A new species of Hortipes ( Araneae , Corinnidae ) , the first spider with an insertable retrolateral tibial apophysis on the male palp

Hortipes gigapophysalis (araneae, corinnidae) is a new species described from both sexes from montane forest on Mt nimba, eastern Guinea. The species is remarkable for its long, whip-shaped retrolateral tibial apophysis (RTa) on the male palp. The structure apparently has an insertable function as the epigyne of the female contains a separate set of ducts starting from a central concavity that is unique in the genus. This duct system is apparently meant to receive the supple RTa. This type of structural arrangement has never previously been found in spiders.


Introduction
With almost 70 known species, Hortipes Bosselaers & ledoux, 1998 is one of the largest afrotropical genera (Bosselaers & Jocqué 2000a).Most of the species live in the litter layer of dense forests and woodlands, hence its very wide distribution from West africa to ethiopia in the north and to large parts of south africa in the south.However, the genus is entirely absent from miombo woodland and is therefore an excellent example of an afrotropical spider genus with this type of distribution.The genus is also remarkable as an example of the range of complexity of genitalia that is met with in many spider genera.In Hortipes, the genitalia range from quite simple structures as in H. silvarum ledoux & emerit, 1998 to complex configurations such as in H. sceptrum Bosselaers & Jocqué, 2000a.The genus can therefore be considered as suitable for testing the evolution of genitalia in the context of sexual selection hypotheses (e.g.eberhard & Huber 1998;Huber 1995a;Jocqué 2002).
The present paper describes a new species in the genus that has a remarkable male palp provided with a very long retrolateral tibial apophysis (RTa) unlike any that has been found in spiders so far.

Diagnosis
Males of H. gigapophysalis are easily recognized by the long, winding RTa; the female is characterized by the central depression and the separate duct systems in the epigyne.These characters are unique in the genus.

Etymology
The species is called gigapophysalis in view of the extraordinarily long retrolateral tibial apophysis on the male palp.
genitaLia.epigyne with a central concavity and with convoluted tubes visible in transparency.Two funnel-shaped openings leading into a system of wide, very simple, thin-walled tubes, with blind ends (Fig. 2c-F).Thick-walled insemination ducts starting from slit-shaped lateral openings leading to two sets of spermathecae.anterior pair oval, opening posteriorly into ducts leading towards posterior, globular pair.The systems are separate and there is no connection between them (Figs 2c-F, 3a-B).

Distribution
High altitude gallery forest (> 1200 m) on Mt nimba, Guinea.

