2002. The Journal of Arachnology 30:588±600

COURTSHIP AND SPERM TRANSFER IN THE WHIP SPIDER PHRYNUS GERVAISII (, ): A COMPLEMENT TO WEYGOLDT'S 1977 PAPER

Alfredo V. Peretti: CONICET- CaÂtedra de Diversidad I, Facultad de Ciencias Exactas, FõÂsicas y Naturales, Universidad Nacional de CoÂrdoba. Avda. VeÂlez Sars®eld 299 (5000), CoÂrdoba, Argentina. E-mail: [email protected]

ABSTRACT. The aim of this study was to provide descriptive and quantitative data regarding behaviors involved in courtship and in sperm transfer of the whip spider Phrynus gervaisii (Pocock 1894) in order to complete the previous description for this same species given by P. Weygoldt. The specimens were captured in anthills of Paraponera clavata, on Barro Colorado Island, Panama. Ten courtship and ®ve sperm transfer sequences were recorded. Four out of ®ve mating sequences with sperm transfer ocurred between adults with similar body size and in the other case the female was smaller than the male. Sexual interactions did not occur between very small adults. Two male behavior patterns that have not been reported were observed during the initial stage of courtship: ``pedipalp rubbing'' and ``female operculum rubbing''. Contrary to Weygoldt's description, in this study the female never performed ``shaking'' move- ments with her antenniform legs. It was observed that the two distal horn-like extensions of the sper- matophore facilitate the females movements during the sperm transfer. The distal part of the spermatophore stalk provides a suspension area when the female rests on those horns. It was veri®ed that the female can move the claw-like sclerites of the gonopods in all directions. The male executed copulatory courtship and successfully transferred sperm in ®ve analyzed sequences. The female did not pick up the sperm packages when copulatory courtship was not performed. Males that lacked one antenniform leg were able to mate, however they had to perform vibrations more intensely with their non-injured leg for a longer duration. The data are compared with those previously obtained in other whip spiders. Some functional characteristics of the spermatophore and female genitalia of P. gervaisii are also discussed.

RESUMEN. El objetivo de este trabajo es aportar datos descriptivos y cuantitativos sobre patrones de comportamiento que ocurren durante el cortejo y transferencia espermaÂtica del amblipõÂgido Phrynus ger- vaisii (Pocock 1894) con la ®nalidad de completar la descripcioÂn previa de P. Weygoldt para esta misma especie. Los especõÂmenes fueron capturados en hormigeros de Paraponera clavata, en la Isla de Barro Colorado, PanamaÂ. Se registraron 10 secuencias de cortejo y cinco de transferencia espermaÂtica. Cuatro de las cinco secuencias de apareamiento con transferencia espermaÂtica completa ocurrieron entre adultos de tamanÄo corporal similar mientras que en el otro caso la hembra fue maÂs pequenÄa que el macho. No se produjeron interacciones sexuales entre adultos muy pequenÄos. Durante la etapa inicial del cortejo fueron observados dos patrones de comportamiento masculinos que no habõÂan sido citados con anterioridad: ``roces de pedipalpos'' y ``roces al opeÂrculo genital femenino''. Al contrario de la descripcioÂn de Weygoldt, en el presente estudio la hembra nunca realizo movimientos de ``latigueo'' con sus patas anteniformes. Se observo que las dos expansiones distales con forma de cuerno del espermatoÂforo facilitan los movi- mientos de la hembra durante la transferencia espermaÂtica. La parte distal del tallo del espermatoÂforo ofrece un aÂrea de suspensioÂn cuando la hembra se apoya sobre estos cuernos. Se veri®co que la hembra puede mover los escleritos en form de unÄa de sus gonoÂpodos hacia todas las direcciones. El macho efectuo cortejo copulatorio en cinco secuencias analizadas, en ellas la transferencia espermaÂtica fue existosa. Por el contrario, la hembra no recogio los paquetes espermaÂticos cuando no existio cortejo copulatorio. Los machos que carecõÂan de una pata anteniforme tambieÂn fueron capaces de aparear. Sin embargo, ellos tuvieron que realizar maÂs intensamente las vibraciones con sus patas no danÄadas, y sobre todo durante un tiempo maÂs prolongado para evitar que la hembra se alejara. Se comparan los datos aquõ registrados con aquellos previamente obtenidos en otros amblipõÂgidos. Se discuten algunas caracterõÂsticas funcionales del espermatoÂforo y genitalia femenina de P. gervaisii. Keywords: Phrynus gervaisii, Amplypygi, courtship, sperm transfer, spermatophore

Whip spiders (Amblypygi) represent an in- natural history of which are still poorly teresting group of , the biology and known. Whip spiders have strong, raptorial

