ANIMAL BEHAVIOUR, 2006, 72, 487e502 doi:10.1016/j.anbehav.2006.02.008 Suppression of escape behaviour during mating in the cricket Acheta domesticus K . A . KI LLIA N, L. C. S NELL, R. A MM ARELL & T. O. CRI ST Department of Zoology, Miami University (Received 3 June 2005; initial acceptance 17 August 2005; final acceptance 27 February 2006; published online 7 July 2006; MS. number: A10182) We examined behavioural switching in the cricket Acheta domesticus. Animals are constantly exposed to sensory information that must be integrated by the nervous system and transformed into an appropriate behavioural response. Often, the same sensory inputs can play a crucial role in different behaviours. For example, in isolated crickets, tactile activation of specific cercal sensory receptors can trigger escape, but these same sensory inputs are also important during mating. We mechanically stimulated crickets before, during and after copulation and found that most touch-evoked escape responses are suppressed in copulating males. The behavioural switch from escape to mating occurs following a male’s chemosensory contact with a female and requires the continued presence of the female for its full expression. We removed the antennae from male and female crickets to examine whether chemosensory cues detected by the antennae are necessary for this escape suppression and mating initiation. Although the antennae are the primary source of this chemosensory information, we determined that the maxillary palps are another important source. Removal of male antennae did not significantly impact mating success. The loss of the female antennae, however, did have a significant negative effect on both female and male receptivity and mating behaviour. Ó 2006 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. Animals are exposed to a constant flow of sensory information from their internal and external environments that must be integrated by the central nervous system and transformed into an appropriate behavioural response. Often, sensory inputs and central neurons can be shared components of the neural circuits responsible for the generation of different behaviours (Morton & Chiel 1994; Kristan & Shaw 1997). Accordingly, an animal must be able to choose, based on the circumstances in which it finds itself, how to respond to such shared inputs. Although the neural systems involved in the production of specific behavioural responses have been well characterized in some cases, the underlying mechanisms contributing to this decision-making process are still not well understood. A hierarchical scheme has been used to explain the process of behavioural choice where higher-ranking behavioural functions have precedence over, and are able to Correspondence and present address: K. A. Killian, Department of Zoology, Miami University, 212 Pearson Hall, Oxford, OH 45056, U.S.A. (email: [email protected]). L. C. Snell is now at the Biology Department, McLennan Community College, 1400 College Drive, Waco, TX 76708, U.S.A. 0003e3472/06/$30.00/0 control the production of, lower-ranking behaviours (Tinbergen 1950, 1951; Davis 1979). Such a hierarchical arrangement can help to explain how an animal prioritizes its behavioural responses in the face of conflicting demands. The relatively simpler nervous systems and stereotyped behaviours of invertebrates have provided researchers with an opportunity to directly investigate the underlying mechanisms controlling this prioritizing of behaviour. For example, in the marine molluscs Pleurobranchaea (Davis et al. 1974a) and Clione (Norekian & Satterlie 1996), the leech Hirudo (Misell et al. 1998) and the crayfish Procambarus (Krasne & Lee 1988), feeding behaviour is considered to rank more highly than withdrawal in each animal’s behavioural repertoire since all of these animals show a decrease in their response to mechanical stimulation while feeding. Mutual inhibitory interactions among the neural circuits responsible for the production of conflicting behaviours influence this behavioural choice (Kovac & Davis 1980; Huang & Satterlie 1990; Jing & Gillette 1995, 2000; Norekian & Satterlie 1996; Esch & Kristan 2002). However, the position of a particular behaviour within an animal’s hierarchy is also flexible and dependent on an animal’s motivational state. For example, feeding behaviour no longer has priority over withdrawal or escape in a well-fed Pleurobranchaea (Davis 487 Ó 2006 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. 488 ANIMAL BEHAVIOUR, 72, 2 et al. 1977; Gillette et al. 2000), and a hormonally induced repression of feeding behaviour has been shown to occur in Pleurobranchaea during egg laying (Davis et al. 1974b). We have been using a simple insect model, the cercal sensory system of the cricket Acheta domesticus, to investigate behavioural choice and neural decision making. The cerci, two cone-shaped sensory appendages located on the terminal abdominal segment of orthopteran insects, are covered by mechanosensory receptors that can trigger escape behaviours such as running (Gras & Horner 1992), kicking and jumping (Huber 1965; Dumpert & Gnatzy 1977). For example, when a cricket’s cerci are contacted by a predatory wasp, touch-sensitive campaniform sensilla are activated (Dumpert & Gnatzy 1977), resulting in defensive kicking of the metathoracic legs (Gnatzy & Heusslein 1986; Hustert & Gnatzy 1995). However, these same cercal campaniform sensilla also stimulate cricket mating behaviour (Snell & Killian 2000). Our goal was to determine the underlying mechanism that allows these same cercal sensory inputs to influence two such widely conflicting behaviours as mating and escape. Cricket mating behaviour consists of an orderly sequence of events: male stridulation and pair formation; courtship behaviour; copulation and transfer of the spermatophore; and male guarding behaviour (Alexander 1961; Loher & Rence 1978; Loher & Dambach 1989; Adamo & Hoy 1994). In this study, we focused on behavioural changes occurring during courtship and copulation. Chemosensory antennal contact between a male and female cricket is usually sufficient to elicit courtship behaviour (Rence & Loher 1977; Hardy & Shaw 1983; Balakrishnan & Pollack 1997). During this initial stage of mating, the male produces courtship song, flattens his body to the ground, and rocks rhythmically from side to side in an attempt to stimulate female mounting. During mounting, the female comes up behind the male and contacts his cerci with her maxillary palps. Instead of eliciting defensive kicking or jumping, however, this contact triggers a ‘backward slipping’ motor response from the male (Sakai & Ootsubo 1988). During backward slipping, the male holds his body lowered towards the ground and moves backward beneath the female as she mounts onto his back. As she mounts, the male hooks his epiphallus onto the female’s subgenital plate to transfer the spermatophore. It takes approximately 2.5 min for spermatophore transfer to be completed and during this time the male holds his cerci adjacent to the female’s abdomen where they produce small, repetitive flicking movements. Tactile activation of campaniform sensilla on the male’s cerci during these movements provides important feedback on female body position to the male’s motor circuitry responsible for transfer, and the removal of these sensory inputs can significantly decrease mating success (Snell & Killian 2000). In this study, we investigated the role of behavioural context in an animal’s responsiveness to touch. First, we hypothesized that mating would rank highly in the cricket’s behavioural repertoire such that escape responses to tactile stimulation would be suppressed during mating. Huber (1965) reported that male Gryllus campestris crickets show fewer kicking responses to tactile stimulation of the cerci when courting a female. Here, we asked whether there is a more widespread decrease in a mating male’s touch-evoked escape responses. For example, tactile stimulation of the wings normally triggers jumping (Hiraguchi & Yamaguchi 2000), however, this response must be suppressed during copulation in order for the female to assume her mounted position on the back of the male. To test this hypothesis, we mechanically stimulated male crickets before and during copulation on five different body locations. Second, we tested the hypothesis that the initiation of courtship song and behaviour by a male would signify that a behavioural switch had occurred and would be accompanied by a suppression of the male’s touch-evoked escape responses. We also examined the role of the female in the initiation of this switch. Finally, we removed the antennae from male or female crickets to determine whether chemosensory cues detected by the antennae are needed to trigger a cricket’s switch from escape to mating behaviour. METHODS Animals We used male and female Acheta domesticus crickets in all experiments. Immature crickets were raised in our laboratory culture or purchased from Fluker’s Cricket Farm, Baton Rouge, Louisiana, U.S.A. Groups of approximately 50 immature males and females were housed together in large clear plastic containers within an incubator with controlled temperature (29 C) and light (12:12 h light:dark cycle) exposure. Upon reaching adulthood (10th instar), males and females with all body parts intact were placed in individual plastic deli containers (8 cm high, 11 cm wide) and 10e21 days elapsed before they took part in a mating trial. Isolated crickets were housed in incubators with other similarly stored crickets so that visual and auditory, but not tactile, interactions were possible. All crickets had free access to water and dry dog chow. Behavioural Tests For each mating trial, we randomly chose a pair of isolated, virgin, age- and size-matched male and female crickets. Each cricket was only used in one trial. Males were only used if a spermatophore was present and calling song was produced. We carefully placed (without direct handling) each mating pair into a clear, round Plexiglas arena (15 cm diameter). A removable piece of opaque Plexiglas in the centre of the arena separated the male from the female. White paper on the floor of the arena provided traction and was replaced after each trial. All trials were performed at room temperature (20 C) under dim lights during the