Discussion
The RTa is usually a solid, strongly sclerotized structure.Its main function was long thought to be internally stabilising the expanded male palp during copulation, which means that the RTa locks either specific sclerites of the bulbus or between the bulbus and cymbium in order to arrest further rotation or movement (Huber 1995a: 151;sierwald & coddington 1988: 264).However, the RTa has been found to fulfill many different functions.Huber (1994) mentions four different functions for the RTA in one single spider family (agelenidae).In Textrix denticulata (Olivier, 1768), the RTA fixes the male palp to the female epigyne; in Histopona torpida (c.l.Koch, 1834) the role of the RTa is unclear, since the patellar apophysis arrests the male bulbus internally; in Agelena gracilis and Agelenopsis spp. the RTa locks the male palp internally in two different ways, as described below (Gering 1953;osterloh 1922).Jäger (2006: 58) confirms this diversity in function, showing that the RTA anchors in the epigastric furrow in sparassidae but anchors in the epigyne in Gnaphosidae.Huber (1995a) points out that the function of the RTA is highly species specific and may be involved in mate choice by physical contact during copulation (1995a: 152), rather than internally stabilising the male palp.Huber (1995aHuber ( , 1995b: 698, fig. 5: 698, fig. 5) demonstrated that the RTA is used to fix the male pedipalp to the female epigyne for six species from six different families of the RTa-clade (coddington & levi 1991).Huber (1995a: 159-160) also states that most data inferring that the RTa arrests the male bulb internally during copulation are doubtful for methodological reasons.Either the authors used artificially expanded genital bulbs, or they studied genital bulbs fixed in copula, but sprung away from the epigyne.The vast majority of research using reliable methodology (observations of animals in copula, sections of animals fixed in copula) led to the conclusion that the RTA fixes the male pedipalp to the female externally.Such fixation to the female epigyne may be necessary because of the lack of innervation of the male bulbus (eberhard & Huber 1998).
Huber cites the studies of osterloh (1922) and Gering (1953) as exceptions to this generalisation, and lists them in his table 1 as an RTa that arrests the genital bulb internally in Agelena gracilens c.l. Koch, 1841 and Agelenopsis spp.Giebel, 1869, respectively.More specifically, Osterloh (1922: 348) states that the RTa is inserted between the palpal bulbus ("Kapsel") and the palpal terminal sclerites ("stema") in copula, and that only the terminal sclerites ("Alles, was zum Stema gehört") are introduced into the epigyne.locking of the male palpus to the epigyne in these two species is performed by the patellar apophysis, not the RTa (1922: 390).Gering (1953: 13) states: "The ectoproximal margin of the cymbium has a weakly produced depression, which normally is somewhat more heavily pigmented and sclerotized than the surrounding area (figs 11-12, 39).The tibial process rests in this depression when the palpus is locked."We can conclude that neither author clearly advocates that the RTa is introduced into the internal female genital organs.In Gnaphosidae the RTa, a "primary anchoring device to the epigynal pocket" (Senglet 2004: 87) does not take part in intromission, which takes place after the palpus is fixed (senglet 2004: 91).
However, in the present contribution, an RTa is described which has exactly this very unusual function: intromission into the female genital organs.To the best of our knowledge, such a use of the RTa has never been mentioned before.as the RTa of H. gigapophysalis sp.nov. is very long, thin and supple, it is very unlikely that it is used as a rigid locking device as described above.Its analogy with the embolus, a long whiplike, supple structure, made us suspect that it might be an insertable sclerite.detailed study of the epigyne indeed reveals the presence of two independent tubular systems.The first is the usual insemination duct starting at a lateral slit-like entrance and leading to two sets of spermathecae (Fig. 2c, d).The second is composed of wide, very simple, thin-walled tubes starting from funnelshaped entrances in a central concavity, ending blind at the end of several bends without connection with the first system (Fig. 2E, F).It is assumed that the second system is meant to receive the RTA.
Hortipes is a genus with an amazing genitalic variability.nevertheless, not a single species with an RTa that is clearly insertable in the female genitalia has been described to date.among the known species (Bosselaers & Jocqué 2000a;ledoux & emerit 1998), only a few species of the orchatocnemis clade (node 12 in Bosselaers & Jocqué 2000a, fig.4) have vulvae that, in view of their weakly sclerotised entrance chamber (Bosselaers & Jocqué 2000a: 14, character 65) show possible evidence of additional sclerites being inserted during copulation.H. merwei Bosselaers & Jocqué, 2000, for example, has an elaborate lobed entrance chamber (Bosselaers & Jocqué 2000a: fig.20e) that seems especially suited for insertion of the exceptionally large, flexible and coiled MA of the male palp (Bosselaers & Jocqué 2000a: fig. 19a, b).Jäger (2012: figs 19, 23) describes Sinopoda vulvae with similar thin-walled sacs.In this genus, these sacs proved to be expandable, but Jäger does not mention any role for them.The entrance chamber of H. aelurisiepae Bosselaers & Jocqué, 2000, also has peculiar thin-walled lobes, lateral in this case, that might receive an additional palpal sclerite from the unknown male.However, none of the Hortipes species described to date have an RTa of considerable length that matches a vulvar structure in which it could be inserted.Moreover, as a result of the fact that the Hortipes species of western africa are very poorly sampled (Bosselaers & Jocqué 2000a: map 1), no intermediate stages are known between the genitalia of the species already described and the insertable RTa of H. gigapophysalis sp.nov.Based on leg spination and genitalic structure, H. gigapophysalis sp.nov.seems closest to H. hesperoecius Bosselaers & Jocqué, 2000 and H. silvarum.Repeating the cladistic analysis of Hortipes species known from both sexes (Bosselaers & Jocqué 2000a), including H. gigapophysalis sp.nov., confirms this (Fig. 5): H. gigapophysalis sp.nov.turns out to be the sister species of H. silvarum and is situated at the base of the Hortipes clade in the strict consensus of 18 trees of fit -53.061 (PAUP values), obtained under implied weighting (Fig. 5).In the strict consensus of 1656 shortest trees of length 265 obtained under equal weighting, the salticola clade (node 3 in Bosselaers & Jocqué 2000a, fig. 4) holds a still more basal position within Hortipes, but the sister species relationship between H. silvarum and H. gigapophysalis is confirmed.However, although H. silvarum has a relatively slender and apically coiled RTA (Bosselaers & Jocqué 2000a: fig. 12e, f), it has no vulvar tubes in which the RTA could be inserted (Bosselaers & Jocqué 2000a: fig. 13d).Only more extensive sampling of the forest litter layer of tropical western africa may produce the as yet unknown Hortipes species allowing to reconstruct the phylogeny of this most unusual genitalic development.