588 PERETTIÐMATING BEHAVIOR IN PHRYNUS GERVAISII 589 pedipalps armed with sharp spines. Since goldt (1990, etc.) has mentioned the presence whip spiders continue to molt and grow after of antenniform tapping along with other male- reaching sexual maturity, adults of the same female contact during the sperm pick-up, it is species may be different sizes (Weygoldt neccesary to add more details to increase our 1995). The courtship behavior and spermato- knowledge of this behavior (e.g., do all the phore morphology are two of the most fasci- males of a same species always perform cop- nating subjects to be analyzed in these ani- ulatory courtship?, can any male traits, such mals. Whip spiders use vibration of the as body size, affect copulatory courtship be- ``anteniform'' ®rst pair of legs during court- havior?). ship (Thomas & Zeh 1984; Weygoldt 1990). The aim of this study was to provide de- After a prolonged dance, the male turns until scriptive and quantitative data regarding be- facing away from the female and deposits a haviors involved in courtship and in sperm stalked spermatophore. Then he turns towards transfer of the whip spider Phrynus gervaisii the female again and lures her to approach the (Pocock 1894) in order to complement the spermatophore and pick up the spermatozoa previous general description given by Wey- (Weygoldt 1990, 1998). Few amblypygid spe- goldt (1977) for this species. Weygoldt stud- cies have been observed (i.e., Weygoldt and ied the mating behavior and spermatophore of others have studied the courtship of 19 out of Tarantula palmata Kraepelin 1899, using approximately 125 described species). Al- adults specimens captured in Santa Martha, though the literature contains general descrip- Colombia. Quintero (1981), in the important tions of the mating, we lack many descriptive revision of the amblypygid genus Phrynus in and quantitative details of behaviors involved the Americas, synonymized this species with in each stage of the courtship and sperm trans- Phrynus gervaisii. Thus, in this study Wey- fer. In fact, the elaborate displays involved in goldt's observations are adopted as the most the courtship and spermatophore deposition of important and direct antecedents, comparing whip spiders show the importance of more de- the data with those published by this author tailed observations not only to increase our and, mainly, adding new details to that general general knowledge but also to lend insight description, in particular with respect to be- into the sexual selection that shaped the dis- havioral variation. plays. Spermatophore morphology and sperm METHODS transfer mechanisms vary among genera and Capture site and laboratory condi- families and provide useful characters for sys- tions.ÐTwenty females and 13 males of P. tematics (Weygoldt, 1998). One of these fam- gervaisii were collected in October, 1996 in ilies, the Phrynidae, contains medium to lar- Barro Colorado Island (Smithsonian Tropical ged-sized whip spiders and occurs in tropical Research Institute -STRI), Panama, where the and semitropical areas of the Americas (Quin- observations were carried out under labora- tero, 1981). The spermatophores of phrynids tory conditions (see details below). The spec- are large and have triangular, heavily sculp- imens were captured in the morning and in the tured heads (Weygoldt 1969, 1974, 1977, early afternoon in anthills of Paraponera cla- 1990, 1999). The female genitalia are char- vata (Hymenoptera, Formicidae) inside the acterized by the existence of two gonopods, parcel of 50 hectares that STRI has on the each equipped with a claw-like sclerite that is island. Externally, these anthills consist of ex- used to pick up the small sperm packages cavations at the base of a tree. Phrynus ger- from the spermatophore (Weygoldt 1990, vaisii lives inside galleries of the anthills, go- 1999). It is currently unclear how these claw- ing out at night to feed (PeÂrez Mendieta 1996; like sclerites come into contact with the sperm Peretti, pers. obs.). In order to capture packages. Similarly, it would be of interest to specimens I hit the base of a tree containing understand the behaviors, such as copulatory anthills several times with a metal rod. This courtship, that occur during and after the made many ants to run out very quickly and sperm transfer. In arachnids copulatory court- made some whips spiders leave the anthills ship has been observed in spiders (Eberhard and climb the tree; immediately afterward the 1994), scorpions and solpugids (Peretti 1997, specimens were captured (by covering them pers. obs.). Although in whip spiders Wey- with plastic containers 9.6 cm in diameter). 590 THE JOURNAL OF ARACHNOLOGY