last 6 h of the light cycle (1200e1800 hours), when A. domesticus shows maximum sexual activity (Nowosielski & Patton 1963). We allowed each mating pair to acclimate to the arena for 15 min before removing the divider. Each trial was given a 15-min time limit, with removal of the divider signalling the start of the trial. All control and experimental mating pairs made physical contact during all trials. KILLIAN ET AL.: BEHAVIOURAL SWITCHING IN CRICKETS Successful pairs were pairs in which a spermatophore was completely transferred from male to female within the allotted 15 min. Mating failure could occur at any one of the four sequential steps within a trial, which we defined according to the male’s role in the mating sequence as follows. (1) Failure in courtship song (CS): the male failed to produce continuous courtship song and to show the lowered body posture and rhythmic rocking associated with courtship. (2) Failure in backward slipping (BWS): the female did not successfully mount onto the back of the male. All males produced courtship song and rhythmic rocking. Included in these failures were pairs in which the male failed to successfully back under the female as she tried to mount him, as well as pairs in which the female did not show interest, and thus never attempted to mount the male. (3) Failure in hooking: the male was able to successfully backward slip under the female as she mounted, but was unable to attach his epiphallus to the female’s subgenital plate. (4) Failure in transfer: the male was unable to successfully thread and transfer the spermatophore to the mounted female following successful hooking. Such failures were confirmed when the female dismounted without an attached spermatophore. Mechanical stimulation We mechanically stimulated male crickets with a small (size zero), soft bristled paintbrush on five different body locations: the distal antenna; the dorsal femur of the mesothoracic leg; the dorsal femur of the metathoracic leg; the dorsal surface of the two forewings at the point of hinge overlap; or the dorsal surface of the cercus. To prevent response habituation, paintbrush stimuli were applied to each cricket with an interstimulus interval of at least 10 s and in the following sequence: cercus, wings, metathoracic leg, antenna, mesothoracic leg. The application of stimuli alternated between the left and right side of the body such that all appendages on one side of the body were stimulated before those on the opposite side. This procedure ensured that at least 100 s would elapse before the reapplication of paintbrush stimulation to each cercus, antenna or leg, while at least 40 s elapsed between wing stimulations. The side to which stimuli were first applied varied between animals and was selected randomly. We visually observed and recorded the behavioural response of each animal to each mechanical stimulus. One of six possible reactions could occur following a mechanical stimulus: (1) forward locomotion; (2) backward locomotion; (3) jump (full extension of both metathoracic legs); (4) withdrawal; (5) kick (partial to full extension of one metathoracic leg); or (6) no response. We included as withdrawal responses a lateral movement of the body or appendage away from the stimulus source following mesothoracic or metathoracic leg stimulation, a lateral withdrawal of the animal’s body away from a touched cercus, or a lateral turning movement of the animal’s head and/or body away from a stimulated antennae. During copulation, some animals responded to the paintbrush with a small jerking, or shivering, movement of the body or stimulated appendage. We called these slight movements ‘flinches’ and included them with withdrawal responses. Occasionally, an animal would move forward or backward a few steps before eliciting a kick or a jump. In those cases, we considered the kick or jump to be the primary response and only included that response in our tabulations. Under some experimental situations, a male could respond with mating movements, such as backward slipping or abdominal hooking movements, during tactile stimulation. Control mating pairs Male and female crickets used in control mating trials did not receive mechanical stimulation or any other experimental manipulation. We used these mating pairs (N ¼ 40) to determine an overall mating success rate, a control level of failure for each stage of the mating sequence (Fig. 1), and a control duration for each stage of the mating sequence (Table 3). Each trial was recorded on videotape with a Videolab Flexcam camera and a Panasonic AG-1960 videocassette recorder. Experiment 1: suppression of male escape behaviours We hypothesized that all male touch-evoked escape behaviours such as jumping, running and kicking would be suppressed during copulation. To test this hypothesis, we applied mechanical stimuli to male crickets at three different time periods: before, during and immediately after copulation. We used ‘premating’ stimuli to characterize and quantify the range of behavioural responses shown by males before they came into contact with a female and commenced mating behaviour, and to determine the primary response evoked during mechanical stimulation of each of the five body regions. We used ‘during-mating’ stimuli to determine whether the occurrence of each of these characteristic responses was altered during copulation. During-mating stimuli were applied to each male during the period of spermatophore transfer (i.e. female mounted on top of the male and male epiphallus attached to the female). ‘Postmating’ stimuli, stimuli applied to each male within 2e10 min of successful spermatophore transfer and female dismount, were used to determine whether there was a return of each behavioural response to its premating level following a successful copulation. Two groups of males were tested that differed in the time the males received premating stimulation. In one group (N ¼ 42), premating stimuli were applied 15e60 min before initiation of a mating trial. Individual males were placed into an arena and allowed to acclimate for 15 min before the onset of premating stimulation. Each of the five selected body regions was stimulated five times for a total of 25 premating stimuli delivered to each male. Immediately following male stimulation, we placed the opaque barrier in the centre of the arena and added a female cricket to the side opposite the male. Following a 15e60-min period of acclimation, this barrier was removed. In the second group (N ¼ 40), we returned each male to its container after it received premating stimulation and waited 24e48 h before allowing the male to take part in a mating trial. 489 ANIMAL BEHAVIOUR, 72, 2 (a) 1 0.8 0.6 0.4 Proportion of pairs that successfully reached each mating stage 490 0.2 Control (40) 15–60 min (42) 24–48 h (40) 0 (b) 1 0.9 0.8 0.7 Control (36) CS + Fem (54) CS – Fem (42) 0.6 (c) 1 0.8 0.6 0.4 0.2 0 Contact Control (40) MAA (49) FAA (53) CS BWS Hooking Transfer Mating stage Figure 1. Effect of (a) the time at which males received premating tactile stimulation, (b) tactile stimulation of the male in the presence (CS þ Fem) or absence (CS Fem) of the female and (c) male (MAA) or female (FAA) antennae ablation on the proportion of mating pairs that successfully reached and completed each stage in the mating sequence. Mating stages are shown in order of occurrence and include: physical (i.e. chemosensory) contact; courtship song (CS) with lowered posture and rhythmic rocking of the body; backward slipping (BWS) with female mounting; hooking of the male epiphallus onto the female subgenital plate; transfer of the spermatophore. All control and experimental mating pairs made physical contact during all trials. Numbers in parentheses indicate number of individual mating trials within each group. The same group of control mating pairs was used in each comparison. Note that there are fewer control pairs in (b) because only males that initiated courtship behaviour (CS) were used in this comparison. Also note the different scale in (b). Means 1 binomial SE are shown. Experiment 2: activation of the behavioural switch from escape to mating Chemosensory contact between male and female crickets elicits male courtship song and courtship behaviour (Rence & Loher 1977; Hardy & Shaw 1983; Balakrishnan & Pollack 1997). We hypothesized that this contact is necessary and sufficient to trigger the behavioural switch from escape to mating, and that the initiation of courtship song and rhythmic rocking by a male would signify that a switch had occurred in that animal. To investigate the time to suppression of male escape behaviour and the role of the female in this suppression, we recorded the behavioural responses of three groups of male crickets to mechanical stimulation. (1) Courtship song with female present (CS þ Fem). We used 54 mating pairs. The male of each pair was mechanically stimulated with a paintbrush immediately after courtship song was elicited by male and female chemosensory contact. The female was allowed to stay in close proximity and have additional contact with the male throughout stimulus application. Because our tactile stimulation of the male did not inhibit male or female mating behaviour, a variable number of stimuli were delivered (range 2e25) before the onset of male backward slipping and female mounting, at which point we ended our stimulation of the male. (2) Courtship song without female present (CS Fem). We used 42 pairs of crickets. Immediately after male courtship song was elicited by chemosensory contact with the female, we prevented further contact between each pair by separating the male and female with the opaque divider. We then stimulated each male with a paintbrush until a total of 25 stimuli were applied (five stimuli to each of five body regions). Immediately upon completion of stimulation, we lifted the divider and allowed the trial to resume. (3) No courtship song with female present (No CS þ Fem). The males in these 10 mating pairs made chemosensory contact with a female, but failed to begin courtship song and rhythmic rocking within the 15-min time limit allotted for each trial. We began tactile stimulation of each male at the end of the 15-min trial, and this stimulation was done in the presence of the female. We delivered five stimuli to each of the five body regions