suggest a rather close relationship of Halodromus and Rhysodromus, but it is not clear whether Halodromus is sister to Rhysodromus or rather a distal clade within Rhysodromus.Reconstructions are further hampered by the problematic delineation of Rhysodromus/the Philodromus histrio group.In the limits of Szita and Logunov (2008) it is most probably a paraphyletic assemblage.Th e Ebo-like philodromid genera recognized in this study have the following biogeographic origins.Schick (1965) identifi ed Ebo as an "Austral element", with a distribution centre south of the Nearctic coniferous forests.Ebo s. str. is widely distributed in temperate North America, while Titanebo is an eremial element with primary centre of distribution in the deserts of the southwestern Nearctic ("Sonoran element").Halodromus is also allocated to the eremial fauna, with a distribution centre in the Afro-Syro-Eremial regions south of the Mediterranean.A remarkable result of this study is the uncovering of the vast Afroeremial distribution of several species that have been considered narrow endemics of Israel or the Canary Islands respectively.Rhysodromus species are widely distributed from boreal to eremial regions in the Holarctic.Th e highest diversity is recorded from the eastern Palaearctic (Szita and Logunov 2008), but a better knowledge of the African and South-Asian philodromid fauna may result in deeper insights.
of the host Christine Rollard (MNHN) contributed to the success of this project.I would like to thank Peter Michalik (University of Greifswald) for access to technical resources in his department and for taking the SEM photographs.Th is study would not have been possible without the generous loan of specimens from the following persons and institutions: Robert Bosmans (CB), Efrat Gavish-Regev (TAU), Ambros Hänggi (NHMB), Peter Jäger and Julia Altmann (SMF), Barbara Th aler-Knofl ach (CTh ), Norman Platnick (AMNH), Christine Rollard and Elise-Anne Leguin (MNHN), and Jörg Wunderlich (CJW).Access to copies of rare and old literature was given by courtesy of Peter Jäger (SMF).Th anks are due to Christophe Herve (MNHN) for his help with identifying labels and localities from the Simon collection.I am deeply indebted to Dmitri Logunov (the Manchester Museum) for the loan of comparative material, fruitful discussion and comments on an earlier draft.Th e comments of two anonymous referees helped to improve a former version of the manuscript.