The were housed individually in cag- stained with methylene blue and examined es of different sizes, all furnished with a piece with a light microscope. Genitalia of live and of tree bark set vertically so that they could ®xed (in alcohol) females were compared and climb. They were fed crickets and locusts ®t with unused spermatophores in order to de- once a week. Moist cotton balls were used to termine the correspondence among the struc- maintain high humidity and the temperature tures that are closely in contact during the varied from 24±32 ЊC. Following Weygoldt & sperm transfer. Finally, each individuals will- Hoffmann (1995), all animals were kept under ingness to perform a new mating was ob- a light/dark cycle of 12:12 hours. In the pre- served daily. sent work the dark phase started at 1100 h. One preserved individual has been depos- This synchronized the activity rhythms of the ited as a voucher specimen at the Arachno- animals with each other and with our activity. logical Scienti®c Collection of the CaÂtedra de Analysis of mating and genitalia.ÐOb- Diversidad Animal I, Facultad de Ciencias servations of the mating sequences were made Exactas, FõÂsicas y Naturales, Universidad Na- in a 40x30x30 cm mating arena that was fur- cional de CoÂrdoba and one additional speci- nished with a vertical piece of tree bark from men was deposited at the Smithsonian Tropi- the capture site. The substrate consisted of cal Research Institute, Panama. soil, bark and some stones. All the elements of the arena were replaced with fresh materi- RESULTS als prior to each use. A 40 W red lamp al- Body size and mating.ÐIn the anthills lowed observation without disturbance. Oral containing individuals of P. gervaisii (n ϭ 28 recordings of 10 mating sequences were made of 37 anthills checked) there were one or two on audio cassettes; seven were described in adult whip spiders (®ve cases), but they were more detail and ®ve had complete sperm never of the same sex. These individuals var- transfer. These matings were always per- ied in size, but all of them were sexually ma- formed using different males and females. The ture (e.g., large and small females with off- female was introduced ®rst into the arena and spring were collected). The individuals mated 10±15 min later, the male was introduced. in the lab with others captured in either the Three males lost the distal part of one of their same anthill or elsewhere. In all the sequences antenniform legs due to the ant attacks ob- the female was smaller than the male (Table served in the ®eld. These males were useful 1): in four out of ®ve matings with complete in determining whether the amputation could sperm transfer the male:female size ratio was affect their ability to perform courtship and 1.08 Ϯ 0.05 (mean Ϯ SD) and in the other sperm transfer. Timing and frequency of be- case it was 1.61. The size ratio was 1.04 for haviors were recorded. Behavior patterns were another sequence where the female did not identi®ed following Martin & Bateson (1993), pick up the sperm packages. The duration of Weygoldt (1977) and Weygoldt & Hoffmann the courtship was only 9.10 Ϯ 3.10 min when (1995). In all the specimens prosoma length males were smaller than females (size ratio: was used as an index of body size (Quintero 0.78 Ϯ 0.20; n ϭ 3); in two sequences the 1981). male moved away ®rst. Sexual interactions Every male-female interaction was record- did not occur between very small adults (i.e. ed, with details of the sperm transfer phase. prosoma length less than 4.3 mm), and in Two matings were interrupted in order to ob- 100% of the trials (n ϭ 9) they moved away tain fresh spermatophores still containing from each other after a brief contact with their spermatozoa. All the spermatophores were antenniform legs. The mean duration of the preserved in 70% alcohol. One day before and complete matings was 3.20 Ϯ 0.73 h; range immediately after a mating I observed in a ϭ 2.22±4.46 h. Males were larger than fe- stereomicroscope if the female had sperm un- males (ratio 1.14) in the two shortest matings der the claw-like sclerites by slighty opening while the body size ratio was lower in the two the genital operculum. Approximately eight longest sequences (Table 1). hours after a complete sperm transfer, three Courtship.ÐThis stage represented 92.12 females were sacri®ced and dissected to locate Ϯ 2.27% (mean Ϯ SD) of the duration of an spermatozoa in the spermathecae. The con- average complete mating. In seven out of 10 tents extracted from each spermatheca were courtship sequences males remained motion- PERETTIÐMATING BEHAVIOR IN PHRYNUS GERVAISII 591 up the Ð No Yes Yes Yes Yes Yes 80% courtship Presence: Copulatory Ј Ј Ј Ј Ј Ј Ј 49 Ð Ð Ϯ 3h 7 Sperm 3h 12 4h 22 2h 50 2h 8 3h 03 Latency (h, min) uptaking packages Ј Ј Ј (a) (b) Ј Ј Ј Ј Ј Ј 45 Ϯ 3h 1 phore 4h 13 2h 41 1h 57 Latency (h, min) Figure 1.ÐCourtship in Phrynus gervaisii. The 3h 03 2h 54 3h 45 Spermato- deposition 2h 37 male (left) is tapping the females right pedipalp with his left antenniform leg (arrow). This male was able to mate in spite of lacking a part of his right 1 2 1 1 1 3 1 0.8 1.4 antenniform leg and third leg. Ϯ

less at the base of a vertical piece of tree bark rubbing Pedipalp before the ®rst sexual contact took place. 1.1 44.5 22.4 45 28.1 53.2 32.3 29.1 36.3 (min) Frequency Ϯ Males and females oriented ``face to face'' Latency (Fig. 1) and 4.43 Ϯ 1.05 cm apart. In some cases (3 of 10) the female began the contact, 0 2 1 0 0 3 2 approached the male and stayed within one 1.2 1.1

Ϯ cm of him. As Weygoldt observed, initial ``tapping'' with the antenniform legs was very Female

rubbing light and intermittent. A typical sequence was . Latency is the time from the beginning of the courtship to the onset of the ®rst operculum 9.2 Ð Ð Ð performed repeatedly (5 Ϯ 2.8 times for court- 13.2 34.7 24.5 18.6 22.7 (min) Frequency Ϯ

Latency ship, n ϭ 10): the male stepped forward, per- forming tapping and unfolding his pedipalps. Then, the female also did the same sequence 3 4 2 3 0 5 2 1.5 2.7 (for a more detailed description of each be-

Ϯ havior, contact the author). Phrynus gervaisii The male sometimes tried to grasp the fe- male with her unfolded palps (``grasping with 7 5.7 4.5 8.6 Ð 3.3 8.3 Grasping with 10.3 13.5 unfolded palps

(min) Frequency unfolded palps'', latency: 8.3 Ϯ 3.3 min) (Ta- Ϯ Latency ble 1). In some sequences (6 of 10) the male performed a behavior that has not been de- Ј Ј Ј Ј Ј Ј Ј scribed in either P. gervaisii or other whip spi- 49 Ð Ð

Ϯ ders (``female operculum rubbing'') (Fig. 2). 3h 9 3h 12 2h 53 2h 13 3h 16 4h 28 Mating (h, min) duration This behavior involves lifting the females ca- parace and then touching her, from genital operculum to chelicerae, 2±4 times with his 0.2 Size 1.2 1.04 1.61 1.04 1.41 1.03 1.14 1.14 ratio palps. The occurrences of ``grasping with un- Ϯ folded palps'' and ``female operculum rub- 1.1 7.1 7.5 5.5 7.7 5.6 8.0 7.7 7.6 bing'' per mating was higher in the largest

Ϯ males (Table 1). (mm) length Contrary to Weygoldt's observations, in Prosoma 0.4 8.3 7.8 8.8 8.1 7.9 8.3 8.8 8.7

Male Female this study the female never performed ``shak- Ϯ ing'' movements with her antenniform legs. After 36.37 Ϯ 10.39 min (from the onset of