for a total of 25 stimuli delivered to each male. For the males of all three groups, each behavioural response to a stimulus was placed into one of three categories: escape response (forward locomotion, backward locomotion, jumping, kicking, or withdrawal); mating response (backward slipping or abdominal hooking movements); or no response. For male mating responses, hooking consisted of upward thrusting movements of the abdomen, and backward slipping was a backward movement that occurred in conjunction with a lowered body posture. Experiment 3: role of the antennae Male and female antennal contact triggers courtship song and intense posture in male crickets. To determine whether chemosensory cues detected by the antennae are necessary for successful mating, we removed the antennae from male and female crickets 1e2 days after the adult moult. Each animal was anaesthetized on ice for 10e15 min and microdissecting scissors used to sever each antenna at its base. Each animal was then returned to its container for an additional 9e20 days before taking part in a mating trial. KILLIAN ET AL.: BEHAVIOURAL SWITCHING IN CRICKETS For these mating trials, we either paired antennalablated males (MAA) with control females (N ¼ 49), or paired antennal-ablated females (FAA) with control males (N ¼ 53). We did not apply paintbrush stimuli. We recorded each trial on videotape and measured the duration of each mating stage and the number of physical contacts, and the sites of those contacts, for each pair by analysing slow motion frame-by-frame playback of each trial. Data Analyses Probability of success or failure across the mating sequence The probability that cricket pairs successfully completed each step of the mating sequence was analysed as a correlated binomial response variable. Each pair was considered a subject, and the probability of successful completion of each mating step (1 ¼ success, 0 ¼ failure) within subjects was analysed as a repeated measures response. The response profiles of subjects were compared among treatments and controls in three separate analyses: (1) control and premating stimulus treatments (15e60 min and 24e48 h); (2) control and the presence or absence of the female on the responses of males that had initiated courtship song (CS Fem and CS þ Fem); and (3) control and antennal-ablation treatments (FAA and MAA). Correlated binomial responses were analysed using generalized estimating equations (Myers et al. 2002). Maximum likelihood estimation was first used to provide initial parameter estimates for the effects of treatment, which assumes homogeneous variances. A revised empirical estimate of the standard errors was conducted using a correlation matrix of the repeated observations. This approach, called quasilikelihood estimation because it involves an iterative procedure, results in larger estimates of standard errors especially if repeated observations are correlated (Myers et al. 2002). We used this approach to analyse the three experiments using Proc GENMOD in SAS 9.1 software (SAS Institute 2002, Cary, North Carolina, U.S.A.). The binomial variable of success or failure at each mating step was modelled using the logit link function. The control was set as the reference group for parameter estimation. We used an autoregressive correlation matrix in SAS to account for the conditional dependence in the probability of success between adjacent mating steps. To test for treatment differences in the probability of success, we performed pairwise contrasts between treatments and the control using Wald’s chisquare test. Comparison of behavioural responses before, during and after mating To determine whether the touch-evoked behavioural responses produced by tactile stimulation of each of five body regions were altered during and after mating, we used one-way repeated measures analyses of variance (ANOVA) with BonferronieDunn post hoc tests to compare the premating behavioural responses of male crickets to their responses during and after mating (experiment 1). Since we predicted that escape responses would decrease and ‘no responses’ would increase during mating, we tested the hypothesis that the occurrence of the six possible behavioural responses would differ between the two treatments (i.e. premating versus during-mating stimulation or premating versus postmating stimulation). In experiment 2, each behavioural response to a tactile stimulus was placed into one of three categories: escape response (forward locomotion, backward locomotion, jumping, kicking, or withdrawal); mating response (backward slipping or hooking); or no response. A one-way ANOVA with individual animals as subjects was performed to determine whether the presence of courtship song was necessary for the male’s switch from escape to mating responses (comparison of CS þ Fem pairs to No CS þ Fem pairs) and to determine whether the continued presence of the female during a mating trial was necessary for male escape suppression (comparison of CS þ Fem pairs to CS Fem pairs). Data were subjected to arcsine square-root transformation prior to analysis to stabilize the variation in proportional responses across the entire response range of each experiment (Neter et al. 1996). ANOVA was performed using PROC GLM in SAS 9.1 (SAS Institute). We used ANOVAs so that we could test the hypotheses of mating or treatment (courtship song or female presence) effects as the mean proportional response using individual animals as replicates. All values are reported as means 1 SE unless otherwise indicated. RESULTS Experiment 1: Suppression of Male Escape Behaviours During Mating Males receiving tactile stimulation 24e48 h premating The mating behaviour of male crickets that received tactile stimulation 24e48 h before a mating trial was unaffected by this prior stimulation (Fig. 1a), with the proportion of mating pairs successfully reaching each stage of the mating sequence and the overall mating success similar to those of control pairs (Wald’s chi-square: c21 ¼ 0:23, P ¼ 0.63). Of the 40 males that were stimulated 24e48 h before their mating trials, 29 (73%) were able to successfully complete all stages of mating, whereas 31 of 40 control males (78%) mated successfully. In addition, the proportion of experimental males that made the transition from hooking to successful completion of transfer was not different from the unstimulated control males, indicating that paintbrush stimulation of the experimental males during this stage had no effect on their ability to transfer the spermatophore. Behavioural responses elicited by tactile stimulation of the antennae, mesothoracic legs, metathoracic legs, wings and cerci before, during and after copulation for 28 of these successful male crickets are shown in Fig. 2. Premating stimulation consisted of four to five stimuli delivered to each of five body locations for a total of 694 stimuli delivered to these 28 males. Premating stimulation of each body part produced a primary, or dominant, behavioural response (Fig. 2). Most premating stimuli applied to the antennae produced either no discernible evasive response 491 ANIMAL BEHAVIOUR, 72, 2 1 Antennae 0.75 *** Pre During Post ** 0.5 ** 0.25 ** 0 1 0.75 Meso Legs * *** 0.5 *** 0.25 Proportion of behavioural occurrences 492 0 * ** *** 1 0.75 *** Meta Legs * 0.5 0.25 0 * *** * *** ** ** 1 Wings 0.75 ** 0.5 0.25 0 1 ND ND ND ND ND Cerci ND *** 0.75 0.5 0.25 0 *** Fwd Loc Bwd Loc ** ** Jump Withdraw *** Kick No Resp Figure 2. Comparison of the touch-evoked behavioural responses of successfully mated males (N ¼ 28) that had received premating stimulation 24e48 h before their mating trials (Pre) to their responses during spermatophore transfer (During) and within 10 min after female dismount with a spermatophore (Post). The proportion (mean 1 SE) of touches resulting in forward locomotion (Fwd Loc), backward locomotion (Bwd Loc), jumping, withdrawal, kicking, or no response (No Resp) is shown for each of five different body locations. Mating responses, such as backward slipping or abdominal hooking movements, were never observed in these males. ND ¼ not determined: wings could not be touched because of the mounted position of the female. *P < 0.05; **P < 0.01; ***P < 0.001. from the animal, or a withdrawal of the animal’s head or body away from the stimulus source. Withdrawal of the leg or body occurred most often during premating stimulation of the mesothoracic legs, whereas the primary responses to metathoracic leg stimulation were forward locomotion and jumping. Tactile stimulation of the wings primarily evoked jumping responses whereas a tactile stimulus applied to a cercus usually resulted in the male moving forward away from the stimulus or kicking the paintbrush with the ipsilateral metathoracic leg (Fig. 2). KILLIAN ET AL.: BEHAVIOURAL SWITCHING IN CRICKETS Mating responses, such as backward slipping or abdominal hooking movements, were never observed during premating stimulation. During copulation, there was a significant decrease in most touch-evoked escape responses, with the majority of paintbrush stimuli producing no visible response from these animals (Fig. 2, see Table 1 for test statistics). Since spermatophore threading and transfer is completed by the male in approximately 2.5 min (Snell & Killian 2000), we were only able to successfully deliver 12.5 0.5 stimuli (range 8e17 stimuli) to each animal during this stage of mating, for a total of 352 stimuli delivered to these 28 males. The wings could not be stimulated because of their position beneath the mounted female. During copulation, 26 of the 154 stimuli (16.9%) applied to the mesothoracic and metathoracic legs of these males resulted in small flinching movements, which we Table 1. Analysis of the proportion of total touch-evoked responses of 28 successful male crickets stimulated 24e28 h prior to a mating trial to their responses during and after copulation (ANOVA) During mating F df Antennae Fwd Loc Bwd Loc Jump Withdraw Kick No Resp 1.00 2.17 d 14.55 d 22.28 1,13 1,13 1,13 1,13 1,13 1,13 Meso Legs Fwd Loc Bwd Loc Jump Withdraw Kick No Resp 10.20 5.95 33.71 15.86 d 77.89 P Postmating F df P 0.34 0.16 d 0.002 d 0.0004 3.24 0.38 0.00 7.89 d 7.96 1,27 1,27 1,27 1,27 1,27 1,27 0.08 0.54 1.00 0.009 d 0.009 1,25 0.004 1,25 0.02 1,25 <0.0001 1,25 0.0005 1,25 d 1,25 <0.0001 1.33 0.11 1.12 6.78 d 