The Philippine hair wax spiders and their relatives : revision of the Pholcus bicornutus species group ( Araneae , Pholcidae )

We revise the Southeast Asian Pholcus bicornutus group in which males are characterized by a unique pair of horns on their ocular area, each of which carries at its tip a brush of hairs. In two species, the two hair brushes are ‘glued’ or ‘waxed’ together by an unidentified substance into a very consistently curved and pointed single median tip. In the other five species known, the hairs are unglued. We present a first revision of ocular modifications in Pholcidae and identify twenty supposedly independent origins. Most cases are in Pholcinae, and all but one case are limited to the male, suggesting sexual selection as the main driving force in the evolution of ocular modifications in Pholcidae. Previously, the Pholcus bicornutus group consisted of four species limited to the Philippines. We describe four new species, including three species from the Philippines (P. olangapo Huber, sp. nov.; P. kawit Huber, sp. nov.; P. baguio Huber, sp. nov.) and the first representative from outside the Philippines (P. mulu Huber, sp. nov. from Sarawak, Ne borneo) and provide new records and SeM data for three previously described species.


Introduction
When eugène Simon returned from his expedition to Luzon in 1891, he brought with him, among others, a highly unusual pholcid spider species from caves near Manila.unlike any other pholcid known at that time, males had a pair of seemingly segmented horns at their ocular region.Simon was clearly struck by this feature, and devoted the only two figures in the original description of this new species Pholcus bicornutus Simon, 1892 to the male ocular region (Simon 1892: figs 3-4).One year later, in his epochal Histoire Naturelle des Araignées, he used one of these figures again (Simon 1893: fig.455).In both editions, the two-segmented horns are illustrated as being more or less parallel in frontal view, with separate tips.
More than 100 years later, when preparing the first revision of Pholcus Walckenaer, 1805 (Huber 2011a), one of us (bAH) studied Simon's type material of P. bicornutus.While one mystery of Simon's 'segmented horns' could be solved, another one appeared.Scanning electron microscopy revealed that what had looked to Simon like 'distal segments' were in fact brushes of hairs that originated at the tips of the 'basal segments'.Strikingly, however, the tips of these hair brushes were not separate as illustrated by Simon but appeared like 'glued' or 'waxed' together at their tips (Huber 2011a(Huber : figs 1566(Huber , 1568)).Simon had clearly interpreted this as an artifact, and the latter author tended in the same direction, especially because in all specimens of two newly described close relatives of P. bicornutus (P.arayat Huber, 2011;P. pagbilao Huber, 2011) the hair brushes on the otherwise very similar ocular horns did not look like they were glued together.In addition, sample sizes were small and no scanning electron images could be made of the two new species, so Huber (2011a) did not further dwell on the mysteriously joined hair brushes of P. bicornutus.
The present paper revises the Pholcus bicornutus group based on much larger samples than previously available and finally presents strong evidence that the 'glued' or 'waxed' hair brushes of P. bicornutus are not an artifact.First, in new material of P. bicornutus and of a seemingly very closely related species (P.olangapo sp.nov.), all twelve available males have the exact same configuration, with the hair brushes joined together and curved in a highly consistent way.Second, photographs of live males of P. olangapo sp.nov.show that the joined tips are not an artifact of fixation and preservation.And finally, in none of over 100 studied males of P. arayat, P. pagbilao, and three newly described species, do the hairs appear glued together in the light microscope, and this impression is supported by scanning electron micrographs of three of these species.