Table 1. Main features of seven mating sequences in courtship), staying 8±7 cm from the female, Mean Values occurrence of the behavior pattern. Frequency refers to the total number of occurences of the behavior pattern per mating. (a): the female did not pick sperm packages; (b): this sequence was interrupted to obtain a fresh spermatophore. 4 5 6 7 Sequences 1 2 3 the male emited an audible sound by rubbing 592 THE JOURNAL OF ARACHNOLOGY

Figure 2.ÐFlow diagram of a representative courtship sequence in P. gervaisii. Each symbol represents a behavior: circles ϭ female, squares ϭ male. The behaviors that happened more than once are shown in gray. Arrows indicate transitions between behaviors. Abbreviations: BACK ϭ backward motion; CLE ϭ antenniform legs cleaning; GUP ϭ graspping with unfolded palps; MOV ϭ forward motion; PER ϭ pedipalp rubbing; RFO ϭ female operculum rubbing; SHA ϭ antenniform legs shaking; TAP ϭ tapping with antenniform legs; UNF ϭ palps unfolding; VIB ϭ antenniform legs vibration. Other references: (1) beginning of the courtship, the female approachs the male and performs TAP; (2) The rectangle includes behaviors that occur while the male executes intense TAP; (3) in this courtship sequence the female moved backwards after a male approached. Thereafter she moved close to the male again; (4) male turning: this represents the beginning of the sperm transfer stage. the spines of the distal region of his pedipalp his observations might correspond to what I tibiae (``pedipalp rubbing'', its frequency of call ``grasping with the unfolded palps''. occurrence was higher in the two largest Then, staying to 3±4 cm in front of the fe- males) (Table 1). This behavior has not been male, the male stepped forward for a last time cited before, although Weygoldt (1977) had and performed ``tapping'' intensively (no vi- mentioned sounds like ``scrapings''; however brations included). Just as described by Wey- PERETTIÐMATING BEHAVIOR IN PHRYNUS GERVAISII 593 goldt, the male turned until he faced away the sperm masses (Fig. 3). The dorsal append- from the female: this behavior constitutes the ages of the spermatophore sloped a little down beginning of the second stage, the sperm when the female placed her body over them transfer. (Fig. 4). Finally, the female rose and let her Sperm transfer behavior.ÐDuring this body down twice more on the spermatophore stage, the male stayed always below the fe- before moving away. male on the vertical piece of tree bark. As also Copulatory courtship: The male usually observed by Weygoldt, approximately ten continued vibrating his antenniform legs just minutes after turning (10.60 Ϯ 3.30 min, n ϭ as the female took up the sperm packages 7), the male deposited the spermatophore (Figs. 3±4). During this behavior the males while his antenniform legs crossed each other antenniform legs intermittently touched those backwards and vibrated rhythmically (``vibra- of the female. In contrast, she neither moved tion'') while touching the female. In general, her antenniform legs nor touched the male the latency to the ``spermatophore deposition'' during this phase (however, before this phase, was shorter in the largest males (Table 1). At she did it while the male touched her during the end of this behavior, the male did not cross the deposition of the spermatophore). Wey- his antenniform legs; and he executed ``shak- goldt (1977) did not mention in his general ing'' 4 or 6 times together with ``vibration'' description that those male vibrating move- (Fig. 3). Weygoldt (1977) reported that the in- ments occured again when female picked up tense movements of the males antenniform the sperm packages. Copulatory courtship was legs during the deposition of the spermato- executed in ®ve analyzed sequences, and in phore were similar to those in courtship. Ner- all of these the sperm transfer was successful. vertheless in the present study ``shaking'' was The female did not pick up the sperm pack- exclusively performed during the deposition ages in another sequence where no copulatory of the spermatophore and also when the fe- courtship was performed by the male. Al- male approached it. It could be possible that though the copulatory courtship sometimes this difference between the studies re¯ects the could include typical movements of ``tap- difference in the descriptions or perhaps some ping'' and ``shaking'', these behaviors were variability in the sexual behavior because of more common during the precopulatory court- the male's sexual receptivity. After turning un- ship. til facing the female, the male remained over Post-sperm transfer stage.ÐAfter picking the spermatophore for 8.31 Ϯ 1.01 min (n ϭ up the sperm packages, the female raised her 5) (``spermatophore protection''), perhaps body and ended any contact with the sper- providing protection for the spermatophore matophore (Fig. 3). Weygoldt (1977) did not until it hardened. In this study I did not ob- offer details of this stage. Agonistic behavior serve the male grabbing the deposited sper- did not occur either before or after the sepa- matophore as was reported by Weygoldt ration. In two cases the female moved away (1977). ®rst, while the male remained still and then When ``vibration'' and ``shaking'' became he ate the emptied spermatophore almost more vigorous the female moved her anten- completely (the basal part of the stalk re- niform legs forward so that they were touched mained attached to the soil). In contrast, the intermittently by the male's. The combination female never ate the spermatophore. The in- vibration-shaking was performed 10±15 times gestion of the post-insemination spermato- from the beginning of spermatophore protec- phore by the male has not been mentioned be- tion until the end of the female approach to fore for P. gervaisii. In species of Charinus the spermatophore. Although my description Simon 1892 the male performed the same be- of the female approach to the spermatophore havior whereas in another Phynidae, P. mar- is similar to that of Weygoldt (1977), many ginemaculatus C. L. Koch 1841, it was the details can be added. The female performed female that ate the empty spermatophore gentle movements up and down before pick- (Weygoldt 1969, 1977, 1990). A male was ing up the sperm packages. Then, the male able to mate again and deposit another sper- increased the intensity and duration of vibra- matophore six days after mating. The females tion-shaking. Immediately afterwards, she mated again one or two days after mating with moved over the spermatophore and picked up the same or a different male. 594 THE JOURNAL OF ARACHNOLOGY ϭ artner female ϭ male turnning; ϭ 0), showing the time ϭ spermatophore deposition; FAP ϭ individual withdrawal (male: m; female: f). ϭ male tapping with antenniform legs; TUR ϭ preparatory stage for spermatophore deposition; STI-f ϭ , beginning with the male ``turnning'' (TUR) (Time P. gervaisii sperm packages uptaking; TAP-m ϭ spermatophore protection; Pre-DES ϭ STI-f over used spermatophore; SUP ϭ combination of the male behavior patterns ``antenniform legs vibration and shaking''; WIT ϭ SHA ϩ Figure 3.ÐDiagrammatic chronology of the sperm transfer stage in Other references: (1) theis male more was than performing 10 TAP cm before turnning; away (2) from the the intensity transfer of site. the vibrating movements increased. Each ends (3) when either p female approaching to the spermatophore; PSE stillness; STI-ems course of a typical mating sequence. The vertical grayVIB area shows the copulatory courtship stage. Abbreviations: DES PERETTIÐMATING BEHAVIOR IN PHRYNUS GERVAISII 595