1.00 1,27 1,27 1,27 1,27 1,27 1,27 0.26 0.75 0.30 0.01 d 0.33 Meta Legs Fwd Loc 158.68 1,27 <0.0001 6.84 1,27 0.01 Bwd Loc 7.36 1,27 0.01 7.16 1,27 0.01 Jump 28.81 1,27 <0.0001 12.24 1,27 0.002 Withdraw 0.25 1,27 0.62 0.40 1,27 0.53 Kick 12.31 1,27 0.002 1.51 1,27 0.23 No Resp 239.82 1,27 <0.0001 0.00 1,27 0.98 Wings Fwd Loc Bwd Loc Jump Withdraw Kick No Resp ND ND ND ND ND ND Cerci Fwd Loc 67.34 Bwd Loc 1.00 Jump 11.78 Withdraw 0.19 Kick 70.71 No Resp 322.76 ND ND ND ND ND ND ND ND ND ND ND ND 1,27 <0.0001 1,27 0.33 1,27 0.002 1,27 0.66 1,27 <0.0001 1,27 <0.0001 0.99 2.25 9.70 3.37 1.00 0.33 1,27 1,27 1,27 1,27 1,27 1,27 0.33 0.15 0.004 0.08 0.33 0.57 1.61 0.00 8.42 0.04 0.05 0.01 1,27 1,27 1,27 1,27 1,27 1,27 0.22 1.00 0.007 0.84 0.82 0.90 Fwd Loc: forward locomotion; Bwd Loc: backward locomotion; No Resp: no response. See text for definition of responses; Dash: did not occur; ND: not determined. Not all males received the full complement of touches during copulation. tabulated as withdrawal responses. These flinches, which accounted for 100% of withdrawal responses observed in response to leg stimulation during spermatophore transfer, were not observed during premating or postmating stimulation of the male. Postmating stimulation of these 28 males (N ¼ 698 total stimuli) revealed that the proportional occurrence of most responses returned to premating levels following female dismount (Fig. 2, Table 1). However, a decrease in touch-evoked jumping responses relative to premating levels was observed for all five body locations, with this decrease significant for metathoracic leg, wing and cercal stimulation (Fig. 2, Table 1). As during premating stimulation, all stimuli applied during and after copulation failed to evoke mating responses in these males. Males receiving paintbrush stimulation 15e60 min premating Significantly fewer males that received paintbrush stimulation 15e60 min before a mating trial were able to complete all stages of mating and successfully transfer a spermatophore to a female (Fig. 1a) when compared to both unstimulated control males (Wald’s chi-square: c21 ¼ 5:67, P ¼ 0.02) and to males stimulated 24e48 h before their mating trials (Wald’s chi-square: c21 ¼ 3:94, P ¼ 0.047). Most failures occurred at the first two stages of mating (Fig. 1a), with nine of the 42 males failing to initiate courtship behaviour even after multiple contacts with a female, and an additional nine pairs failing during backward slipping and female mounting. We used paintbrush stimuli to determine whether the successful 15e60-min prestimulated males would show escape suppression during copulation. We delivered a total of 550 premating stimuli (25 stimuli per animal) to the 22 successful males in this group. Premating stimuli applied to each body region evoked one or two primary responses and the proportional occurrence of each response was similar to that observed for males receiving 24e48 h of premating stimulation (data not shown). We delivered 281 stimuli (12.8 0.9, range 4e20 stimuli) to these 22 males during copulation and found a significant decrease in all of the primary touch-evoked escape responses (data not shown). Instead, most stimuli evoked no discernible response (X SE proportion of no responses: antennae, 1.0 0.0; mesothoracic legs, 0.85 0.07; metathoracic legs, 0.92 0.03; wings, not determined; cerci, 0.99 0.06). Flinches were the only response we observed when these males were stimulated during copulation. As with 24e48-h prestimulated males, paintbrush stimulation had no effect on the ability of the 15e60-min males to successfully complete transfer (Fig. 1a). We also stimulated these males after female dismount (25 per animal for a total of 550 stimuli). Again, as with males stimulated 24e48 h before mating, the proportional occurrence of most responses returned to premating levels once transfer was complete (data not shown). However, these males also showed a decrease, relative to premating levels, in the proportion of jumping responses evoked by stimulation of all five body regions, with this decrease significant for mesothoracic leg (ANOVA: F1,21 ¼ 5.17, P ¼ 0.03) and wing (ANOVA: F1,21 ¼ 6.30, P ¼ 0.02) stimulation, and not 493 494 ANIMAL BEHAVIOUR, 72, 2 quite significant for cercal stimulation (ANOVA: F1,21 ¼ 3.08, P ¼ 0.09). All stimuli delivered either before, during, or after copulation also failed to evoke mating movements in these males. Experiment 2: Activation of the Behavioural Switch from Escape to Mating Since we found that most touch-evoked responses were inhibited in copulating males, we asked whether courting males would also show this suppression. We hypothesized that male escape suppression would occur in conjunction with the initiation of male courtship song and behaviour. We used three groups of mating pairs to test this idea. In the CS þ Fem group, male crickets were tactilely stimulated immediately after the initiation of courtship song and in the presence of a female. All stimuli were applied to these males prior to mounting of the female. In the CS Fem group, the male was separated from the female by an opaque barrier immediately after courtship song initiation and then stimulated. In the No CS þ Fem group, we recorded the touch-evoked behavioural responses of males that failed to produce courtship song even after multiple contacts with the female and so never progressed into the mating sequence. These males were stimulated in the presence of the female. Application of paintbrush stimuli to courting males did not affect the subsequent ability of these males to copulate. The proportion of CS þ Fem pairs and control pairs that successfully transferred a spermatophore (Fig. 1b) was not significantly different (Wald’s chi-square: c21 ¼ 1:07, P ¼ 0.30). Similarly, the mating success of pairs in which males were stimulated in the absence of the female but immediately after the initiation of courtship (CS Fem; Fig. 1b) was not significantly different from that of control pairs (Wald’s chi-square: c21 ¼ 0:87, P ¼ 0.35) or CS þ Fem pairs (Wald’s chi-square: c21 ¼ 0:00, P ¼ 0.97). We compared the proportion of escape responses, mating responses, or no responses elicited by tactile stimulation of the males for the successful CS þ Fem, successful CS Fem, and the No CS þ Fem mating pairs (Fig. 3, Table 2). Since we did not act to prevent male backward slipping and female mounting during our application of paintbrush stimuli to the CS þ Fem males, a variable number of stimuli were applied to each male (8.71 0.93 stimuli/animal, range 2e25 stimuli) and some males did not receive tactile stimulation to all five body locations. In contrast, because the males of the CS Fem pairs were separated from the females during tactile stimulation, we were able to apply 25 stimuli, five to each of the five body parts, to each male before we removed the barrier and allowed each pair to continue mating. The males that failed to initiate courtship song were stimulated in the presence of the female (No CS þ Fem), however, because these pairs never initiated mating behaviour, we were also able to apply 25 stimuli to each of these 10 males. Courting males stimulated in the presence of the female showed significantly fewer escape responses than courting males that were separated from the female when stimulated or than males that failed to initiate courtship (Fig. 3, Table 2). Mating responses were only observed in CS þ Fem males and primarily occurred in response to tactile stimulation of the wings or cerci (Fig. 3). Of the 109 total stimuli applied to the cerci of CS þ Fem males, 41 (38%) resulted in backward slipping while only one stimulus produced hooking. In contrast, the only mating response shown by these males to wing stimulation was hooking, with 57 (60%) of the 95 total stimuli applied to the wings producing this response. We also compared the specific escape responses evoked by tactile stimulation of the wings and cerci. Tactile stimulation of the cerci of CS Fem and No CS þ Fem males produced significantly more forward locomotion (ANOVA: F2,94 ¼ 11.32, P < 0.0001) and kicking (ANOVA: F2,94 ¼ 10.59, P < 0.0001) responses than stimuli applied to the cerci of CS þ Fem males (data not shown). During wing stimulation, the only escape response shown by CS þ Fem males was forward locomotion. In contrast, both CS Fem males and No CS þ Fem males had a tendency to respond with more forward locomotion (ANOVA: F2,93 ¼ 2.68, P ¼ 0.07) and significantly more backward locomotion (ANOVA: F2,93 ¼ 15.73, P < 0.0001) and jumping (ANOVA: F2,93 ¼ 15.22, P < 0.0001) responses to wing stimulation than did CS þ Fem males (data not shown). Experiment 3: Role of the Antennae To determine whether chemical cues detected by the antennae are required to trigger the successful transition from escape to mating behaviour, we removed both antennae from male or female crickets and determined the effect that this antennal ablation had on subsequent mating success (Fig. 1c). For all control and experimental mating pairs that we tested, physical (i.e. chemosensory) contact between the male and female was required for courtship behaviour to be initiated, and we used the first expression of courtship song in conjunction with lowered body posture by the male of each mating pair as our criterion for mating initiation. Visual cues alone were never sufficient to initiate courtship behaviour. Significantly fewer pairs in which the females lacked antennae (FAA) were able to successfully complete all stages of mating when compared to both control pairs (Wald’s chi-square: c21 ¼ 22:39, P < 0.0001) and male antennal-ablated (MAA) pairs (Wald’s chi-square: c21 ¼ 12:01, P ¼ 0.0005). The majority of the 39 FAA pairs that failed to mate did so at the first two stages of mating, with 22 males failing to initiate courtship and 15 pairs failing at backward slipping and mounting (Fig. 1c). The 22 failures at courtship were not due to a lack of contact since the number of contacts made by these unsuccessful pairs (8.7 1.04, range 3e19) was similar to the number made by the 14 successful FAA pairs characterized in Table 3 (10.9 1.6, range 3e25). Instead, eight of the 22 pairs that failed at courtship did so because neither the male nor the female was receptive, eight failed because the female was receptive and approached the male but the male