Material and methods
Most of the material studied herein was collected during recent expeditions to the Philippines (Feb.-Mar. 2014) and to northern borneo (Jul.-Aug. 2014).Further specimens came from a student project at Mindanao State university -Iligan Institute of Technology dealing with ecological aspects of pholcid diversity in the Philippines.The material is currently deposited at Mindanao State university -Iligan Institute of Technology (MSu-IIT), Philippines; Sarawak Museum, Kuching (SMK), Malaysia; and Zoologisches Forschungsmuseum Alexander Koenig, bonn (ZFMK), Germany.Additional material came from the California Academy of Sciences, San Francisco (CAS), u.S.A.; the Netherlands Centre for biodiversity Naturalis, Leiden (rMNH); and the Museum of Zoology, Turku university (ZMT), Finland.
Methods and terminology used are as in Huber (2011a).Measurements are in mm unless otherwise specified.eye measurements are approx.± 5 µm.epigyna were cleared in warm NaOH solution and stained with chlorazol black.For SeM photos, specimens were dried in hexamethyldisilazane (HMDS) (brown 1993) and photographed with a Hitachi S-2460 scanning electron microscope.SeM data are presented within the descriptions but are not based on the holotype specimens described.Locality coordinates are in round brackets when copied from labels and original publications or when received directly from collectors, in square brackets when originating from some other source (such as online gazetteers, Google earth, etc.).Distribution maps were generated with ArcMap 10.0.Pholcus -Huber 2011a: 124-126.

Natural history
Seven of the eight named species were observed in the field.They occupied a variety of microhabitats, ranging from caves (close to the ground; P. bicornutus) and wet rock walls (P.olangapo sp.nov.) to dark sheltered spaces among rocks (P.mulu sp.nov., P. baguio sp.nov., P. pagbilao), large holes at tree bases (P.kawit sp.nov.), abandoned buildings (P.pagbilao), and open space among vegetation exposed to direct sunlight (P.arayat).Webs mainly consisted of a domed sheet with a diameter of up to ~50 cm (P.mulu sp.nov.), but were much smaller in the rock-dwelling P. olangapo sp.nov.(~15 cm).In P. baguio sp.nov., webs were connected to each other and formed large communal structures up to 2 m in length.When disturbed, most species were observed to run toward the periphery of the web, seeking shelter at the substrate rather than staying and whirling in the web.

Distribution
The Pholcus bicornutus group is largely restricted to the Philippines, with only one species in Ne borneo (Figs 1-2).

Diagnosis (updated)
Distinguished from most similar known relative (P.olangapo sp.nov.) by absence of dark lateral bands on carapace (in males and females), and by procursus shape (prolatero-dorsal process of proximal part longer and more slender; entire distal hinged part of procursus longer and more slender; distinctive shapes of procursus tip and of uncus; figs 1560-1561 in Huber 2011a).From other species of the P. bicornutus group (P.pagbilao, P. arayat, P. schawalleri, P. baguio sp.nov., P. mulu sp.nov., P. kawit sp.nov.) by hinged procursus, small epigynal plate, undivided dark band ventrally on abdomen (also in P. kawit sp.nov.and P. mulu sp.nov.), and by presence of slightly curved hairs on legs (especially on tibiae and metatarsi 1-2).

Note
Simon (1892) described this species from two caves, "Cueva de San Mateo" and "Cueva de Antipolo".The first is possibly identical to what is now called "Pamitinan Cave" (14.731°N, 121.190° e) which in turn may be identical to what in Huber (2011a) is cited as "Montealban Cave" (correct spelling: "Montalban Cave").The second is presumably identical to what is now called "Mystical Cave" (14.606°N, 121.209° e).The distance between these two caves is just 14 km.However, as noted previously after examination of Simon's type specimens (Huber 2011a), males from the two caves differ slightly in details of the procursus, and the specimens from Mystical Cave are assigned tentatively to this species (the lectotype, designated in Huber 2011a, is from "Cueva de San Mateo").

Natural history
The spiders were found within the cave close to the cave entrance; no specimen was found deeper in the cave.The spiders built their domed sheets among rocks close to the ground.