Figure 4.ÐSchematic lateral view of the position and movements of the female over a spermatophore during the sperm transfer stage. *: location of the gonopods in the internal face of genital operculum. The male continues performing VIB. Other abbreviations: FAP ϭ female approaching; SUP ϭ sperm package uptaking. Parts of spermatophore: 1 ϭ foot, 2 ϭ basal region of stalk, 3 ϭ medial-distal region of stalk which withstands pressure from the females body, 4 ϭ large region of the horn, 5 ϭ distal region of the horn, 6 ϭ copulatory grooves.

Sexual behavior in injured males.Ð latory groove'' in each of the two sides (Fig. Males of P. gervaisii were able to mate even 6). These grooves contain in their basal-me- if one of their antenniform legs was injured. dial parts the two small sperm packages (Figs. The three males captured in this condition 7±8). The apical part of the spermatophore (prosoma length: 8.04 Ϯ 0.72 mm) were able head has two well developed dorsal horn-like to mate normally with non-injured females appendages (see Figs. 5±6). The wider region (prosoma length: 7.46 Ϯ 0.26 mm). One of of these horns has irregular edges, unlike the those males also lacked the tarsus of the third drawing by Weygoldt (1977) in which they pair of legs (Fig. 1). In addition, one of them appear very uniform. Further, the distal region mated again and deposited a new spermato- of each horn is like a ``®nger'' with its dorsal phore after a week. In an injured male, the part slightly ¯attened, but not oriented toward typical light touches of ``tapping'' became the internal side of the spermatophore as in more similar to ``vibration''. Such a male al- Weygoldt's ®gures. Inside the horns and the ways touched the females antenniform legs distal part of the stalk is a whitish substance with only his non-injured leg. When an anten- that contains abundant spherical cells with niform leg was partially amputated, its healthy many granular bodies. The quantity of the part moved simultaneously with the other leg whitish substance is variable among different during the deposition of the spermatophore, spermatophores. As with other phrynids, in P. vibration and shaking. A female captured in gervaisii each of the females gonopods is similar condition did not approach the sper- equipped with a distal, claw-like, hard and matophore. dark sclerite (Figs. 9±10). At the base of each Pre-insemination spermatophore and fe- is a seminal receptacle (for other details see male genitalia.ÐThe basal region of the stalk Weygoldt, 1990, 1998). In this study it has of the intact, pre-insemination spermatophore been veri®ed that the female could move the is inclined at approximatelly 30Њ in relation to gonopods in all directions, mainly towards the soil (Fig. 4). The two parts of the medial both sides. These movements are possible be- region of the stalk are slightly separated from cause the females are able to selectively in- each other. The medial-distal region forms an ¯ate parts of the soft bases of their gonopods, angle of 50±55Њ with the basal region. The probably by contracting the adductor muscles. distal region is wrinkled and has one ``copu- Sperm transfer mechanism.ÐDuring 596 THE JOURNAL OF ARACHNOLOGY