acted aggressively by jerking, chirping aggressively, kicking or running forward in response to each contact KILLIAN ET AL.: BEHAVIOURAL SWITCHING IN CRICKETS 1 Antennae ** CS + Fem 0.75 No CS + Fem CS – Fem 0.5 ** 0.25 0 1 0.75 Meso Legs ** *** 0.5 Proportion of behavioural occurrences 0.25 * 0 1 0.75 Meta Legs ** ** *** *** 0.5 0.25 ** *** 0 1 Wings *** *** 0.75 0.5 0.25 *** 0 1 Cerci 0.75 *** *** *** 0.5 0.25 0 Escape response ** *** Mating response No response Figure 3. Courtship song (CS) and the presence of the female (Fem) are both required for the suppression of escape and activation of mating behavioural responses to tactile stimulation. The touch-evoked responses of male crickets stimulated after the initiation of courtship song and in the presence of the female (CS þ Fem, N ¼ 49) were compared to males that failed to initiate courtship in the presence of the female (No CS þ Fem, N ¼ 10) and to males in which we removed the female immediately following the initiation of male courtship song (CS Fem, N ¼ 40). The proportion (mean 1 SE) of touches resulting in escape behaviours (forward locomotion, backward locomotion, jumping, withdrawal, kicking), mating behaviours (backward slipping, hooking), or in no response is shown for each of five different body locations. *P < 0.05; **P < 0.01; ***P < 0.001. by the female. The remaining six pairs failed because the male initially appeared receptive, following and repeatedly contacting the female with his antennae, but the female was not receptive and ignored the male. As a result, the male never advanced into courtship song and courtship behaviour. An additional 15 of the female antennal-ablated pairs failed at the backward slipping/mounting stage of mating (Fig. 1c). Even though all males reaching this stage performed courtship song, lowered body posture and rhythmic rocking, seven of the 15 pairs that failed did so because the male would kick back or move forward away 495 496 ANIMAL BEHAVIOUR, 72, 2 Table 2. Analysis of effect of a lack of courtship song (CS) or absence of the female (Fem) on the proportion of total male cricket touchevoked responses (ANOVA) No CS þ Fem F Antennae Escape Mating No Resp df 2.17 1,44 d 1,44 2.17 1,44 P CS Fem F df P 0.15 d 0.15 10.69 1,74 0.90 1,74 12.26 1,74 Meso Legs Escape 10.77 1,36 Mating 1.51 1,36 No Resp 9.62 1,36 0.002 0.23 0.004 36.29 1,66 <0.0001 6.17 1,66 0.02 31.86 1,66 <0.0001 Meta Legs Escape Mating No Resp 0.007 d 0.007 16.57 1,75 d 1,75 16.57 1,75 8.05 1,45 d 1,45 8.05 1,45 0.002 0.35 0.001 0.0001 d 0.0001 Wings Escape 53.01 1,54 <0.0001 114.63 1,84 <0.0001 Mating 20.51 1,54 <0.0001 83.11 1,84 <0.0001 No Resp 0.51 1,54 0.48 0.37 1,84 0.55 Cerci Escape 41.63 1,55 <0.0001 Mating 9.18 1,55 0.004 No Resp 2.51 1,55 0.12 46.20 1,85 <0.0001 37.20 1,85 <0.0001 0.01 1,85 0.92 No Resp: no response; Dash: did not occur. Not all CS þ Fem males received the full complement of touches for this comparison. from the female when she approached the male from behind and contacted his cerci or abdomen with her palps. The remaining eight pairs failed because the female did not attempt to mount. Instead, antennal-ablated females often appeared agitated, spending the majority of the trial walking around the perimeter of the arena and attempting to climb up the arena wall. In contrast, mating pairs in which the male was antennal ablated tended to be less successful at mating than control pairs, with 13 of the 49 antennal-ablated males failing to initiate courtship compared to four of 40 Table 3. Effect of male (MAA) or female (FAA) antennal ablation on cricket mating behaviour Control (N¼30) MAA (N¼28) Mating stage duration (s): ContacteCS 79.112.3a 78.918.3a CSeBWS 30.98.5 58.316.4 10.70.8b BWSeHk 9.00.9a HkeTransfer 194.714.6 185.114.2 Total 313.720.4a 332.927.1a Contacts needed 2.30.3a 3.40.5a to evoke CS FAA (N¼14) 243.644.1b 117.055.7 11.40.8b 194.514.2 566.570.0b 10.91.6b All values are means SE. Different superscript letters indicate significant differences between values within a row. CS: courtship song; BWS: backward slipping; Hk: hooking; Total: total time from initial contact to completion of transfer. See text for a complete description of each mating stage. control males (Fig. 1c); however, this difference was not significant (Wald’s chi-square: c21 ¼ 3:53, P ¼ 0.06). Again, these failures were not due to a lack of chemosensory contact, with these 13 pairs making 13.6 1.59 contacts (range 4e23) during the 15-min trial. Of the 13 pairs that failed at courtship, five failed because neither the male nor the female was receptive, whereas the remaining eight failed because the male would kick, run and aggressively chirp and jerk his body when contacted by the female. To examine more closely the influence of antennal sensory inputs in mating success, we compared the duration of each stage of mating for successful control pairs and male antennal-ablated and female antennalablated pairs (Table 3). Female, but not male, antennal ablation significantly increased the amount of time that elapsed between the first physical contact between a mating pair and the initiation of courtship song and courtship behaviour by the male (ANOVA: F2,69 ¼ 7.30, P ¼ 0.001; Table 3), as well as the number of actual physical contacts required to evoke male courtship song (ANOVA: F2,69 ¼ 25.44, P < 0.0001; Table 3). Four primary types of physical contacts would usually occur between the male and female of a mating pair: antennaleantennal (controls only); antennalebody; palpsepalps; palpsebody. The part of the body contacted was usually the wings. Palpebody contact could be initiated by either the male or the female of a mating pair. Control mating pairs primarily relied on the antennae for sex recognition, with 85% of all contacts made involving the antennae and only 15% the palps (N ¼ 67 total contacts for 30 pairs). Male antennal-ablated pairs relied more heavily on chemosensory information from the palps, with 45% of all contacts (N ¼ 93 total contacts for 28 pairs) involving those appendages. In contrast, 72% of all contacts for mating pairs in which the female was antennal ablated (N ¼ 153 contacts for 14 pairs) involved male antennal contact with the female wings, head or legs while only 28% involved the palps, and with the majority of palpal contacts initiated by the female. The majority of contacts were initiated by the intact males of these pairs because the initial response of an antennal-ablated female was to ignore the male while he followed and antennated her as she moved about the arena. Interestingly, the males of pairs in which either the male or the female was antennal ablated tended to show more aggressive responses than did control males, often chirping aggressively, jerking their bodies and kicking when contacted by the female. These displays of aggression, which did not occur as frequently during the mating trials of control pairs, usually ended once courtship behaviour was initiated by the males of the experimental pairs. The mean time that elapsed from the first expression of courtship song/lowered posture by the male to the initiation of male backward slipping and female mounting (CSeBWS, Table 3) was slightly longer and more variable for experimental pairs, but did not differ significantly between the three groups (ANOVA: F2,69 ¼ 2.23, P ¼ 0.12). This variability could be attributed to the failure by several pairs at their first attempt at backward slipping either because the female overran the male during mounting, or because she was improperly positioned at the start of KILLIAN ET AL.: BEHAVIOURAL SWITCHING IN CRICKETS mounting, causing the male to move forward as the female contacted his cerci, abdomen or leg with her palps. Both antennal-ablated males and males attempting to mate with an antennal-ablated female took significantly longer to successfully attach, or hook, their epiphallus to the female subgenital plate than did control males (BWSeHooking; ANOVA: F2,69 ¼ 7.03, P ¼ 0.002; Table 3). At the start of backward slipping, intact males held both antennae directed posteriorly towards the female. As the female mounted and walked forward over the male, the male would move his antennae forward until the female’s head was positioned almost directly above his head. Many antennal-ablated males, however, had difficulty in hooking because females would advance too far forward. For female antennal-ablated pairs, more male hooking attempts were also necessary because these females tended to be misaligned and misplaced laterally on the back of the male during mounting. However, once the male’s epiphallus was successfully hooked, the time it took for transfer of the spermatophore (HookingeTransfer; Table 3) did not differ significantly between the three groups (ANOVA: F2,69 ¼ 0.37, P ¼ 0.69) even though many of these experimental pairs remained misaligned. Finally, when total duration of trials was compared, only pairs in which the female was antennal ablated took significantly longer to successfully mate (ANOVA: F2,69 ¼ 9.78, P ¼ 0.0002; Table 3). DISCUSSION Touch-evoked Escape Behaviours Tactile stimulation of the antennae, legs, wings and cerci of nonmating male A. domesticus crickets evoked a primary behavioural response when each body part was touched. Responses could include simple avoidance reflexes such as withdrawal of an appendage away from the stimulus source, or more complex evasive responses such as running, jumping or kicking. For example, isolated male crickets primarily responded to mesothoracic leg stimulation with withdrawal of the leg or body and to metathoracic leg stimulation with forward locomotion or jumping. Insect legs are covered with touch-sensitive hairs (Hustert 1985; Newland 1991), and activation of such hairs probably triggered these reactions. Most stimuli applied to the wings elicited jumping, as previously reported for the cricket Gryllus bimaculatus (Hiraguchi & Yamaguchi 2000). Small mechanosensory hairs on the hindwings triggers this jumping (Hiraguchi et al. 2003) and similar hairs are located on the basal region of the cricket forewings (Fudalewicz-Niemczyk & Rosciszewska 1972), the point at which we stimulated the wings in our experiments. We found that most paintbrush stimuli applied to the cricket antenna evoked no discernible escape response. Tactile stimulation of the antennal