Etymology
The species name is derived from the type locality; noun in apposition.
Color.Carapace pale ochre to light brown, with wide median dark brown mark including ocular area and connecting posteriorly to wide lateral marginal bands (Fig. 3); clypeus pale ochre; sternum brown, with three pairs of light ochre marginal marks at bases of coxae 2-4; legs light brown to ochre, dark rings subdistally on femora and tibiae and in patella area, tips of femora and tibiae lighter; abdomen ochregray, dorsally with several pairs of indistinct cuticular marks, with distinct internal darker marks visible through cuticle dorsally and laterally; ventrally with undivided wide brown band between gonopore and spinnerets.
Body.Habitus as in Figs 3 and 5; ocular area raised, each eye triad on additional short hump directed towards lateral (Fig. 12), with pair of long processes between eye triads, each with distal brush of hairs that are 'glued' or 'waxed' together to form consistently curved pointed tip (Figs 13,15); carapace without median furrow; clypeus unmodified; sternum wider than long (1.00/0.75),unmodified.ALS with one widened, one pointed, and six cylindrically shaped spigots of variable sizes (Fig. 21).Gonopore with four epiandrous spigots (Fig. 22).
CheliCerae.As in Fig. 9, with large lateral apophyses and low frontal humps proximally, dark distal apophyses near median line provided with two small modified (cone-shaped) hairs each (Fig. 20); without stridulatory ridges.

Female
In general similar to male but eye triads closer together (distance PMe-PMe 210 µm), without processes between eye triads (Fig. 14).Tibia 1 in 11 females from type locality: 8.6-10.5 (mean: 9.2).epigynum mostly weakly sclerotized, small posterior plate with small 'knob' ; some females with strongly protruding epigynal area (apparently expandable); internal genitalia as in Figs 11 and 26.Spinnerets as in Fig. 19, ALS with one widened, one pointed, and six cylindrically shaped spigots of variable sizes, PMS with two spigots each.The single female specimen from Mt. banahaw shares the carapace pattern and is thus tentatively assigned to this species rather than to P. bicornutus (in which the female genitalia appear indistinguishable; compare figs 1539 and 1564 in Huber 2011a with Figs 11 and 24 herein); tibia 1: 9.7.

Natural history
At the type locality, specimens were collected from small webs attached to a perpendicular, dripping wet rock wall beside a small stream (Figs 3-4).

Distribution
Known from two localities in Luzon only (single female specimen from Mt. banahaw assigned tentatively, see above; Fig. 1).Huber, sp. nov ) by large round uncus with pointed process (Fig. 40) and by large prolateral sclerite distally on procursus (Fig. 40); from most species (except P. kawit sp.nov.) also by undivided median dark band ventrally on abdomen (Fig. 37).From other species of the P. bicornutus group (P.bicornutus and P. olangapo sp.nov.) by unhinged procursus, presence of appendix, large epigynal plate, and by absence of curved hairs on legs.

Etymology
The species name is derived from the type locality, noun in apposition.
Color.Carapace pale ochre, with wide median brown mark including ocular area, without lateral marks (Fig. 36); clypeus not darkened; sternum dark brown; legs in live specimens bluish, in alcohol ochre to light brown, tips of femora and tibiae whitish; abdomen pale gray, dorsally with dark cuticular marks that are fused above spinnerets, with indistinct internal darker marks visible through cuticle dorsally and laterally; ventrally with undivided wide brown band between gonopore and spinnerets.
Body.Habitus as in Figs 36-37; ocular area raised, each eye triad on additional short hump directed towards lateral, with pair of processes between eye triads, each with distal brush of hairs; carapace without median furrow; clypeus unmodified; sternum wider than long (1.00/0.85),unmodified.
CheliCerae.As in Fig. 42, with small lateral and frontal apophyses proximally and dark distal apophyses near median line provided with two small modified (cone-shaped) hairs each; without stridulatory ridges.distinct ventral process; tibia large, with small but distinct ventral cavity (for proximal bulbal sclerite); procursus complex distally, with distinctive prolateral sclerite and membranous structures; bulb with large rounded uncus with pointed process, weakly sclerotized and distally widening embolus, appendix with small prolateral spine, distally curved towards retrolateral.legs.Without spines or curved hairs; few vertical hairs; retrolateral trichobothrium on tibia 1 at 3.5%; prolateral trichobothrium absent on tibia 1, present on other tibiae; tarsus 1 with >30 pseudosegments, only distally about 20 fairly distinct.