Figures 5±10.ÐScanning electron micrographs of spermatophore (5±8) and female genitalia (9±10) of P. gervaisii. 5. Lateral view of the spermatophore head. 6. Dorsal view of spermatophore head; showing the well developed dorsal appendages resembling horns. 7. Dorsal-anterior view of the left copulatory groove. 8. View of the right copulatory groove. 9. Internal view of the posterior end of the genital operculum, showing the gonopods with their soft bases and distal claw-like sclerites. 10. Posterior-lateral view of the left gonopod, showing the slight curvature presented in the medial part of claw-like sclerite. Abbreviations: cg ϭ copulatory groove, cs ϭ claw-like sclerite of gonopod, dp ϭ horns of spermatophore, s ϭ place of the sperm package, sa ϭ spine-like apophysis of spermatophore. Scale bars: Figs. 5, 6 ϭ 0.75 mm; 7±9 ϭ 0.25 mm; Fig. 10 ϭ 100 ␮m. sperm transfer, the female moved the genital each copulatory groove of the spermatophore. operculum slightly forwards (Fig. 4) so the Wrinkles appeared in the medial-distal region two ®nger-like regions of the spermatophore's of the stalk as the stalk buckled under the horns entered her genital atrium. Meanwhile pressure of the females body during the cop- the ventral side of the opisthosoma, just be- ulatory movements. When the female's body hind the operculum, rested on the large ex- was completely over the spermatophore, the pansions of the horns. Then the female moved spine-like apophysis of the spermatophore her body up and down slightly and rhythmi- (Fig. 6) touched the ventral side of her opis- cally (Fig. 4), inserting a claw-like sclerite in tosoma. Each claw-like sclerite pulled one PERETTIÐMATING BEHAVIOR IN PHRYNUS GERVAISII 597 sperm package into the apical part of the cop- characteristic differs, at least in the studied ulatory groove. Finally, she lifted her body population, from interspeci®c aggression. and the sperm packages remained under her Only individuals of Phrynus gervaisii inhab- claw-like sclerites in the genital atrium. In one ited the nests of the ant Paraponera clavata case a sperm package, containing both encap- whereas in a near fallen trunk a male of Pa- sulated and dried spermatozoa, remained at- raphrynus laevifrons (Pocock) was captured tached under the claw-like sclerites for one (Peretti pers. obs.). A female of P. gervaisii week. The gonopods of one captured female killed and ate that P. laevifrons male after contained completely dried sperm packages. they were placed together in a terrarium. Ag- Post-insemination spermatophore.ÐDif- onistic behavior among P. gervaisii females ferences between pre-insemination and post- included strong touches with their pedipalps insemination spermatophores are small but and chelicerae. well de®ned. For instance, the wide regions of The various parts of courtship probably the horns of the post-insemination spermato- serve speci®c functions. For example, the ini- phore are partially oriented downwards (i.e., tial stage may be used to assess sexual recep- they are not horizontal in relation to the sur- tivity. During this stage, ``tapping'' and face) because of the pressure exerted by the ``grasping with unfolded palps'' could signal female during her movements over the sper- each individual's inclination to continue the matophore. The ®nger-like region of each courthip. And the male may use additional be- horn is more ¯attened and the stalk is also haviors to elicit female cooperation. For ex- more inclined toward the surface. In addition, ample, ``female operculum rubbing'' may the apical part of the stalk usually became have a stimulatory function since the female more folded on itself when the females move- always became more receptive after the male ments were very intense. performed it. Because this behavior has not been observed in other amblypigid species, DISCUSSION comparison cannot be made for the Order. Sexual behavior.ÐAlthough some of the However, comparing this behavior with those differences reported here with respect to Wey- of other arachnids, such as scorpions, ``female goldt's work are only due to the fact that the operculum rubbing'' resembles ``tickling'' of previous descriptions were more general, a some buthids (Polis & Sissom 1990). Indeed, number of the observations in the current males of Zabius fuscus (Thorell) often per- study are truly novel, such as ``pedipalp rub- form ``tickling'' with the ®rst pair of legs on bing'' and ``female operculum rubbing'' and the females genital operculum. In relation to could indicate the existence of geographical other behaviors of P. gervaissi, further studies variation between the two populations. Fur- are needed to determine whether the audible ther analysis in different populations of other sound emited during ``pedipalp rubbing'' can whip spider species may elucidate the degree be detected by the female (directly or via the of geographical variability in the sexual be- substratum). havior of this group. When the courtships of phrynid Amblypygi The courtship is usually begun by the male are compared, all contain ``vibrations'' or the in arachnids (e.g., spiders: Foelix 1996; scor- equivalent (e.g., with more or less shaking pions: Polis & Sissom 1990), although the fe- movements). This similarity demonstrates the male can perform some mechanical signals to uniformity existing in the courtship of the attract a vagrant male (Polis & Farley 1979). whip spiders (Weygoldt 1990, 1998, etc.). However, like other whip spiders, females of Like in other arachnids (Eberhard 1994; Pe- P. gervaisii began the courtship (e.g., in a se- retti 1997, pers. obs.) the copulatory courtship quence she approached the male and tapped of P. gervaisii could function to stop the fe- to him as if alerting him to her presence). male from leaving before picking up the Probably, the ``female initiative'' is more fa- sperm packages. However, the ultimate func- vored in whip spiders because of the lack of tion of the copulatory courtship in a context strong intersexual aggressiveness which is of cryptic female choice (Eberhard 1991, very common in other groups such as camel 1996) has not yet been studied. spiders and scorpions (Punzo 1998; Peretti & It was surprising that the males that had Acosta 1999; Peretti et al. 1999). This last only one antenniform leg were still able to 598 THE JOURNAL OF ARACHNOLOGY perform pre- and post-copulatory courtship. The female genitalia of P. gervaisii as well Healthy females even cooperated with them as those of other Phrynidae show mechanical (e.g., by moving an antenniform leg close to properties (e.g., genital appendages that can the males broken one). Weygoldt & Hoffmann pick up the sperm packages) which may per- (1995) has photographs (®gs. 5±10) that show mit female control of immediate post-copula- an injured male of Phrnynichus cf. ceylonicus tory processes by exercising some type of (C.L. Koch 1843) (Phrynichidae) mating with ``cryptic female choice'' (Thornhill 1983; a healthy female. The same has been observed Eberhard 1996). For example, the sperm pack- in another phrynichid, Damon gracilis Wey- ages can be transfered immediately from the goldt 1997 (Weygoldt 1998: ®gs. 9±13). It gonopods to the seminal receptacles or they would be very interesting to study experimen- can be stored in place which results in dehy- tally the courtship of these accidentally asym- dration. Although the movements of the go- metrical males. The position that a P. gervaisii nopods are limited, they would be enough to male occupies on the tree bark (always below manipulate and squeeze the sperm packages the female during sperm transfer) agrees with in order to facilitate the expulsion of the se- that observed by Weygoldt & Hoffmann men (Weygoldt 1977). In addition, the move- (1995) in Damon diadema (Simon 1876) (Da- ment of the adductor muscles could induce the monidae). That position could facilitate a bet- movement of the spermatozoa towards the ter penetration of the claw-like sclerites in the seminal receptacles (Weygoldt 1990, 1999). copulatory grooves. In fact, the opened parts Weygoldt (1999) mentioned that clear inter- of the latter stay facing upwards so that the speci®c differences in spermatophores exist claw-like sclerites can pull out the sperm among the Phrynidae. The ``species mate rec- packages without losing them. ognition system'' hypothesis (Paterson 1985) Sperm transfer mechanism and genita- is hardly acceptable to explain spermatophore lia.