flagellum of the cricket often produces a lateral movement of the flagellum away from the stimulus (Balakrishnan & Pollack 1997), however, we did not include such movements in our counts of evasive withdrawal responses. The cockroach Periplaneta americana, in contrast, produces a fairly robust evasive response when the antennae are touched (Comer et al. 1994). However, initiation of these responses may require the activation of mechanoreceptors located in the basal segments of the cockroach antennae rather than on the flagellum (Comer et al. 2003). Similar receptors have been identified in the cricket (Kammerer & Honegger 1988; Gebhardt & Honegger 2001), however, the low frequency of evasive responses observed in our animals suggests that tactile stimulation of the midregion of the antennae usually does not activate these basal receptors. However, it is also possible that our animals responded with an unusually low number of evasive responses to antennal stimulation because they were taken from populations that had been cultured at high densities over several generations. When we mechanically stimulated the cerci of male crickets, the primary responses observed were kicking or forward locomotion. Campaniform sensilla activated during tactile stimulation of the cerci are scattered over the surface of each cercus (Killian et al. 1993) and can trigger kicking of the cricket’s metathoracic legs (Dumpert & Gnatzy 1977). Cercal campaniform sensilla are also important during mating because they trigger male backward slipping (Sakai & Ootsubo 1988) and provide sensory feedback to the male on female body position during both hooking and spermatophore transfer (Sakai & Ootsubo 1988; Snell & Killian 2000). Suppression of Escape Behaviours During Mating Following mounting of the female and hooking of the genitalia, both the male and female become relatively rigid and immobile. All body movements cease except for slight movements of the male abdomen, genitalia and cerci associated with spermatophore transfer. We found that male crickets showed a lack of responsiveness to touch when stimulated during this stage of mating. In addition, many males showed a significant decrease in jumping responses to touch when stimulated immediately following copulation. We do not think that habituation from repetitive tactile stimulation is responsible for these observed changes since these same males showed robust evasive responses when repetitively stimulated in isolation, before their mating trials. What produces the immobility and almost complete suppression of touch-evoked responses during copulation? One possibility is that sensory feedback arising from the position of the animals’ limbs or from the male-to-female contact occurring during copulation plays a role. For example, by allowing the antennae of the cockroach P. americana to contact a mechanical barrier, an immobile or ‘quiescent’ state can be induced in the animal during which escape running responses normally triggered by touch or wind are suppressed (Watson & Ritzmann 1994). Interestingly, Watson & Ritzmann (1994) found that quiescent roaches initiate antennal waving immediately prior to leaving this immobile state and regaining their responsiveness to stimulation. Similarly, copulating A. domesticus crickets remain quiescent until spermatophore transfer is complete and begin antennal waving just prior to unhooking of the genitalia and female 497 498 ANIMAL BEHAVIOUR, 72, 2 dismount (personal observation). A similar immobile, unresponsive state can be induced in the cricket G. bimaculatus. This immobile state, or thanatosis, usually lasts 2e4 min and is initiated by mechanical restraint of the limbs (Nishino & Sakai 1996; Nishino 2004). Interestingly, the duration of this mechanically induced thanatosis is similar to the duration of spermatophore transfer reported for A. domesticus mating pairs (Snell & Killian 2000). A second possibility is that the release of neurohormones or neuromodulators within the central nervous system following male and female antennal contact is responsible for this widespread change in the efficacy of escape responses. For example, in the cricket G. bimaculatus, haemolymph levels of the biogenic amine octopamine increase during courtship, and antennal contact with a conspecific is necessary to trigger this increase (Adamo et al. 1995). Biogenic amines may also regulate male copulatory behaviour during mating (Matsumoto & Sakai 2001), as well as the duration of male guarding behaviour following mating (Ureshi et al. 2002). In addition, the escape behaviours of both crickets (Stevenson et al. 2000) and cockroachs (Goldstein & Camhi 1991; Casagrand & Ritzmann 1992; Hill & Blagburn 2001) can be modulated by these amines. We also found that tactile stimulation of males 15e60 min before the initiation of a mating trial negatively impacted mating success. Failures were either due to a failure of the male to initiate courtship or because the male showed escape responses when contacted by the female. We suggest that repetitive tactile stimulation of these males just before their mating trials induced the release of neuromodulators that may have acted to inhibit the switch from escape to mating behaviour. Mechanical handling stress can increase haemolymph octopamine levels in crickets (Woodring et al. 1988), locusts and cockroaches (Davenport & Evans 1984), and octopamine is called the ‘fight or flight’ hormone of insects (reviewed in Roeder 1999). Interestingly, when males were stimulated approximately 15 min before mating, but immediately after female contact and the initiation of courtship (experiment 2, CS Fem pairs), this decrease in mating success was not observed. This suggests that prior chemosensory contact with a female could act to prevent any stress-induced enhancement of escape behaviour elicited by repetitive tactile stimulation. We also observed a decrease in the jumping responses of males immediately following copulation and during the onset of male guarding behaviour. During guarding, the male attempts to remain in close contact with the female, will often rest his antennae across the female’s body, and will jerk his body and chirp aggressively in response to any movements made by the female (reviewed in Zuk & Simmons 1997). A decrease in male jumping responses to touch during guarding may facilitate these interactions. Activation of the Switch from Escape to Mating Behaviour We asked whether touch-evoked escape responses were inhibited during the period between courtship initiation and copulation. Contact of the female mouthparts, or palps, with the male’s cerci or abdomen is needed to initiate male backward slipping, and the female must also mount onto the male’s back and wings. Jumping and kicking responses to wing or cercal stimulation would thus need to be suppressed before these events could take place. In addition, Huber (1965) reported that isolated G. campestris males primarily kick during tactile stimulation of the cerci, but that both courting and guarding males respond to cercal stimulation with copulatory movements. Accordingly, we asked whether the switch from escape to mating behaviour is triggered during the initial period of sex recognition, when the male and female first make chemosensory contact. First, our results confirmed the importance of male courtship behaviour since all males that failed to initiate courtship song even after 15 min of interaction with a female (No CS þ Fem pairs) failed to copulate. Male courtship song is a prerequisite for the initiation of female receptivity and mounting behaviour in A. domesticus (Crankshaw 1979; Nelson & Nolen 1997) as well as other cricket species (Loher & Rence 1978; Adamo & Hoy 1994; Balakrishnan & Pollack 1996). In addition, we found that the touch-evoked escape responses of nonsinging males stimulated in the presence of the female were not suppressed, indicating that these males had failed to successfully make the transition from escape to mating behaviour. We also found that only courting males in constant contact with a female (CS þ Fem) would produce copulatory responses to tactile stimulation of the wings and cerci. They also produced significantly fewer kicking and jumping responses than did courting males stimulated in the absence of the female (CS Fem). These results suggest that not only is male and female contact required to initiate courtship, but that continued contact between the male and female is necessary for mating to progress from courtship to copulation. Our results conflict with those of Sakai & Ootsubo (1988), who reported that courting G. bimaculatus males isolated from females would produce backward slipping and hooking movements when mechanically stimulated. However, these researchers did not indicate how long their mating pairs interacted before separation and stimulation of the male. In our study, each CS Fem male was separated from the female by an opaque divider immediately after the initiation of courtship song by the male, whereas CS þ Fem males were in constant contact with the female during the entire stimulation period. This continuous female contact may have provided the reinforcing stimulus needed to allow those males to progress into copulatory behaviour. In support of this, Matsumoto & Sakai (2000) reported that the number of courting males that responded with copulatory movements to tactile stimulation of the cerci increased over time from 0% (when males were stimulated immediately at the onset of courtship and following separation from the female) to 30% (when males were stimulated after 15 min of interaction with a female). The reinforcing stimuli experienced by the CS þ Fem males may have been visual, chemical or tactile cues provided by the females. Hardy & Shaw (1983) found that KILLIAN ET AL.: BEHAVIOURAL SWITCHING IN CRICKETS upon initial antennal contact, an A. domesticus male will respond aggressively towards a female and that continued contact is required for sex discrimination. These authors suggested that males require additional information other than contact chemoreception, perhaps visual cues supplied by the sexually receptive female’s body position or movements, to initiate courtship (Hardy & Shaw 1983). These cues may be derived from the female antennae since we found that female antennal ablation can negatively