Natural history
Specimens were found in large domed sheet webs (diameter up to ~50 cm) among rocks near the ground.When disturbed they moved very rapidly towards the periphery of the web.

Diagnosis
easily distinguished from most similar known relatives (species with horns between eye triads carrying brushes of unglued hairs and with simple, unhinged procursi: P. arayat, P. pagbilao, P. schawalleri, P. baguio sp.nov., P. mulu sp.nov.) by distal modifications of procursus (dorsal pointed process; Fig. 46); from most species also by female genitalia (epigynum large trapezoidal plate -Fig.30; similar in P. pagbilao and P. mulu sp.nov.); from most species (except P. mulu sp.nov.) also by undivided dark median band ventrally on abdomen (Fig. 39); from P. pagbilao also by simpler appendix (Fig. 45); from P. arayat also by wider abdomen (Figs 38-39); from P. schawalleri also by more slender uncus (Fig. 45).From other species of the P. bicornutus group (P.bicornutus and P. olangapo sp.nov.) by unhinged procursus, presence of appendix, large epigynal plate, and by absence of curved hairs on legs.

Etymology
The species name is derived from the type locality; noun in apposition.
Color.Carapace pale ochre to orange, with wide median dark brown mark including ocular area and small brown submarginal marks laterally (Fig. 38); clypeus light brown; sternum monochromous light brown to orange, labium darker; legs light brown to ochre, dark rings subdistally on femora and tibiae and in patella area, tips of femora and tibiae lighter; abdomen ochre-gray, dorsally with several pairs of dark cuticular marks that are fused above spinnerets, with distinct internal darker marks visible through cuticle dorsally and laterally; ventrally with undivided wide brown band between gonopore and spinnerets.Body.Habitus as in Fig. 38; ocular area raised, each eye triad on additional short hump directed towards lateral, with pair of short processes between eye triads, each with distal brush of hairs; carapace without median furrow; clypeus unmodified; sternum wider than long (1.05/0.95),unmodified.ALS with one widened, one pointed, and three to four very small conical spigots (~1.5-3 µm wide and ~3-6 µm long).
CheliCerae.As in Fig. 47, with small lateral and frontal apophyses proximally and dark distal apophyses near median line provided with two small modified (cone-shaped) hairs each; without stridulatory ridges.
PalPs.As in Figs 45-46; coxa unmodified; trochanter with short conical retrolateral process and longer ventral apophysis with distinctive tip; femur with finger-shaped retrolateral process proximally and distinct ventral apophysis; tibia with dark dorsal band, also ventrally darkened, with small but distinct ventral cavity (for proximal bulbal sclerite); procursus complex distally, with distinctive pointed process dorsally (more sclerotized and at different angle than in P. baguio sp.nov.); bulb with large uncus, weakly sclerotized wide embolus, simple appendix with main branch curved towards retrolateral and shorter prolateral side-branch.legs.Without spines or curved hairs; few vertical hairs; retrolateral trichobothrium on tibia 1 at 4%; prolateral trichobothrium absent on tibia 1, present on other tibiae; tarsus 1 with >30 pseudosegments, only distally about 10 fairly distinct.

Female
In general similar to male but sternum dark brown, eye triads closer together (distance PMe-PMe 230 µm), without processes between eye triads.Tibiae 1 missing.epigynum large trapezoidal plate surrounded by whitish cuticle (Fig. 30); with anterior 'knob'; internal genitalia as in Figs 32 and 49.

Natural history
Adult specimens were only found in large holes at tree bases near a brook in the forest.Juveniles were more common in any dark sheltered spaces along the brook.