ÐCopulatory grooves of the spermato- diversity since different types of sexual sig- phore could act as conductors leading the nals (e.g., tapping, shaking, vibrations, etc.) claw-like sclerites of the females gonopore to- are typically produced by both sexes during wards the sperm packages. The two horn-like courtship. Although the sculpting of the sper- dorsal appendages of the spermatophore could matophore creates a large, hard and elastic have a double function: their ®nger-like an- head with a minimum of material (Weygoldt terior region ®ts tightly in the female genital 1999), the need for structural integrity does atrium; meanwhile their wide posterior region not explain the great diversity in shapes since offers an effective base in which the female structural integrity could also be achieved by can rest to coordinate her movements. In ad- a more or less uniform morphology. Might the dition, the distal part of the stalk provides a differently sculpted structures have evolved to suspension area when the female rests on the aid the female in detecting the different parts horns. The whitish substance observed inside of the spermatophore and ®nding the sperm the horns and part of the stalk could help in packages? From a sexual selection perspec- the copulatory process by absorbing the pres- tive, the ``internal courtship'' hypothesis sure exerted by the females body. The spine- clearly predicts a conspicuous interspeci®c di- like apophysis of the spermatophore might versity in male genitalia and a relative unifor- function as a communicatory device since it mity in the females (Eberhard 1985, 1990, contacts the ventral face of the female just as 1996). However, future studies are necessary her gonopods reach the sperm packages. It to support this idea for whip spiders since would be interesting to determine whether the their spermatophores do not seem to have size of the spermatophore including the cop- structures to stimulate females. The other sex- ulatory grooves varies among the males, since ual selection hypothesis, ``the mechanical ®t'' the size of the claw-like sclerites varies among (Eberhard 1985; Huber 1993, 1995), might females from different post-adult stages. In a not be directly applied to amblypygids be- scorpion, Bothriurus bonariensis (C.L. Koch cause of the apparent lack of close mopholo- 1842) the spermatophore tends towards an av- gical correlation between the male and fe- erage size that allows all females -independent males genitalia. An evaluation of this of their body sizes- to use them without dif- hypothesis requires an analysis of all intraspe- ®culty (Peretti et al. 2001). ci®c differences since, at least in the studied PERETTIÐMATING BEHAVIOR IN PHRYNUS GERVAISII 599 population of P. gervaisii, slight differences talia. Bolletino de la Accademia Gioenia de were detected among spermatophores of dif- Scienza Naturale 26:209±214. ferent males (e.g., in the horn-like distal ex- Huber, B.A. 1995. The retrolateral tibia apophysis tensions and in the copulatory grooves). Per- in spiders. Shaped by sexual selection?. Zoolog- haps these differences affect the uptake of the ical Journal of the Linnean Society 113:151±163. Martin, P. & P. Bateson. 1993. Measuring Behavior sperm package by the female just as variations (2nd ed.). Cambridge University Press., Cam- in lamella structures of scorpion spermato- bridge. phores affect sperm transfer (Peretti 1993, Paterson, H.E.H. 1985. The recognition concept of 2000; Peretti et al. 2001). species. Pp. 21±29. In Species And Speciation (E. Vrba, ed.). Transvaal Museum Monographs ACKNOWLEDGMENTS No. 4. I thank William G. Eberhard for giving Peretti, A.V. 1993. Estudio De La BiologõÂa Repro- whole-hearted help and useful suggestions ductiva En Escorpiones Argentinos (Arachnida, during my visit and realization of this study Scorpiones): Un Enfoque EtoloÂgico. Unpub- in Barro Colorado Island. Rolando A. PeÂrez lished Ph. D. Thesis, Universidad Nacional de CoÂrdoba, Argentina. Mendieta helped me ®nd the nests of Para- Peretti, A.V. 1997. Evidencia de cortejo copulatorio ponera clavata containing whip spiders. I en el orden Scorpiones (Arachnida), con un anaÂ- thank Peter Weygoldt and two anonymous ref- lisis en Zabius fuscus (Buthidae). Revista de la erees for useful criticism and comments that Sociedad EntomoloÂgica Argentina 56(1±4):1±10. improved this article. I am also grateful to Peretti, A.V. 2000. Scorpion genitalia: which theory Luis Acosta for his help with the translation can best explain the diversity? Abstract XXI In- of the German. I thank Petra Sierwald, Robert ternational Congress of Entomology, Brazil, Suter, James Berry and Carla Teruzzi for sug- Book II:771. gestions with manuscript preparation and as- Peretti, A.V. & Acosta, L.E. 1999. Comparative sistance in the English. Financial and logistic analysis of mating in scorpions: the post-transfer support for this work was provided by the stage in selected Argentinian bothriurids (Cheli- cerata, Scorpiones, Bothriuridae). Zoologischer Smithsonian Institution, the Smithsonian Anzeiger 237(4):259±265. Tropical Research Institute and the Consejo Peretti, A.V., L.E. Acosta & T.G. Benton. 1999. Nacional de Investigaciones Cientõ®cas y TeÂc- Sexual cannibalism in scorpions: facts or ®ction? nicas de la RepuÂblica Argentina. I express my Biological Journal of the Linnean Society 68: sincere thanks to the authorities of the STRI 485±496 for excellent assistance during my visit in Peretti A.V., M.L. Depiante & M. BattaÂn-Horens- Panama, in particular to Mrs. MarõÂa Leone for tein. 2001. Allometry and asymmetry of body friendly collaboration. characters and spermatophores in Bothriurus bo- nariensis (C. L. Koch) (Scorpiones, Bothriuri- LITERATURE CITED dae). In Scorpions 2001 In Memoriam Gary A. Eberhard, W.G. 1985. Sexual Selection And Animal Polis (V. Fet & P. Selden, eds.). British Arach- Genitalia. Harvard University Press, Cambridge. nological Society, London pp. 345±355. Eberhard, W.G. 1991. Copulatory courtship and PeÂrez Mendieta, R.A. 1996. DistribucioÂn, Mortali- cryptic female choice in insects. Biology Review dad y AsociacioÂn Con Plantas De Nidos De Pa- 66:1±31. raponera clavata (Formicidae) En Una Parcela Eberhard, W.G. 1994. Evidence for widespread De Cincuenta HectaÂreas De Bosque De La Isla courtship during copulation 131 species of in- De Barro Colorado. Unpublished Thesis, Uni- sects and spiders, and implications for cryptic fe- versidad de Panama. male choice. Evolution 48(3):711±733. Polis, G.A. & W.D. Sissom. 1990. Life History. Pp. Eberhard, W.G. 1996. Female Control: Sexual Se- 161±223. In The Biology Of Scorpions. (G. A. lection By Cryptic Female Choice. Princeton Polis, ed.). Stanford University Press, Stanford, University Press, Princeton. California. Eberhard, W.G. 1997. Cryptic female choice insects Punzo, F. 1998. The Biology Of Camel-spiders and spiders. Pp. 32±57. In The Evolution Of (Arachnida, Solifugae). Kluwer Academic Press, Mating Systems Of Insects And Spiders. (J.C. Boston. Choe & B. Crespi, eds.). Cambridge University Quintero, D. 1979. Comportamiento ritualõÂstico: Press, Cambridge. amenaza y sumisioÂn en Paraphrynus laevifrons Foelix, R.F. 1996. Biology Of Spiders (2nd ed.). Ox- (Pocock) (Amblypygi: Phrynidae) de PanamaÂ. ford University Press, New York. Cuaderno de Ciencias No. 3:5±14 Ed. Universi- Huber, B.A. 1993. Female choice and spider geni- taria, PanamaÂ. 600 THE JOURNAL OF ARACHNOLOGY