affect male receptivity. Similarly, sexually receptive G. bimaculatus males given the opportunity to contact and court anaesthetized females will usually ignore them (Adamo & Hoy 1994). However, blinded and deafened Teleogryllus commodus males had no difficulty in courting and copulating with females (Loher & Rence 1978) and Teleogryllus oceanicus males readily initiated courtship when allowed to contact anaesthetized females with their antennae (Balakrishnan & Pollack 1997). Role of the Antennae We found that removal of the female antennae significantly decreased mating success, whereas male antennal ablation had little effect. For all mating trials, antennal contact was required for sex recognition and the initiation of courtship song and courtship behaviour by the male. Visual cues alone were not sufficient and we found no evidence during our experiments that males or females responded to airborne pheromones. Similarly, male and female antennal contact is the primary means of sex recognition for T. commodus (Loher & Rence 1978), G. bimaculatus (Adamo & Hoy 1994) and T. oceanicus (Balakrishnan & Pollack 1997) crickets. This contact activates chemoreceptors on the antennae, allowing the male to confirm the presence of a female and providing the signal that initiates courtship behaviour (Rence & Loher 1977; Hardy & Shaw 1983; Balakrishnan & Pollack 1997; Tregenza & Wedell 1997). Adamo & Hoy (1994) also reported that removal of a male’s antennae 3 days before pairing with intact females had no effect on the mating success of G. bimaculatus crickets. In contrast, Murakami & Itoh (2003) recently reported a significant decrease in the number of G. bimaculatus males that would court intact females 7 days after male antennal removal, and they suggested that the longer elapsed time between antennal ablation and their behavioural tests may explain these differing results. Loher & Rence (1978) also found that T. commodus males paired with intact females 10 days after surgical removal of their antennae failed to initiate courtship song and instead either ignored or acted aggressively towards females. Even though our A. domesticus males had their antennae removed 9e20 days before behavioural trials, their courtship behaviour was not as severely affected, suggesting that antennal inputs may be of primary importance for sex recognition and courtship initiation for the males of some, but not all, species of crickets. Chemosensory contact is crucial for the initiation of male courtship behaviour. Nevertheless, the overall success rate of mating pairs in which the males had their antennae removed was not significantly different from that of control pairs. Pairs that did fail to mate did so because the antennal-ablated male failed at courtship behaviour. Instead, many of these males chirped aggressively, jerked their bodies, kicked or ran when contacted by a female. These behaviours were less prevalent in control males and most, if not all, aggressive responses ended once courtship behaviour was initiated. Similarly, deantennated G. bimaculatus males show increased levels of aggression towards females (Adamo & Hoy 1994). The fact that we saw no significant effect of male antennal ablation on mating success, or on the time or number of contacts required to initiate courtship for successful pairs suggests that chemosensory cues derived from the maxillary palps could be sufficient to initiate courtship behaviour from antennaeless males. Adamo & Hoy (1994) reported that G. bimaculatus males and females also increase the number of contacts they make with their palps following removal of the antennae. Klein (1981), who described the various sensilla covering the tips of the maxillary palps of the cricket, concluded that most palpal sensilla function as contact chemoreceptors. In the cricket brain, both maxillary palp afferents (Ignell et al. 2000) and antennal afferents (Staudacher & Schildberger 1999) terminate within the deutocerebrum. This anatomical arrangement, together with our observation that maxillary palp information can initiate courtship behaviour in the absence of antennal inputs, suggests that afferents from both pairs of appendages may be involved in the activation of common neural circuits within the brain, possibly at the level of higher integrative centres such as the mushroom bodies (Frambach & Schurmann 2004). Removal of the female antennae negatively affected the ability of pairs to copulate successfully. Interestingly, more males failed to initiate courtship behaviour when the female lacked antennae and the male was intact than when the male was antennal ablated and the female was intact. In addition, for the 14 female antennal-ablated pairs that were successful, it took more than five times as many contacts between the male and female to initiate male courtship, resulting in a significant increase in the time that elapsed between the first maleefemale contact and the initiation of courtship. In contrast, the number of contacts and time required to initiate the courtship behaviour of antennal-ablated males were similar to those of intact males. These findings again suggest that visual and/or tactile cues provided by the female antennae are important in the expression of male courtship behaviour. Female antennal ablation also significantly decreased the number of pairs successfully completing male backward slipping and female mounting. About half of these failures were because the female never became sexually receptive and instead ignored the courting male throughout the 15-min trial. Antennal ablation also significantly decreases the mounting responses of T. commodus (Loher & Rence 1978) and T. oceanicus (Balakrishnan & Pollack 1997) females, however, these failures were apparently not due to a lack of interest by the female. The remainder of the female antennal-ablated pairs that failed at backward slipping did so because the male would kick, or run forward, when the female contacted his cerci 499 500 ANIMAL BEHAVIOUR, 72, 2 or abdomen. Females normally use their palps to contact the male’s cerci and use their antennae to contact the male’s wings and abdomen, and this contact elicits male backward slipping and female mounting behaviour (Loher & Rence 1978; Sakai & Ootsubo 1988). As the female mounts, the male swings his antennae back and forth in front and behind him, contacting the female’s head and antennae while she holds her antennae out in front of her (Alexander 1961; Loher & Rence 1978; Adamo & Hoy 1994). We have also observed that when a female cricket approaches a male from behind and contacts his cerci with her palps, he will sometimes move forward a few steps and then immediately resume courtship, perhaps indicating that the female was not in a proper position for mounting. Males showed an increase in such responses when paired with a deantennated female and, when successful mounting did occur for such pairs, the females were often misaligned. The importance of antennal sensory feedback during mounting is also indicated by the significantly longer time required for hooking of the genitalia by both the male and the female antennal-ablated pairs. In summary, we have demonstrated that male touchevoked escape behaviours are suppressed during mating. Mating and escape are incompatible behaviours and understanding the factors controlling their production can provide insight into the process of this behavioural switching. The switch from escape to mating occurs following antennal contact with a female and requires the continued presence of the female for its full expression. The chemosensory information responsible for triggering this switch in the male is primarily supplied by the antennae, although information derived from maxillary palp chemoreceptors may also be important. However, loss of the female antennae had a significant negative effect on both female and male receptivity and mounting behaviour. Future work needs to focus on the underlying neural mechanisms mediating these behavioural changes. Acknowledgments We thank Michael Hughes, Manager, Miami University Statistical Consulting Center for his expert advice and statistical assistance. We thank Rachel Orr, Christa Nagel, Jace Perkerson and Erin Wernke for their technical assistance in the laboratory and Kelly Pogorzelski and Michael Sunderman for maintaining our cricket colony. Financial support was provided by a grant from the National Institute of Mental Health (R15 MH6060701A1) and a Miami University Committee for Faculty Research grant. References Adamo, S. A. & Hoy, R. R. 1994. Mating behaviour of the field cricket Gryllus bimaculatus and its dependence on social and environmental cues. Animal Behaviour, 47, 857e868. Adamo, S. A., Linn, C. E. & Hoy, R. R. 1995. The role of neurohormonal octopamine during ‘fight or flight’ behaviour in the field cricket Gryllus bimaculatus. Journal of Experimental Biology, 198, 1691e1700. Alexander, R. D. 1961. Aggressiveness, territoriality, and sexual behavior in field crickets (Orthoptera: Gryllidae). Behaviour, 17, 130e223. Balakrishnan, R. & Pollack, G. S. 1996. Recognition of courtship song in the field cricket, Teleogryllus oceanicus. Animal Behaviour, 51, 353e366. Balakrishnan, R. & Pollack, G. S. 1997. The role of antennal sensory cues in female responses to courting males in the cricket Teleogryllus oceanicus. Journal of Experimental Biology, 200, 511e522. Casagrand, J. L. & Ritzmann, R. E. 1992. Biogenic amines modulate synaptic transmission between identified giant interneurons and thoracic interneurons in the escape system of the cockroach. Journal of Neurobiology, 23, 644e655. Comer, C. M., Mara, E., Murphy, K. A., Getman, M. & Mungy, M. C. 1994. Multisensory control of escape in the cockroach Periplaneta americana. II. Patterns of touch-evoked behavior. Journal of Comparative Physiology, 174, 13e26. Comer, C. M., Parks, L. & Halvorsen, M. B. 2003. The antennal system and cockroach evasive behavior. II. Stimulus identification and localization are separable antennal functions. Journal of Comparative Physiology, 189, 97e103. Crankshaw, O. S. 1979. Female choice in relation to calling and courtship songs in Acheta domesticus. Animal Behaviour, 27, 1274e1275. Davenport, A. P. & Evans, P. D. 1984. Stress-induced changes in the octopamine levels of insect hemolymph. Insect Biochemistry, 14, 135e143. Davis, W. J. 1979. Behavioral hierarchies. Trends in Neurosciences, 2, 5e7. Davis, W. J., Mpitsos, G. J. & Pinneo, J. M. 1974a. The behavioral hierarchy of the mollusk Pleurobranchaea. I. The dominant position of the feeding behavior. Journal of Comparative Physiology, 90, 207e224. Davis, W. J., Mpitsos, G. J. & Pinneo, J. M. 1974b. The behavioral hierarchy of the mollusk Pleurobranchaea. II. Hormonal suppression of feeding associated with egg-laying. Journal of Comparative Physiology, 90, 225e243. Davis, W. J., Mpitsos, G. J., Pinneo, J. M. & Ram, J. L. 1977. Modification of the behavioural hierarchy of Pleurobranchaea. I. Satiation and feeding motivation. Journal of Comparative Physiology, 117, 99e125. Dumpert, K. & Gnatzy, W. 1977. Cricket combined mechanoreceptors and kicking response. Journal of Comparative Physiology, 122, 9e25. Esch, T. & Kristan, W. B. 2002. Decision-making in the leech nervous system. Integrative and Computational Biology, 42, 716e724. Frambach, I. & Schurmann, F. W. 2004. Separate distribution of deutocerebral projection neurons in the mushroom bodies of the cricket brain. Acta Biologica Hungarica, 55, 21e29. Fudalewicz-Niemczyk, W. & Rosciszewska, M. 1972. The innervation and sense organs of the wings of Gryllus domesticus L. (Orthoptera). Acta Biologica Cracoviensia, 15, 35e51. Gebhardt, M. & Honegger, H.