Distribution
Known from type locality in Mindanao only (Fig. 2).Huber, sp. nov

Etymology
The species name is derived from the type locality; noun in apposition.
Color.Carapace ochre to orange, with wide median brown mark including ocular area and clypeus (except rim), connecting posteriorly with lateral submarginal brown bands (Fig. 50); sternum monochromous light brown to orange, labium darker; legs light brown, dark rings subdistally on femora and tibiae and in patella area, tips of femora and tibiae lighter; abdomen ochre-gray, dorsally with several pairs of dark cuticular marks that are fused above spinnerets, laterally with indistinct internal darker marks visible through cuticle; ventrally with wide brown band divided into anterior and posterior parts separated by light V-shaped space.
Body.Habitus as in Fig. 50; ocular area raised, each eye triad on additional short hump directed towards lateral (Fig. 63), with pair of long processes between eye triads directed towards posterior, each with distal brush of hairs curved upwards, hairs not 'glued' together ; carapace without median furrow (Fig. 64); clypeus unmodified; sternum wider than long (1.10/0.85),unmodified.ALS with one widened, one pointed, and three very small conical spigots (Fig. 74).Gonopore with four epiandrous spigots.
CheliCerae.As in Fig. 60, small lateral and frontal apophyses proximally and dark distal apophyses near median line provided with two small modified (cone-shaped) hairs each (Fig. 73); without stridulatory ridges.

Female
In general similar to male but sternum dark brown, eye triads closer together (distance PMe-PMe 250 µm), without processes between eye triads.Tibia 1 in 16 females: 9.6-11.6(mean 10.7).epigynum relatively small, roughly oval plate (Figs 33,75) with pair of dark brown areas and anterior 'knob' weakly sclerotized in most specimens; entire epigynal area strongly protruding in some specimens; internal genitalia as in Figs 35 and 62. Spinnerets and spigots as in male (Fig. 76).

Natural history
The spiders were found in partly extremely high densities in a dry brook bed near the road, among rocks and shaded vegetation.The individual domed sheets were interconnected to form large communal webs of up to 2 m length.

Diagnosis (updated)
Distinguished from most similar known relatives (species with horns between eye triads carrying brushes of unglued hairs and with simple, unhinged procursi: P. pagbilao, P.

Description (amendments)
As noted in Huber (2011a), the holotype is apparently artificially darkened.The new material supports this assumption: carapace in males and females light brown with darker median band (sometimes indistinct; in females medially divided) and lateral bands (sometimes barely visible); sternum light brown to orange in both males and females.Hairs on male ocular processes not 'glued' together 80).Tip of procursus extremely complex .Male palpal tibia with small but distinct ventral cavity.Distal male cheliceral apophyses with two modified (cone-shaped) hairs each (Fig. 85).Male gonopore with four epiandrous spigots (Fig. 86).Tarsus 4 comb-hairs as in Fig. 87.ALS in both sexes with one widened, one pointed, and three very small conical spigots (Figs 88,90).

Variation
Tibia 1 in 15 newly examined males: 10. 1-13.5 (mean 11.9); in 12 newly examined females: 9. 2-11.1 (mean: 10.3).The specimens from bohol are overall very similar to those from Luzon but the male differs in some details of the palp: tip of trochanter apophysis not so strongly directed proximad; distal semitransparent process on procursus not clearly bifid; distance between tip of uncus and subterminal process larger.They are thus assigned tentatively.

Natural history
The spiders were found both in sheltered spaces (among rocks and logs) and among vegetation, with their domed webs sometimes directly exposed to the sun.

Diagnosis (updated)
Distinguished from most similar known relatives (species with horns between eye triads carrying brushes of unglued hairs and with simple, unhinged procursi: P.