Quintero, D. 1981. The amblypygid genus Phrynus Weygoldt, P. 1990. Arthropoda -: Sperm in the Americas (Amblypygi, Phrynidae). Jour- transfer. Pp. 77±119. In Reproductive Biology Of nal of Arachnology 9:117±166. Invertebrates. Vol. IV, Part B. Fertilization, de- Thomas, R.H. & D.W. Zeh. 1984. Sperm transfer velopment, and parental care. (K.G. & R.G. Adi- and utilization strategies in Arachnids: ecological yodi, eds.). Oxford & IBH Publishing Company, and morphological constraints. Pp. 179±221. In New Dehli, Bombay, Calcutta. Sperm Competition And The Evolution Of Ani- Weygoldt, P. 1998. Mating and spermatophore mor- mal Mating Systems. (R.L. Smith, ed.). Academ- phology in whip spiders (Phynichodamon scullyi ic Press, Orlando. (Purcell, 1901), Damon gracilis nov. spec., Da- Thornhill, R. 1983. Cryptic female choice and its mon variegatus (Perty, 1834), and Euphrynichus implications in the scorpion¯y Harpobittacus ni- bacillifer (Gerstaecker, 1873)) (Arachnida: griceps. The American Naturalist 122:765±788. Amblypygi: Phrynichidae). Zoologischer Anzei- Weygoldt, P. 1969. Beobachtungen zur Fortp¯an- ger 236:259±276. zungsbiologie und zum Verhalten der Weygoldt, P. 1999. Spermatophores and the evolu- Geiûelspinne Tarantula marginemaculata C.L. tion of female genitalia in whip spiders (Cheli- Koch (Chelicerata, Amblypygi). Zeitshrift fuÈr cerata, Amblypygi). Journal of Arachnology 27: Morphologie der Tiere 64:338±360. 103±116. Weygoldt, P. 1974. Vergleichende Untersuchungen Weygoldt, P. & P. Hoffmann. 1995. Reproductive an zwei Heterophrynus (Admetus) Arten, H. lon- behavior, spermatophores, and female genitalia gicornis Butler und H. batesii Butler (Arachnida, in the whip spiders Damon diadema (Simon, Amblypygi, Tarantulidae). Zoologischer Anzei- 1876), Phrynichus cf. ceylonicus (C.L. Koch, ger 192:175±191. 1843) and Euphynichus alluaudi (Simon, 1936) Weygoldt, P. 1977. Kampf, Paarungsverhalten, (Chelicerata: Amblypygi). Zoologischer Anzei- Spermatophoren-morphologie und weibliche Ge- ger 234:1±18. nitalien bei neotropischen Geiûelspinnen (Amblypygi, Arachnida). Zoomorphologie 86: Manuscript received 20 March 2001, revised 20 271±286. November 2001.