-W. 2001. Physiological characterization of antennal mechanosensory descending interneurons in an insect (Gryllus bimaculatus, Gryllus campestris) brain. Journal of Experimental Biology, 204, 2265e2275. Gillette, R., Huang, R.-C., Hatcher, N. & Moroz, L. L. 2000. Costebenefit analysis potential in feeding behavior of a predatory snail by integration of hunger, taste, and pain. Proceedings of the National Academy of Sciences, U.S.A., 97, 3585e3590. Gnatzy, W. & Heusslein, R. 1986. Digger wasp against crickets. I. Receptors involved in the antipredator strategies of the prey. Naturwissenschaften, 73, 212e214. Goldstein, R. S. & Camhi, J. M. 1991. Different effects of the biogenic amines dopamine, serotonin and octopamine on the KILLIAN ET AL.: BEHAVIOURAL SWITCHING IN CRICKETS thoracic and abdominal portions of the escape circuit in the cockroach. Journal of Comparative Physiology, 168, 103e112. Gras, H. & Horner, M. 1992. Wind-evoked escape running of the cricket Gryllus bimaculatus. I. Behavioural analysis. Journal of Experimental Biology, 171, 189e214. Hardy, T. N. & Shaw, K. C. 1983. The role of chemoreception in sex recognition by male crickets: Acheta domesticus and Teleogryllus oceanicus. Physiological Entomology, 8, 151e166. Hill, E. S. & Blagburn, J. M. 2001. Presynaptic effects of biogenic amines modulating synaptic transmission between identified sensory neurons and giant interneurons in the first instar cockroach. Journal of Comparative Physiology, 187, 633e645. Hiraguchi, T. & Yamaguchi, T. 2000. Escape behavior in response to mechanical stimulation of hindwing in cricket, Gryllus bimaculatus. Journal of Insect Physiology, 46, 1331e1340. Hiraguchi, T., Yamaguchi, T. & Takahata, M. 2003. Mechanoreceptors involved in the hind-wing evoked escape behaviour in cricket, Gryllus bimaculatus. Journal of Experimental Biology, 206, 523e534. Huang, Z. & Satterlie, R. A. 1990. Neuronal mechanisms underlying behavioral switching in a pteropod mollusk. Journal of Comparative Physiology, 166, 875e887. Huber, F. 1965. Brain controlled behaviour in orthopterans. In: The Physiology of the Insect Central Nervous System (Ed. by J. E. Treherne & J. W. L. Beament), pp. 233e246. London: Academic Press. Hustert, R. 1985. Multisegmental integrations and divergence of afferent information from single tactile hairs in a cricket. Journal of Experimental Biology, 118, 209e227. Hustert, R. & Gnatzy, W. 1995. The motor program for defensive kicking in crickets: performance and neural control. Journal of Experimental Biology, 198, 1275e1283. Ignell, R., Anton, S. & Hansson, B. S. 2000. The maxillary palp sensory pathway of Orthoptera. Arthropod Structure and Development, 29, 295e305. Jing, J. & Gillette, R. 1995. Neuronal elements that mediate escape swimming and suppress feeding behavior in the predatory sea slug Pleurobranchaea. Journal of Neurophysiology, 74, 1900e1910. Jing, J. & Gillette, R. 2000. Escape swim network interneurons have diverse roles in behavioral switching and putative arousal in Pleurobranchaea. Journal of Neurophysiology, 83, 1346e1355. Kammerer, R. & Honegger, H.-W. 1988. The role of mechanoreceptors in the control of antennal tracking movements of crickets. In: Sense Organs, Interface between Environment and Behaviour (Ed. by N. Elsner & F. Barth), pp. 20. Stuttgart: Georg Thieme Verlag. Killian, K. A., Merritt, D. J. & Murphey, R. K. 1993. Transplantation of neurons reveals processing areas and rules for synaptic connectivity in the cricket nervous system. Journal of Neurobiology, 24, 1187e1206. Klein, U. 1981. Sensilla of the cricket palp. Fine structure and spatial organization. Cell and Tissue Research, 219, 253e266. Kovac, M. P. & Davis, W. J. 1980. Neural mechanism underlying behavioral choice in Pleurobranchaea. Journal of Neurophysiology, 43, 469e487. Krasne, F. B. & Lee, S. S. 1988. Response-dedicated trigger neurons as control points for behavioral actions: selective inhibition of lateral giant command neurons during feeding in crayfish. Journal of Neuroscience, 8, 3703e3712. Kristan, W. B. & Shaw, B. K. 1997. Population coding and behavioral choice. Current Opinion in Neurobiology, 7, 826e831. Loher, W. & Dambach, M. 1989. Reproductive behavior. In: Cricket Behavior and Neurobiology (Ed. by F. Huber, T. E. Moore & W. Loher), pp. 43e82. Ithaca, New York: Cornell University Press. Loher, W. & Rence, B. 1978. The mating behavior of Teleogryllus commodus (Walker) and its central and peripheral control. Zeitschrift für Tierpsychologie, 46, 225e259. Matsumoto, Y. & Sakai, M. 2000. Brain control of mating behavior in the male cricket Gryllus bimaculatus De Geer: the center for inhibition of copulatory actions. Journal of Insect Physiology, 46, 527e538. Matsumoto, Y. & Sakai, M. 2001. Brain control of mating behavior in the male cricket Gryllus bimaculatus De Geer: excitatory control of copulatory actions. Zoological Science, 18, 659e669. Misell, L. M., Shaw, B. K. & Kristan, W. B. 1998. Behavioral hierarchy in the medicinal leech, Hirudo medicinalis: feeding as a dominant behavior. Behavioural Brain Research, 90, 13e21. Morton, D. W. & Chiel, H. J. 1994. Neural architectures for adaptive behavior. Trends in Neuroscience, 17, 413e420. Murakami, S. & Itoh, M. T. 2003. Removal of both antennae influences the courtship and aggressive behaviors in male crickets. Journal of Neurobiology, 57, 110e118. Myers, R. H., Montgomery, D. C. & Vining, G. G. 2002. Generalized Linear Models. New York: J. Wiley. Nelson, C. M. & Nolen, T. G. 1997. Courtship song, male agonistic encounters, and female mate choice in the house cricket, Acheta domesticus (Orthoptera: Gryllidae). Journal of Insect Behavior, 10, 557e570. Neter, J., Kutner, M. H., Nachtsheim, C. J. & Wassermann, W. 1996. Applied Linear Statistical Models. 4th edn. Chicago: Irwin. Newland, P. L. 1991. Physiological properties of afferents from tactile hairs on the hindlegs of the locust. Journal of Experimental Biology, 155, 487e503. Nishino, H. 2004. Motor output characterizing thanatosis in the cricket Gryllus bimaculatus. Journal of Experimental Biology, 207, 3899e3915. Nishino, H. & Sakai, M. 1996. Behaviorally significant immobile state of so-called thanatosis in the cricket Gryllus bimaculatus DeGeer: its characterization, sensory mechanism and function. Journal of Comparative Physiology, 179, 613e624. Norekian, T. P. & Satterlie, R. A. 1996. Whole body withdrawal circuit and its involvement in the behavioral hierarchy of the mollusk Clione limacina. Journal of Neurophysiology, 75, 529e537. Nowosielski, J. W. & Patton, R. L. 1963. Studies on circadian rhythm of the house cricket, Gryllus domesticus L. Journal of Insect Physiology, 9, 401e410. Rence, B. & Loher, W. 1977. Contact chemoreceptive sex recognition in the male cricket, Teleogryllus commodus. Physiological Entomology, 2, 225e236. Roeder, T. 1999. Octopamine in invertebrates. Progress in Neurobiology, 59, 533e561. Sakai, M. & Ootsubo, T. 1988. Mechanism of execution of sequential motor acts during copulation behavior in the male cricket Gryllus bimaculatus DeGeer. Journal of Comparative Physiology, 162, 589e600. Snell, L. C. & Killian, K. A. 2000. The role of cercal sensory feedback during spermatophore transfer in the cricket Acheta domesticus. Journal of Insect Physiology, 46, 1017e1032. Staudacher, E. & Schildberger, K. 1999. A newly described neuropile in the deutocerebrum of the cricket: antennal afferents and descending interneurons. Zoology Analysis of Complex Systems, 102, 212e226. Stevenson, P. A., Hofmann, H. A., Schoch, K. & Schildberger, K. 2000. The fight and flight responses of crickets depleted of biogenic amines. Journal of Neurobiology, 43, 107e120. Tinbergen, N. 1950. The hierarchical organization of nervous mechanisms underlying instinctive behaviour. Symposium of the Society for Experimental Biology, 4, 305e312. 501 502 ANIMAL BEHAVIOUR, 72, 2 Tinbergen, N. 1951. The Study of Instinct. Oxford: Clarendon. Tregenza, T. & Wedell, N. 1997. Definitive evidence for cuticular pheromones in a cricket. Animal Behaviour, 54, 979e984. Ureshi, M., Dainobu, M. & Sakai, M. 2002. Serotonin precursor (5-hydroxytryptophan) has a profound effect on the postcopulatory time-fixed sexually refractory stage in the male cricket, Gryllus bimaculatus DeGeer. Journal of Comparative Physiology, 188, 767e779. Watson, J. T. & Ritzmann, R. E. 1994. The escape response versus the quiescent response of the American cockroach: behavioural choice mediated by physiological state. Animal Behaviour, 48, 476e478. Woodring, J. P., Ouida, W. M. & Rose, E. 1988. Effect of development, photoperiod, and stress on octopamine levels in the house cricket, Acheta domesticus. Journal of Insect Physiology, 34, 759e765. Zuk, M. & Simmons, L. W. 1997. Reproductive strategies of the crickets (Orthoptera: Gryllidae). In: The Evolution of Mating Systems in Insects and Arachnids (Ed. by J. C. Choe & B. J. Crespi), pp. 89e109. Cambridge: Cambridge University Press.
© Copyright 2026 Paperzz