Variation
As noted in Huber (2011a), the tip of the procursus differs slightly between males from Negros Island and males from Luzon.Interestingly, males from bohol Island resemble those from Luzon much more closely than those from neighboring Negros.In addition, there is variation among males from Negros Island: males from Mabinay (see Huber 2011a) differ slightly from the newly collected males from Casaroro Falls.However, all these specimens share the distinctive slender semitransparent process distally on the procursus and the shapes of uncus and appendix (cf.figs 1579-1580 in Huber 2011a).Tibia 1 in 46 newly examined males: 9.6-15.3(mean 12.7) (42 of these males: 11.2-14.0); in 38 newly examined females: 9.3-12.8(mean: 11.2).The black ventral band of the abdomen is always clearly divided into anterior and posterior parts but in some specimens (of both sexes) the tip of the posterior part has a narrow connection to the anterior part.Some of the newly collected specimens from bohol are assigned tentatively because they are not accompanied by males and because the ventral abdominal pattern is partly intermediate between that of P. pagbilao and P. arayat.Some of these specimens may actually be conspecific with the specimens assigned tentatively to P. arayat.

Natural history
The spiders were found in sheltered spaces close to the ground, usually among rocks.In the abandoned buildings near the park entrance of Mt.Isarog, the spiders were found in high numbers in all corners at any height from the ground.

Distribution
Apparently widely distributed in the Philippines (Fig. 2).

Discussion
The male ocular horns with brushes of hairs described herein represent a unique synapomorphy of the Pholcus bicornutus group (Huber 2011a), but ocular area modifications in general are widespread in Pholcidae and come in a wide range of shapes (Table 1).Two aspects are particularly remarkable about the overview in Table 1.First, with only one exception (the 'pseudo-lenses' accompanying the secondary eyes in Smeringopus Simon, 1890 and Smeringopina Kraus, 1957) all modifications are restricted to the male, i.e., represent sexual dimorphisms.Second, these modifications appear very unevenly distributed among major taxonomic groups.No case is known in Ninetinae; only six cases are known in Arteminae, Modisiminae, and Smeringopinae together.All other 14 cases are in Pholcinae.each of the twenty cases listed in Table 1 is supposed to represent an independent origin.This is derived from available phylogenies (reviewed in Huber 2011b; see also Huber 2013;Dimitrov et al. 2013;Huber & Nuñeza 2015;Huber et al. 2015) and from partly fundamental differences in the quality of the modifications.We admit that some particular modifications may have originated more than once (e.g., the small median elevations in Pholcus youngae Huber, 2011 andPholcus schwendingeri Huber, 2011; or the pair of pointed processes near the PMe in the Pholcus minang group and in Panjange Deeleman-reinhold & Deeleman, 1983), while others may not represent independent origins (e.g., the pointed horns in the Pholcus ancoralis group and the sculptured horns in the Pholcus calligaster group).
An exact count will require phylogenies with denser taxon sampling and better resolution (which is currently in preparation; J. eberle, A. Valdez-Mondragón, D. Dimitrov & b.A. Huber, unpubl. data), but it seems very unlikely that the number of independent origins will change dramatically.
unfortunately, we know much less about the function of pholcid ocular modifications.In fact, there is only one study (Huber 1997) that deals in some detail with a sexually dimorphic ocular region.
In Modisimus culicinus (Simon, 1893), the male ocular turret is provided anteriorly with a cuticular lobe that is densely set with hairs and with pores through which large glands in the male ocular area discharge their products.During copulation, the female mouthparts contact this cuticular lobe, suggesting gustatorial courtship (Huber 1997).At first sight it would seem that such a gustatorial courtship function might be a valid hypothesis for further cases: first, the copulatory position of Pholcidae consistently brings the male ocular area close to the female mouthparts; and secondly, dense sets of hairs occur on the male ocular areas in many other Pholcidae (actually in most Pholcus species; Huber 2011a).However, glandular pores have never been found on the male ocular areas in any other species, even though many were examined in detail with the scanning electron microscope.This leaves a wide range of possible explanations, including mechanical functions in an intersexual selection context as well as intrasexual selection (male-male contest) scenarios.Obviously, this is a rich field for future observational and experimental studies.